Content uploaded by Frédéric Varone
Author content
All content in this area was uploaded by Frédéric Varone on May 13, 2016
Content may be subject to copyright.
CHAPTER 3:POLICY ACTORS
We take policy to mean a series of decisions or activities resulting from structured and
recurrent interactions between different actors, both public and private, who are involved
in various different ways in the emergence, identification and resolution of a problem
defined politically as a public one.
In this chapter, we focus on the types of actors concerned by policy, while following
chapters will deal with the resources to which these actors have access in order to
represent their interests (Chapter 4), on the one hand, and with the institutional context
that influences their individual and collective behaviour (Chapter 5), on the other. These
three concepts (actors, resources and institutions) constitute the principal focus of our
analysis and are the key factors on which we will base our policy analysis model (Part
III).
3.1 ‘EMPIRICAL ACTORS’
Given that policies embody the results of the interactions between different public and
private actors, we must start by defining the actual concept of an actor. For the purposes
of this study, the term actor can be taken to designate either an individual (a minister,
member of parliament, specialist journalist etc.), several individuals (constituting for
example an office or a section of an administration), a legal entity (a private company, an
association, a trade union etc.) or a social group (farmers, drug users, the homeless etc.).20
Note, however, that a group of several individuals constitutes a single actor insofar as,
with respect to the policy under consideration, they are in broad agreement and share a
common approach as far as the values and interests that they represent and the concrete
aims that they pursue are concerned. This consensus can be arrived at, for example,
through the hierarchical structure or through the democratic process.
Our approach to the concept of actor is inspired by that of Talcott Parsons (1951). In his
view, in order to analyse a social action, we must focus essentially on the simplest unit
that retains the significance of what Parsons terms a ‘unit-act’. This elementary act is
undertaken by at least one actor who has an objective (bringing about a future state of
affairs with a view to which the actions of the actor in question are directed), and who
uses certain means to achieve that objective (Bourricaud, 1977, p.31). Thus, depending
on the individual case, the actor concept can apply to an individual, a group or groups of
individuals or to an organisation, the latter being defined in terms of the shared ideas or
common interests that link its members. As Olson remarks in his book on the logic of
collective action: ‘Without a common point of interest, there is no group’ (Olson, 1978,
p. 29).
Every individual, legal entity or social group is considered as an actor once, by virtue of
their very existence, they belong to the social field regarded as being relevant to the
analysis:
An individual in a given field does not qualify as an actor by virtue of his
understanding of, or control over, events, nor on the basis of his awareness of his
32
interests and scope for action, nor, a fortiori, because he is aware of his place in
history or in the process of social change, or because he participates in ‘the
production of society (Segrestin 1985: 59).
He has this status simply by belonging to the field being studied in so far as his
behaviour can be shown to contribute to the structuring of this field. It is not,
therefore, a problem of awareness, lucidity or identification: it is simply a de facto
situation, which means that this becomes a question of research (Friedberg 1993:
199)21
In this way, every individual or social group concerned by the collective problem
addressed by a policy can be considered as a potential actor capable of being part of the
‘arena’ (cf. 3.3.1.) of this policy. In fact, the actor’s – more or less active – behaviour
influences the way in which the public intervention in question is devised and
implemented.
This broad definition of the actor concept means that the analyst must consider all
individuals and social groups concerned by a specific collective problem. Such a
viewpoint has the advantage of taking account of the fact that public and private actors do
not all intervene actively and visibly at all stages of a policy: their behaviour is
sometimes directly tangible, but equally it is sometimes hard to identify directly. This
depends on, among other factors, the process by which they become aware of their own
interests, their capacity to mobilise resources and form a coalition to defend their rights
and interests and, finally, their strategic decision either to take action or to remain
voluntarily outside the decision-making arena. By adopting the concept of ‘empirical
actors’ proposed by Friedberg, we go along with him in rejecting the distinction made by
several authors between ‘an agent’, who is rather passive and whose behaviour is
determined by the system in which action occurs, and ‘a true actor’, who has become
active and autonomous in relation to the institutional and social context.
