Jump to ContentJump to Main Navigation
Public policy analysis$

Professor Peter Knoepfel

Print publication date: 2011

Print ISBN-13: 9781861349071

Published to Policy Press Scholarship Online: March 2012

DOI: 10.1332/policypress/9781861349071.001.0001

Show Summary Details

Public policy

Public policy

(p.20) (p.21) Two Public policy
Public policy analysis

Peter Knoepfel

Corinne Larrue

Frédéric Varone

Michael Hill

Policy Press

Abstract and Keywords

This chapter defines the constituent elements of a policy. It refers to public policy as the power games in a specific institutional context played out between various public actors who make a concerted effort to resolve a collective problem in collaboration with or in opposition to para-state and private actors. It adds that public policy is a large number of legislative and administrative activities aimed at the resolution of real problems. The constituent elements of public policy comprise a solution to a public problem, the existence of target groups at the root of a public problem, intentional coherence at the very least, the existence of several decisions and activities, the intervention programme, the key role of public actors, the existence of formalised measures, and the decisions and activities that impose constraints. Lastly, the chapter explains the policy cycle.

Keywords:   public policy, public actors, public problem, policy cycle, private actors, para-state actors

2.1 Policy as a response to social problems

All policies aim to resolve a public problem that is identified as such on the governmental agenda. Thus, they represent the response of the political-administrative system to a social reality that is deemed politically unacceptable.

It should be noted here that it is the symptoms of a social problem that constitute the starting point for the realisation of its existence and of a debate on the need for a policy (for example, decline in the state of forests, drug-associated delinquency, high unemployment). At the initial stage of all public intervention, the actual causes of the collective problem have not yet been defined with certainty or defined consensually by public and private actors. The increase in unemployment levels in industrialised countries and the material precariousness of unemployed people prompt the state to create or revise its unemployment benefit system and to take measures to revitalise the labour market. Air pollution arising from industrial production and the consumption of fossil fuels prompts the state to develop an environmental protection policy. Urban criminality and the deterioration of the physical state of drug addicts are the triggers for new policies on the distribution of heroin under medical supervision. Although this interpretation of policies as institutional responses to (changing) social states that are deemed problematic is dominant within policy analysis, this assumption must be relativised.

Firstly, some instances of social change do not give rise to policies, mainly because they are not visible or expressed (for example, non-visibility of consequences, long-term consequences only, lack of political representation of the disadvantaged groups), or because no mode of state intervention proves feasible and consensual (for example, negative electoral impacts, absence of political-administrative implementation bodies, the inability to influence the behaviour of certain private actors in reality). Thus, the pluralist vision whereby the ‘service hatch’ state responds in an egalitarian and automatic fashion to all ‘social demands’ must be rejected.

(p.22) This point raises questions about the ways in which social problems are defined (Dery, 1984; Weiss, 1989), their thematicisation on the governmental agenda (Kingdon, 1984, 1995; Rochefort and Cobb, 1993), the definition of target groups and the eventual decision not to get involved or apply a collective solution (Bachrach and Baratz, 1963). Numerous filtering processes exist at all of these levels and they represent opportunities for organised actors who oppose the political recognition of an instance of social change to keep the latter outside the political-administrative arena.

Secondly, certain policies may be interpreted, not as collective actions aimed at resolving a social problem (adaptation to or anticipation of a social change), but as simple instruments for the exercising of power and domination by one social group over another. As explained in Chapter One, the neo-Marxist authors believe that state policies aim solely to reproduce or emphasise the divides between the social classes. The neo-Weberian school supposes that state intervention can only enable the satisfaction of the internal interests of bureaucratic actors (administrative inertia). The theory of rational choice defines policy as the (re)distribution of the costs and benefits between the electoral groups in exchange for votes and/or partisan support. Seen from this perspective, substantive policies would be merely currency exchanged in electoral competitions. Finally, the neo-corporatist approach believes that policies protect the interests of organised groups who are able to ‘capture’ the political-administrative institutions and establish clientelist relationships with them.