In the same way, our perspective leads to a rejection of a formal opposition between
‘actor’ and ‘non-actor’. In this regard, we do not contest the fact that, in the context of a
policy, different individuals or social groups display different degrees of internal
organisation, control different action resources and have different capacities for the
mobilisation of external interests. However, we maintain that any individual or social
group concerned by the collective problem addressed by the policy under consideration is
an actor (at least potentially), even if (momentarily) unable to undertake concrete action
in one or more phases of a public intervention. In fact, an actor’s passivity, whether
deliberate or the result of a lack of resources or of a failure to realize the importance of
certain issues, is a factor that explains why one policy is ultimately developed and
implemented rather than another.22
There are also opponents of a policy who only come to the fore in the last stages of its
implementation: whether during negotiations with those responsible for the policy
implementation (for example the foresters in the case of the implementation of the
legislation on forest clearing analysed by Padioleau, 1982) or through legal opposition
(cases involving environmental protection organisations which oppose infrastructure
policies). THERE IS NO PAGE 34
33
If analysts only focus on the behaviour of the most dynamic and enterprising actors,
stipulating that passive groups are ‘non-actors’, they run the risk of overlooking certain
factors that are central to an understanding of how a given policy is developed. Analysts
sometimes adopt too elitist an approach to the actors’ game, not taking into account
sufficiently the effects induced by the passivity of certain social groups or political-
administrative actors who are concerned by the collective problem under consideration.
Note however, that for obvious reasons connected to empirical observation, analysts tend
naturally to focus on the behaviour of the most active actors within the context of a
policy. It is certainly easier to identify individuals, informal groups and formal
organisations who, having access to the necessary resources, participate on an ongoing
basis in the conception, adoption and implementation of a policy.
3.2 ‘INTENTIONAL’ ACTORS
Adopting a scheme of intelligibility known as ‘actantial’ (Berthelot, 1990, p. 76), we
acknowledge the intentionality of individual action. This takes place in a social context
that can be perceived alternatively as a system of interdependence (Crozier and
Friedberg, 1977; Boudon, 1979), a historical stage in a process (Touraine, 1984) or a
situation pertaining to the here and now. In each case, an actor’s behaviour is never
reduced to a position, role or other type of fixed category. In other words, in our opinion,
an actor always disposes of a greater or lesser margin of discretion and of manoeuvre,
depending on the situation in question. Our thesis here is that no social or political field is
perfectly structured, controlled or regulated. For this reason, individual and collective
actors deliberately exploit ‘areas of uncertainty’ (to use the expression coined by Crozier,
1963) that are an inherent part of political-administrative organisations, formal
regulations and social norms in order to promote their own values, ideas and interests.
They possess, therefore, a certain degree of freedom but also resources (cf. Chapter 4),
which enables them to develop strategies and tactics, or even to adopt ‘goal-oriented
behaviour’ (Berthelot, 1990, p. 80).
Thus, we do not seek to deny the influence – which is sometimes quite considerable –
that the actors’ institutional and social context has on their decisions and actions. On the
other hand, we believe that these institutional factors do not determine the assessments,
choices and behaviour of public and private actors in an absolute and linear manner (cf.
chapter 5). We reject, therefore, the holistic theses that assume that social phenomena,
policies for example, have their own intrinsic nature and their own laws which inevitably
lead individuals to act in one way rather than another. On the contrary, we propose that
policies should be interpreted as the result of the behaviour of actors who are (partially)
autonomous. Thus, we adopt the principles of methodological individualism as developed
by the sociologists Boudon and Bouricaud (1990, pp. 301-309).