As stated in the first section of this book, our position lies somewhere between these two extreme visions of a neutral ‘service hatch’ state that is attentive to all social demands, on the one hand, and a ‘captive’ state manipulated by an organised group, on the other. Viewed from this perspective, public policies emerge as a response to a public problem that reflects a social state (in transformation), which has been articulated by mediators (for example, the media, new social movements, political parties and/or interest groups) and then debated within the democratic decision-making process (Muller, 1990). This does not take away from the fact that the problem to be resolved is a social and political construct, even in the case of shock events – for example, the nuclear disaster at Chernobyl (Czada, 1991) or the effects of mad-cow disease (BSE) on humans (Greer, 1999) – because it always depends on the perceptions, representations, interests and resources of different public and private actors (Vlassopoulou, 1999).

There is no linear and mechanical institutional response that would be a function of the objective pressure of a collective problem; this is (p.23) always a redistributive exercise, a ‘mobilisation of bias’ (Schattschneider, 1960). Examples of this, which are already well substantiated by empirical studies, include the insufficient police presence in Swiss suburbs with high immigrant populations, which are one of the locations of urban violence, and the priority implementation of traffic-calming measures in residential areas as opposed to working-class areas directly affected by noise and the hazards of road traffic (see in particular Terribilini, 1995, 1999).

2.2 An analytical definition

We have seen that the essential object of policy analysis is not political power in itself, but its use for the purpose of resolving collective problems. Thus, the notion of public policy refers to the power games in a specific institutional context played out between various public actors who make a concerted effort to resolve a collective problem in collaboration with or in opposition to para-state and private actors. Given that these problems are connected with specific areas or sectors, the term ‘policy’, which has been established in common parlance since the growth of state intervention in the 1930s, is often qualified with the name of the sector or area in question (for example, ‘energy policy’, ‘agricultural policy’, ‘economic policy’, ‘social policy’).

There are numerous definitions of the concept of public policy. Thoenig lists at least 40 in the introduction to his 1980 analysis of public policies (Thoenig, 1985, p 3). Without replicating this list in its entirety, we note some of these definitions below, ranging from vagueness at one extreme to efforts to be very precise at the other:

  • “Public policy is whatever governments choose to do or not to do” (Dye, 1972, p 18).

  • “A public policy is the product of the activity of an authority invested with public power and governmental legitimacy” (Mény and Thoenig, 1989, p 129).

  • “A public policy is a programme of action specific to one or more public or governmental authorities within a sector of society or a given area” (Thoenig, 1985, p 6; Mény and Thoenig, 1989, p 130).

  • “A public policy is the product of activities aimed at the resolution of public problems in the environment by political actors whose relationships are structured. The entire process evolves over time” (Lemieux, 1995, p 7).

(p.24) Despite being very different, these definitions tend to emphasise the idea of actors invested with public power (see in particular Dye, Mény and Thoenig cited above; also Sharkansky, 1970, p 1; Heclo, 1972, p 85; Simeon, 1976, p 548), collective problems requiring resolution (see Anderson, 1984, p 3; Pal, 1992, p 2) and on the state solutions provided (see Laswell and Kaplan, 1950, p 71; Jenkins, 1978, p 15).

Ultimately, policy experts agree that an ‘operational’ definition is necessary to qualify the object and field of study in this area (see Muller, 1990, p 24). The definition proposed below, which encompasses the main elements on which consensus exists in the literature (see specifically Thoenig, 1985, p 7; Mény and Thoenig, 1989, pp 131–2; Lagroye, 1997, p 454), is rooted in this perspective.

Thus, within the framework of the approach proposed here, a public policy is defined as a series of intentionally coherent decisions or activities taken or carried out by different public – and sometimes – private actors, whose resources, institutional links and interests vary, with a view to resolving in a targeted manner a problem that is politically defined as collective in nature. This group of decisions and activities gives rise to formalised actions of a more or less restrictive nature that are often aimed at modifying the behaviour of social groups presumed to be at the root of, or able to solve, the collective problem to be resolved (target groups) in the interest of the social groups who suffer the negative effects of the problem in question (final beneficiaries).

Thus, when we use the term ‘public policy’, we are implicitly referring to a large number of legislative and administrative activities aimed at the resolution of real problems. Most modern legislation is only effective when the political, administrative and social actors involved in the different institutional arrangements are involved in the decision making. The desired effects are only attained, however, in the aftermath of a group of complex decisions that form a sequence between the centre and the periphery. It is this set of decisions and activities that we define here as a ‘public policy’ – decisions taken by public (and sometimes private) actors that are aimed at channelling the behaviour of a target population so that a collective problem that society is not in a position to manage on its own can be resolved by public effort. This set of decisions includes the decisions taken at all stages of public action, and also includes general and abstract rules (laws, decrees, ordinances and so on) and the individual acts and concrete products that arise during policy implementation (administrative decisions, authorisations, subsidies etc).

With a few exceptions, the legislation of the original liberal state was primarily limited to the definition of the frame conditions likely (p.25) to facilitate the resolution of problems by the agents of the private sector. Thus, public activities were limited to the production of legislation and its occasional enforcement by the courts in the case of litigation. It was only from the 1930s, and in close association with the development of the state, that there were efforts on the part of the public service to design interventions directly targeted at concrete problems. This state interventionism is at the root of the conception of real public policies in the sense of the above definition. It was on the basis of these policies that politicians, public managers and researchers started asking questions about the effectiveness and efficiency of different regulatory, incentive, economic and, more recently, persuasive and informational instruments (Knoepfel and Horber-Papazian, 1990; Morand, 1991).

Our definition of public policy claims to be primarily analytical. However, public administrations themselves consider and increasingly manage their daily activities with explicit reference to this kind of analytical frame1. This analytical frame gives the observer a view of all of the different activities that prompt the concretisation and implementation of political and administrative decisions, activities that are considered too often in isolation by the public agents concerned. This frame of reference makes it possible to clarify the distribution of the political and administrative functions and responsibilities of each of the instances at different state levels. Finally, it enables the distinction of those public activities associated with the resolution of a concrete problem through the creation of a particular policy from other state activities that are associated with the management of the entire political-administrative system. Thus, our definition applies to policies referred to as ‘substantive’ (Bussmann et al, 1998) as opposed to ‘institutional’ or ‘constituent’ policies (Lowi, 1972; Quermonne, 1985). In effect, the main object of the latter is the promotion, transformation or disintegration of state or social institutions (Quermonne, 1985, p 62) and not – at least not directly – the resolution of a social problem. We also classify budgetary policies as institutional policies even if budgetary tools are part of the panoply of the instruments of substantive public policies. These exclusions may be problematical in some situations in which institutional or budgetary changes have either explicit problem-solving goals or indirectly have this effect. However, situations of this kind will be complex, and not easily subjected to systematic analysis.

(p.26) 2.3 Constituent elements of a public policy

Several of the constituent elements of a public policy can be noted in the following definition. The constituent elements of a public policy comprise:

  • A solution to a public problem: a policy aims to resolve a social problem that is politically acknowledged as public and necessitates the re-establishment of the communication between several social actors that has broken down or is under threat. Thus, the proposed definition presupposes the recognition of a problem, that is, a socially unsatisfactory situation whose resolution is subject to action by the public sector. Nonetheless, problems that have been the object of public policies can return to the private or social sphere and disappear from the political agenda: for example, behaviours that were formerly subject to moral condemnation such as cohabitation were an element of family policy that is no longer relevant today. Furthermore, the situation can arise whereby public bodies created to resolve a given problem are on the lookout for new public problems: hence, the maintenance of a federal Swiss stud farm within the Federal Office for Agriculture’s main ‘research and popularisation’ division which, having initially represented a means of action in the context of military policy, now focuses on a (new) problem: agricultural research.