The area of uncertainty is particularly significant in the context of unforeseen crises (the
accident at the Chernobyl nuclear power station, natural disasters etc.). The actors who
have to intervene in such circumstances are unprepared and have to cope as best they can
with no system to fall back on. When this happens, the different public authorities
concerned can be observed to react in different ways (Czada, 1991; Keller-Lengen et al.,
34
1998; Schöneich et al., 1998), and indeed such disparities can be observed even within
one and the same organisation (Müller et al., 1997).
This in no way implies that we interpret actors as totally rational beings, motivated only
by the maximisation of their personal utility (both material and non-material) and
omniscient, as suggested by the homo economicus model beloved of neoclassical
economics. For cognitive, emotional and cultural reasons etc., the rationality of
individuals and social groups remains necessarily ‘limited’ (Simon 1957). An actor’s
behaviour can never be reduced to its purely instrumental dimension, that is to say to the
accomplishment of a predefined objective based on a considered choice and perfect
realisation of the best possible course of action. Actors are in part calculating and are
motivated by the satisfaction of their personal needs (means-end rationality or
Zweckrationalität according to Max Weber) and partly drawn towards the defence and
promotion of collective values (value rationality or Wertrationalität). It is important to
bear this dual motivation in mind when interpreting the behaviour of actors in the field of
policy.
Thus, for example, while battling on behalf of the unemployed, civil servants can work to
further the interests of their own departments so as to ensure their survival (which might
be endangered, for example, by the privatisation of job-finding schemes for unemployed
people). In the same way, the social services of a church become involved, for altruistic
reasons, in the state system of providing home care for the elderly, while at the same time
pursuing their own agenda of consolidating the role of the church and its message at the
local community level (cf. Gentile, 1995).
We take the view that actors are rational in the sense that they care about the
consequences of their own decisions and actions, even if they are unable to anticipate and
control all of the effects that stem from these, and especially the adverse or undesirable
effects that derive from the cumulative actions and behaviour of individuals (Boudon,
1979). At the same time, we propose a very broad interpretation of the intentions and
interests that underlie all human activity: an actor’s motivations are manifold, especially
because they depend on the experiences and past history of the individual or social group
concerned and also on the situation that pertains at a given moment in time and which
entails certain constraints as well as opportunities for action. This being the case, we can
speak of a ‘situated rationality’: the analyst being faced with interpreting individual and
collective activities in terms of the logic and expectations upon which strategic
calculations are based and, at the same time, in terms of the actors’ ignorance or intuition,
their emotions or feelings and even in terms of the weight and impact of historical factors
(Friedberg, 1993, p. 211).
Briefly, policy actors can be described in the following terms (Crozier and Friedberg,
1977, pp. 55-56):
1. Actors rarely define clear, explicit or coherent objectives. They change them
as they go along, if only because the unanticipated consequences of their own
actions and those of other actors in the domain of policy compel them to
readjust their objectives or re-evaluate their positions. What was a means to an
end at one moment becomes an end in itself at another, and vice versa.
35
2. Although it may sometimes appear erratic, actors’ behaviour always has a
meaning and a logic of its own which the analyst tries to decipher. Instead of
being necessarily rational in relation to predetermined objectives, it is
sometimes quite reasonable given the constraints and opportunities afforded
by a given situation. According to a subjective appraisal of the institutional
context and of the other actors’ strategies, the actor adapts behaviour so as to
be able to participate in and learn the rules of the ‘game’ of a policy and be
acknowledged socially by the other actors involved. This leads to sometimes
strange coalitions or alliances, for example between a company responsible
for producing moderate levels of pollution and environmental protection
organisations, which join forces in order to protest against a company
producing high levels of pollution; or around an issue like wind farms where
local residents, nature protection groups and tourist businesses band together
(cf. Carter, 2001, p. 281).
3. An actor’s ‘strategic instinct’ (to borrow the expression used by Crozier and
Friedberg, 1977) is characterised by two complementary aspects. On the one
hand, an actor tends to go on the offensive when taking advantage of
opportunities to improve a position and further immediate interests (direct
intervention on the substantive components of policy). On the other hand, an
actor adopts a more defensive approach when seeking to maintain and
broaden a margin of freedom, i.e. the capacity to act in the way wanted at a
later stage (indirect intervention on the institutional components of policy).