  • The existence of target groups at the root of a public problem: all public policy aims to channel the behaviour of target groups, either directly or by affecting these actors’ environment. The ‘causality model’ (see Sections 3.3 and 3.4 in Chapter Three) that underpins the coherence of public policy leads to the identification of the target groups of the policy, that is, the social groups who should be able to resolve the problem by changing their behaviour. A political declaration to the effect that air should be clean, public order restored, unemployment reduced that is not accompanied by the identification of the social groups to be called on to change their behaviour with a view to fulfilling these objectives cannot, therefore, be considered a policy. It should be noted, however, that the target groups of a policy can evolve over time: thus water protection policies started out by defining devils, witches and pagans, households, industrial enterprises and, more recently, farmers, as the target groups at whom public intervention is aimed.

  • (p.27) Intentional coherence, at the very least: a public policy is created with a given direction. It presupposes “a theory of social change” (Muller, 1985, 1995; Mény and Thoenig, 1989, p 140) or ‘a causality model’ (Knoepfel et al, 1998, p 74), which the policy will attempt to apply in its attempt to resolve the public problem in question. It also assumes that the decisions or actions taken are connected. Thus, a lack of coherence will manifest itself in the purely occasional coincidence of measures that are aimed at the same target groups but are not connected to each other in accordance with the legislator’s intention. This is the case, for example, when measures to save energy are introduced as part of an energy policy, while at the same time the VAT or sales tax on energy products is increased for exclusively fiscal reasons. In this case, it is not advisable to include the fiscal measure in the energy policy. If the energy policy does not contain any economic measures, due to the lack of intentional coherence, the fiscal measures may not be considered as energy policy means. However, there will be areas of ambiguity, particularly where there are alternative interpretations of policy goals or actors do not make their policy goals explicit. For example, the congestion charge introduced by the Mayor of London was presented, as its name suggests, as a measure to reduce traffic congestion. That did not stop commentators suggesting that the Mayor’s primary motive had been to increase the revenue available for his use. Interestingly this policy has been resisted by the various foreign embassies in London, particularly the United States one, who argue that the charge is a tax and they should not pay it as they are immune from UK taxes. The policy analyst is hardly in a position to judge what the ‘real’ intention was in a situation like this. The solution to this problem, however, must either involve, as in this proposition, the acceptance of stated intentions or the imposition in the analysis of the presumed intention (with the analyst acknowledging that this is their imposed assumption). Hence this example may be quite appropriately explored in terms of the policy’s impact on traffic congestions (as it is in practice) but could alternatively be explored as revenue-raising policy. See Section 7.2.1 for further discussion of issues about alternative problem definitions.

  • The existence of several decisions and activities: public policies are characterised by a group of actions that go beyond the level of the single or specific decision while remaining short of a “general social movement” (Heclo, 1972, p 84). A basic declaration of (p.28) government policy stating that AIDS is a public problem that does not also define the social groups affected by the existence of this problem cannot in itself be considered as a public policy. This kind of declaration may (but does not have to) contribute to the emergence of a new policy if it is followed by legislation and its application.

  • Intervention programme: this group of decisions and actions should – moreover – contain decisions that are to a greater or lesser extent concrete and specific (decisions relating to the programme and its application). Unlike other authors, we are of the opinion that an intervention programme that is specific to one or more authorities (Thoenig, 1985) cannot be considered in itself as a public policy. Thus, Switzerland’s proposed measurement plan for the prevention of atmospheric pollution2 or France’s plan for the protection of the atmosphere within the framework of the Law on Air (1996) can only be considered as an element of a policy in themselves if individual measures that are the object of explicit decisions are (at least partially) implemented. A programme of interventions that has no outcome is not a policy; it is merely a – possibly indispensable – product among the other constituent elements of a public policy (see Section 9.3 in Chapter Nine).

  • The key role of public actors: this group of decisions and actions can only be considered as a public policy to the extent that those who take the decisions act in the capacity of public actors: in other words, the involvement of actors belonging to the political-administrative system or private actors with the legitimation to decide or act on the basis of a delegation based on a legal rule is essential. If this condition is not fulfilled, a group of decisions of this kind (which can, in fact, also impose restrictions on third parties) will be considered as a ‘corporative (associative)’ or even ‘private’ policy. Thus, many of the ‘policies’ adopted by multinational companies (salary scales, environment strategy, environmental management systems) are based on strictly internal decisions and responsibilities.

  • Existence of formalised measures: a public policy assumes the production of acts or outputs intended to channel the behaviour of groups or individuals. In this sense, our definition of a public policy presupposes the existence of a concrete implementation phase for the measures decided on. However, in certain cases, the policy analysis reveals the failure of political-administrative actors to intervene or a lack of recourse to certain intervention instruments. (p.29) In this respect, we differ in part from approaches that also consider groups of non-decisions (Dye, 1972) or non-actions (Smith, 1976, p 13; Mény and Thoenig, 1989, p 152) as public policies. According to our definition, these non-decisions can only constitute a policy if they are simultaneously accompanied by formal decisions, with which they will be associated. This is the case, for example, if a service voluntarily refrains from implementing the procedure for serving notice on a polluting company so as to prompt the company itself to self-regulate.

  • Decisions and activities that impose constraints: traditionally, the majority of actors assume that the decisions made by political-administrative actors are often coercive in nature (Mény and Thoenig, 1989, p 132). If, as Gibert (1985) suggests, the public action is deemed necessary by virtue of the legitimate authority assumed by public power, today, the diversification of the means of action and intervention at the disposal of the political-administrative system is such that this coercive aspect is increasingly less prevalent. Whether they concern the development of conventional public activities (Lascoumes and Valuy, 1996) or contractual activities (Gaudin, 1996; Godard, 1997), the forms of public action adopted today are as likely to be incentive-based as coercive. This has prompted us to modify this aspect of the definition. Thus, many public interventions are currently implemented by means of contractualisation procedures between the state and public authorities (waste management, road maintenance, regional development), between, for example, the state and private or public companies, foundations or cooperatives (service contract for establishments that fulfil public functions such as hospitals, public transport franchise companies, educational establishments etc) (see Chevallier et al, 1981; Finger, 1997).

Figure 2.1 demonstrates the links between the different constituent elements of a public policy. In doing this, it also indicates those elements that will not be considered here as constituent elements of a policy. It is recognised that this seems to eliminate from consideration some actions that are widely seen as public policy decisions. That is not our intention; rather it is to suggest that in these cases systematic policy analysis is probably not feasible. This is an important consideration that highlights elements in the policy process that tend to undermine rational analysis, offering warnings to analysts against getting trapped into pointless evaluative activities or into the legitimation of symbolic or contradictory policies. (p.30)

Public policy

Figure 2.1: The different constituent elements of a public policy

2.4 Policy cycle

Numerous writers (see the review in Parsons, 1995, pp 78–9) have tried to create a diagram conveying the unfolding of the decision and implementation processes involved in the creation of a policy. The overall impression that emerges from the literature is one of a policy ‘cycle’ starting with the emergence of problems and progressing to the evaluation of the results obtained, as shown in Figure 2.2. The questions posed by the analyst can be differentiated for each of the stages in the policy cycle (see Table 2.1).