From this perspective, it can be seen that every actor weighs up the short-term
gains and the advantages of a longer-term investment, this encourages
participation in one or several phases of a policy. In this way, a right-wing
Swiss party which, in spring 1999, aimed to fight against an environmental
policy which it deemed to be too strict, did not ask for the prevailing
environmental regulations to be relaxed (substantive factor), but instead
launched an attack on the environmental protection organisations’ right to
monitor and control the way in which the regulations were implemented
(institutional factor). The party appears to have been perfectly aware that it
had no chance of obtaining a relaxation of the substantive regulations.
3.3 THE TYPES OF ACTORS
Our concept of policy actors leads us first of all to identify and define the parameters of
the arena in which the actors will intervene (3.3.1) Within this arena, policy actors can be
distinguished in terms of their public nature, that is to say political-administrative actors
who are vested with public authority (3.3.2), or in terms of their ‘private’ nature, that is to
say actors who belong to what are known as socio-economic or socio-cultural spheres
(3.3.3). But the concepts of public and private need using with some care, as we show
below. The second group can be broken down into the target groups (the actors whose
behaviour is politically defined as the (in)direct cause of a problem or who are able to
take action to deal with it), the end beneficiaries of a policy (actors who experience the
negative effects of a particular problem and whose situation should be improved
following the implementation of public intervention) and the third-party groups, who are
36
affected indirectly by the policy, either positively (= positively-affected third parties) or
negatively (= negatively-affected third parties). These two latter groups comprise all
actors whose personal situation is altered by a policy without having been directly
targeted by the policy, either as target groups or end beneficiaries. These three types of
actors constitute what we call the triangle of actors (3.3.4). It is quite possible for actors
to be both targets and beneficiaries (though there may be some problems here about the
differences between how they see themselves and how they are seen by political-
administrative actors).
3.3.1. The policy arena
Our intention here is to deal with policy from the perspective of the solution of a problem
considered as pertaining to the public domain. Here we present the different actors found
in a policy context, a context in which crucial interaction takes place between the
different policy actors. The way in which this arena, in which these actors interact, is
structured is neither neutral nor without effect on the behaviour of the different actors or
the selected modes of public intervention. The policy arena is determined by – among
other factors – the logic of the state. Traditionally, and especially since the advent of the
welfare state, ‘public matters’ are managed by the public actors of the political-
administrative system. But the principle of the constitutional state and that of democracy
equally require the involvement of private actors whose interests and objectives are
affected by the collective problem to be resolved in one way or another. Every policy
arena inherently constitutes a framework that, to a greater or lesser extent:
•is structured
•formally defined
•and contains public actors interacting with private actors
thus allowing alternative strategies to be developed.
In this way, a policy is designed and managed by public and private actors who together
constitute an often highly complex network of interactions, which can experience
problems involving horizontal co-ordination (e.g. relationship between actors belonging
to the same government level) and vertical co-ordination (e.g. relationship between
central, regional and local actors). Although they belong to different organisations and
represent interests that are often opposed, if only because of the contentious issues in
which they all have a stake, these various actors form areas of interaction. The
boundaries of these areas in which they interact are often difficult to define, especially if
the analyst focuses on peripheral actors. This is especially true of the initial phases of the
emergence of new policies. It is these ideas that have been given particular attention in
analyses of ‘governance’ (Pierre, 2000; Richards and Smith, 2002).
On the other hand, in every policy arena it is easier to identify a hard core of actors who,
despite potential conflicts, have a (semi-)vital interest in not losing their position and,
consequently, in controlling and even limiting new arrivals’ access to the area in which
they operate. This ‘policy community’ (Richardson and Jordan, 1979) is often subdivided
into different coalitions (the term used by Sabatier and Jenkins-Smith, 1993, is ‘advocacy
37
coalitions’). At the same time that they struggle to assert their own interests or ideas,
these coalitions together try to differentiate themselves from the individuals and groups
outside the policy arena. With this end in view, the actors in a policy arena develop, for
example, a language that is specific to ‘their’ policy, control the way in which
information is circulated or try to avoid becoming politicised in case that might
overwhelm ‘their’ ordered world in which a level of cohesion is preserved despite the
differences that exist between the different coalitions.
Thus, the agricultural policy arena contains actors that vary from the large international
companies that produce animal feed and auxiliary materials (fertilizers, pesticides) to
intensive farmers, small farmers and the food-processing industry, including the
distribution chains. The public health policy arena includes not only doctors’
organisations, hospitals, the public health services and health insurance companies, but
also the manufacturers of pharmaceuticals. The environment policy arena comprises the
various polluting sectors of industry, eco-industries (companies producing anti-pollution
products), environmental protection organisations and the public services.
As a general rule, policy arenas do not change much in their make-up. A coalition that
was formerly a minority one can become dominant, or the power relationship between
central and local actors can undergo changes, but the configuration of actors and, by
extension, the way this is defined in relation to everybody else, is rarely fundamentally
called into question.
Examples of changes among the actors in the policy arena may nonetheless be linked to:
•A radical change in the way the problem in question is perceived: for
example, narcotic drug use has gone from being seen merely as an issue
requiring a prison sentence to being seen as a health issue, as an aid to
survival. This being the case, the policy requires the participation of doctors
(as methadone must be prescribed under medical supervision), or even health
insurance companies.
•Strong opposition to a given project on the part of some actors: for example,
the construction of railway infrastructure which encountered massive
opposition from environmentalists and ecologists (Rail 2000 in Switzerland,
the Mediterranean TGV in France). The transport authorities have found
themselves obliged to systematically include private associations as actors in
the arena of rail transport policy.
•The departure of some actors from the policy arena: small mountain farmers
in Switzerland, who thought that official agricultural policy was too focused
on the interests of lowland farmers and who advocated a radical reorientation
of this policy, turned their backs on the official policy forum in order to create
a new forum where they could express and assert their point of view more
effectively. In France, the Confederation of Small Farmers, a very small
farmers’ union, adopted a similar stance.
38
3.3.2 Public actors
If there is one common denominator that connects all policies it is the fact that they are
constituted by a range of initiatives usually undertaken by public actors. Thus, it would
seem essential to provide as precise a definition as possible of the characteristics of
public actors as opposed to the private actors who are also involved in policies. This is all
the more necessary given the fact that the names and titles by which the actors are
ordinarily known usually provide little information about what they actually do, whether
in a public capacity or not. Furthermore, the definition of public actors must make it
possible to differentiate between policies described as ‘corporatist’ (associative) or
‘private’. It should be noted that, in terms of an analytical approach, this distinction will
hardly fit in with a prescriptive theory that seeks to define the ‘appropriate’ role of the
state in society. In order for it to work, it is imperative that this definition be inclusive,
that is to say that it should not exclude any of the activities in which political-
administrative actors are involved. However this is difficult.
Several approaches are possible.
There seems little point in defining public actors exclusively in terms of the legal
dimension of their actions. For a long time, the basic element of public activity was
considered to be formally constituted by what are known as administrative measures.