This approach based on the policy cycle model should be understood as a framework and not a rigid grid. This is what Muller has in mind when he notes that “the sequential representation of policies should not be used in a mechanical fashion. Instead, policies should be represented as a continuous flow of decisions and procedures, for which it is necessary to find the meaning” (Muller, 1990, p 33). The proposed grid should be understood as an aid in the quest to understand the decisions taken in the context of a policy. (p.31)

Table 2.1: The different sequences of a public policy


1st phase

2nd phase

3rd phase

4th phase

5th phase


Emergence of problems

Agenda setting

Formulation and adoption of the policy programme

Policy implementation

Policy evaluation


Emergence of a problem

Selection (filtering) of emerging problems

Definition of the ‘causality model’

Application of selected solutions

Determination of eventual policy effects

Problem perception

Outline and formulation of causality model

Definition of suitable and acceptable solution(s) to the defined problem

Action of administrative implementation agents

Evaluation of extent of impacts, effectiveness, efficiency, relevance, with respect to the original problem

Definition of the problem and identification of possible causes

Representation of the problem

Responses of public powers to problems recognised as being the necessary object of a policy

Filtering between ideal solutions and available resources

Request for public action

Selection of instruments

Analyst’s main questions

How is an awareness of the problem reached?

What are the factors that will make the government act in response to the problem?

What are the solutions proposed and accepted by the government and parliament? On the basis of which processes are these solutions formulated?

Have the decisions of legislature and the government been implemented?

What are the direct and indirect effects of the policy?


Public policy

Figure 2.2: The policy cycle

We adopt this approach again in a slightly modified form when we present our own theoretical model (Part III, Chapter Six). Therefore, and more precisely:
  • The phase involving the emergence and perception of problems is defined as a situation triggering a collective need, an absence or dissatisfaction, which is identifiable directly or via external manifestations and for which a solution is sought (Jones, 1970, p 53). More generally, a problem exists when there is a difference between the current and desired status of a situation. Nevertheless, a significant number of social problems exist that are not the subject of a public policy. The passage from the existence of a problem to its political processing results from a ‘social [re]construction’ of this problem that itself is related to the extent of the coverage it receives in the media (through scientific knowledge, the dissemination of information, lobbying etc).

  • The agenda-setting phase corresponds to the consideration by the key actors of the political-administrative system of the numerous requests for action made by social groups or even the public services themselves. This agenda setting could be considered as a mechanism for the filtering of problems by public actors.

  • The policy formulation phase presupposes, firstly, the definition of the causality model by the public actors (a definition that is (p.33) influenced to a greater or lesser extent by the social actors) and the formulation of the political-administrative programme (PAP), that is, the selection of objectives, instruments and procedures to be implemented in order to resolve the problem under consideration. Here again, the existence of a filtering and adjustment mechanism may be considered.

  • The implementation phase consists of the adaptation of the policy programme to the concrete situations encountered (production of outputs). This phase is generally a lot more complex than it seems. Here again, several filtering mechanisms will come into play (for example, non-execution, selective application).

  • Finally, the evaluation phase, which we consider as a constituent element of a policy, aims to determine the results and effects of a policy in terms of the changes in behaviour of target groups (impacts) and problem resolution (outcomes).

The interpretation of public interventions as processes (a dynamic vision rather than the static one that is typical of the traditional legal approach) makes it possible to highlight ‘filtering’ phenomena, such as, for example, the failure to take the initially identified beneficiary groups into account during a development process (for example, a motorway construction project that crosses areas populated by immigrants or other disadvantaged groups, or the construction of high-tension lines at the cost of nature conservation) (see Knoepfel, 1997a). Figure 2.3 identifies the position of the different filtering mechanisms throughout the policy cycle: filtering during the perception of problems placed on the political-administrative agenda, adjustment filtering during the policy formulation phase, implementation filtering and finally, evaluation filtering.

While analysis based on the policy cycle offers certain advantages, it also involves a number of restrictions. The advantages include the following elements:

  • The policy cycle approach enables consideration of the existence of retroactive loops throughout the process, for example, the questioning of a PAP as a result of opposition arising during its implementation phase (the case of strong opposition to the setting up of a nuclear power station that results in the redefinition of energy policy; opposition to the extension of a civil airport that affects the basic conception of air transport policy).