According to the definition of federal Swiss administrative law, for example, the latter
are decisions ‘taken by the authorities in individual cases, founded on federal public law
and having the following objectives: (a) to create, modify or annul rights or obligations;
(b) to record the existence, non-existence or scope of rights or obligations; (c) to reject or
declare inadmissible applications tending to create, modify, annul or record rights or
obligations’.23
39
The order for civilians to report for military training, tax returns and planning permission
etc. are all examples of this. Founded on law, these administrative measures must also
comply with it. As such, they may be subject to legal monitoring, usually by an
administrative court. However, it can happen that public actors escape, quite legally, from
being monitored by the administrative jurisdiction. The state’s interest in not being
restricted by public law is obvious: by acting in the sphere of private law, it escapes the
constraints connected with observance of the principles of administrative action (for
example the necessity for a legal basis, equal treatment, proportionality etc) and the
principles of public law which govern the administrative domain (Manfrini, 1996). This
‘flight’ outside the realm of public law prompts the state and the public authorities to
acquire shares as a limited company or to create private law foundations or semi-public
companies (Société d’économie mixte – SEM) through partnership with private actors.
Today, in both France and in Switzerland, and of course in the UK, more and more
organisations based on private law exist that carry out public functions. These are mainly
organisations that provide goods and services in the energy sector (EDF, Energie de
l'ouest Suisse S.A.), the telecommunications sector (France Télécom, SWISSCOM) and
the transport sector, (SNCF, CFF-SA) to the general public and companies. At a more
local level, such companies provide environmental services (household waste-
incineration plants, water distribution services etc.) and transport services (local transport
companies, contracted companies etc.), though it should be noted that in the UK they do
this under contracts to provide services for public authorities.
In the UK these distinctions between ‘public’ and ‘private’ are often even harder to make.
Whilst in modern times an important range of public or administrative law has developed
the starting point is, as an influential textbook on administrative law puts it, that: ‘there is
no formal distinction between public and private law. The ordinary law of the land, as
modified by Acts of Parliament, applies to ministers, local authorities and other agencies
of government, and ordinary courts dispense it’ (Wade, 1982: 12). That may partly
account for the way in which the implications of developments, like those discussed here,
have received rather less attention in the UK. This topic is further discussed below with
reference to Jordan’s exploration of this issue (1994).
By having recourse to private law, the state and the different public bodies can
successfully avoid being subject to administrative and judicial control. However, this
does not prevent them from intervening in their capacity as public authorities. In contrast
to a restrictive judicial approach, it must be emphasised that ‘para-state’ policies pursued
in this context remain policies and that the administration acts as a public actor.
It should be noted, however, that the flight of the state outside the domain of public law
poses a problem in terms of political control: the administrative decisions subject to
public law can, in theory, be the subject of parliamentary debate, even if they usually
come under the jurisdiction of the executive. On the other hand, it is much more difficult
for the legislature to monitor all acts of public authority which are not public decisions at
the formal level. The degree of effective political control is, therefore, also an inadequate
criterion by which to define an actor as public or private, especially in a period of
economic liberalisation and deregulation – indeed privatisation – of the different public
services (most notably the telecommunications services, railways, postal, gas and
electricity services). There are of course many institutional variations around this theme,
and in the case of the UK there is an increasing concern about measures which remove
quite significant policy decisions from parliamentary scrutiny, these cannot be explored
more here but the general point applies across national contexts.
As far as political actors caught up in power struggles are concerned, the line of
demarcation between public and private actors is an important factor in the development
of their strategies. In fact, confronted with a social problem that has not yet been tackled
by the state, political actors can either propose the introduction of a new policy (which
will be politically costly to a greater or lesser degree), or propose that a ‘corporatist’ or
‘private’ policy be established.
There are numerous examples of policies of this latter type. Most notably:
•wage policy (collective agreements),
•the voluntary compliance of companies with manufacturing standards
(categorising of companies according to ISO regulations)
•disclosure agreements entered into by bankers who undertake to monitor the
origins of certain funds which may be linked with money-laundering
operations or tax evasion.
In order to define public actors as a whole, we return once more to the classical term
‘political-administrative system’ initially proposed by Easton. According to Easton, the
political-administrative system comprises all of a country’s governmental
(parliament/government) administrative and legal institutions, which have the capacity,
apparently legitimised by the legal establishment, to structure any sector of society
through decisions of an authoritarian nature. These decisions are the result of political-
administrative processes that are completed according to precise rules of procedure
governing internal and external interactions (Easton, 1965, p. 25).