  • The identification of the stakes and actors involved in each stage of the policy cycle makes it possible to reduce the complexity of (p.34)

    Public policy

    Figure 2.3: The public policy process and filtering mechanisms

    (p.35) the subjects being analysed. Thus, it is possible to analyse the actor constellation (public/private, central/local) and its variation throughout the policy cycle (appearance and disappearance of actors during each sequence).

  • The formulation of analytical questions, hypotheses and partial theories for each stage of the policy cycle makes it possible, in particular, to single out the factors analysed on the basis of different disciplinary fields – sociology, law, political science, economics – and to create sub-disciplines: formulation of public action (‘policy design’), policy implementation research (‘policy implementation’), and policy programme evaluation (‘policy evaluation’).

  • The possibility of combining policy analysis with a rationalising vision of public action (for example, the linear link between the objectives, means and results that is implicit in management strategies such as the Planning, Programming and Budgeting System [PPBS]). This makes it possible, for example, to detect errors in the identification of the problem or to identify gaps in policy implementation.

This heuristic model may be compared to the stages in the resolution of a (private) problem. This similarity prompts several authors to consider public action as a rational enterprise of problem resolution without necessarily taking the distinctive features of the public sector into account (Table 2.2).

However, the ‘policy cycle’ model does have certain limits from an analytical perspective (see for example, Jenkins-Smith and Sabatier, 1993, pp 3–4; Hill and Hupe, 2006; Hupe and Hill, 2006).

Table 2.2: Similarities between the policy cycle and the stages of problem solving


Problem solving

Public policy


Problem recognition

Agenda setting


Proposal of solution

Policy formulation


Choice of solution

Decision making


Putting solution into effect

Policy implementation


Monitoring results

Policy evaluation

Source: Howlett and Ramesh (2003, p 13)

(p.36) In effect:

  • This is a descriptive approach that can be deceptive as the chronological course of the policy process does not necessarily coincide with the order of the different stages in the model. Thus, a programme may be implemented prior to its precise formulation during the emergence of new problems (for example, in the case of efforts to overcome pollution caused by agriculture in France; see Larrue, 2000). Breaks may also occur in the process with the reformulation of the public problem and the solutions before the measures initially planned are implemented and/or evaluated (for example, in the case of political asylum policy in Switzerland; see Frossard and Hagmann, 2000; there are similarities to this in the UK in the case of the rapid succession of political responses to fears about terrorism, with new initiatives occurring before exiting policies have been properly established).

  • This heuristic approach does not enable the development of a true model of the causality of public policies and the identification of logical links between the different stages. It runs the risk of giving an artificial coherence to the policy by prompting the analyst to construct links between elements that do not exist in reality.

  • The policy cycle model is in line with a legalistic interpretation of public action (‘top-down’ approach) and centred on state action, and it fails to take account of an approach that originates with social actors and their context (‘bottom-up’ approach). Thus, one could be led to incorrectly attribute the reduction in electricity consumption to energy-saving measures when it actually results from an increase in prices or downturn in the economy. Similarly, a number of solutions exist that are looking for a problem: a state service that is due to be closed down (for example, the federal Swiss stud farm) will create a new problem in order to survive (the risk of disappearance of traditional horse races that are part of the national heritage).

  • This approach does not make it possible to go beyond a sequential analysis and consider, in particular, several cycles unfolding at the same time or the possibility of incomplete cycles. For example, in order to understand drug policy it is important to dissociate the cycles and identify the different pillars of the policy: crackdown, prevention, survival aid (related to AIDS), medicalisation (with methadone) and medical control.

(p.37) Nonetheless, in the context of our policy analysis model, we believe it is worth staying with this sequential approach while taking its advantages and limits into account. We see it as constituting a reference framework that can be considered as a tool of a pedagogical or heuristic nature, but which also should be complemented by a more cross-sectional analysis of the stages of a policy. This cross-sectional analysis rests on the detailed study of the main keys for reading policies, that is, the actors, their resources and the institutional framework within which they interact (Chapter Three, Four and Five).