Some aspects of this definition of the political-administrative system and the public
actors who constitute its fundamental components, are worth singling out:
•The classical dimension of this definition resides in its affirmation of the
sovereignty of public actors: the state is supposed to be the only entity entitled
to exercise power of restraint over all other sub-systems and over citizens
(legitimate monopoly of power – Max Weber).
•The administrative organisations of the political-administrative system form
an important and relatively independent centre of gravity (for example in
relation to governments and parliaments) in the body of a state’s institutions.
•The notion of interaction suggests that there are reciprocal relationships
between the sub-systems. The political-administrative sub-system transforms
‘societal’ demands (inputs) into restrictive state actions (outputs).
The definition distinguishes between external interactions (relationships of reciprocal
influence between the political-administrative system and its environment, for example
consultation procedures for parliamentary bills, rulings or decrees with regard to people
who will be affected by them or the examination of applications for operating licences)
and internal interactions in the public sector (for example the consultation procedure for
the services linked to ‘environmental impact studies’ in Switzerland, or the procedures
for bringing an issue to the attention of the relevant authorities within the French
administration or the planning inquiry system in the UK. Clear procedural rules are
generally defined for each type of interaction by the legislation and/or regulations.
It should be noted, however, that several private actors to whom the state delegates some
of its privileges belong indirectly to the political-administrative system. In France and
Switzerland these actors are generally designated by the term para-public (or para-state)
administrative bodies. This phenomenon of interpenetration between the public and
private sectors exists in many areas of intervention both in France and in Switzerland
(Mény and Thoenig, 1989; Linder, 1987, pp. 113-116; Germann, 1987). Many of the
examples given below have UK equivalents, even if the legal forms of their designations
differ.
These para-public administrations can exist in various forms:
•Autonomous public establishments created by a law and enjoying a certain
freedom of enterprise: in Switzerland they include the cantonal banks,
universities, the Swiss Broadcasting Company (SSR); and in France they
include public establishments of an administrative nature (such as the water
companies and the National Library), of a scientific, cultural and professional
nature (e.g. the universities) and of an industrial and commercial nature
(establishments that provide all kinds of public services and obtain revenue
from these services: airports, the national electricity supplier - Electricité de
France - EDF - and the national railway - Société Nationale des Chemins de
Fer - SNCF).
•Semi-public and private organisations: local co-operative style companies for
the promotion of the economy, or national organisations like the Swiss
cooperative for the study of radioactive waste deposits (CEDRA) in
Switzerland and Société d’Équipement, the regional or local development
companies, in France.
•Private organisations: the central union for milk producers, pension funds,
health insurance and social security organisations in France, or any French
organisation set up as an association in accordance with the terms of the Law
of 1901, as long as it is under the control of political-administrative actors. In
the UK professional associations licensed to regulate the behaviour of their
members would also come into this category.
•Social organisations: Pro Senectute, Helvetas, various mutual aid
organisations.
In the UK Jordan has explored this issue in a book in which he writes of ‘government in
the fog’. He argues that ‘the public/private divide is more than a matter of an imprecise
boundary and involves a scale of problems that throw the categories into doubt’ (1994, p.
183).
In order to define public actors, we use the notion of the political-administrative
arrangement (PAA, cf. Chapter 8.2.), an entity that is structured by legal rules governing
competencies and administrative procedures and by other relatively informal institutional
rules, and which encompasses all public actors involved in the development and
implementation of a policy. This notion is based on the existence of public responsibility
and, consequently, direct government control of these actors, and it does not, therefore,
include private actors. For this reason, it is distinct from the notion of ‘policy networks’
developed most notably in France by Le Galès and Thatcher (1995) following the Anglo-
Saxon analyses dominated by Richardson and Jordan (1979), Rhodes (1981), and Smith
(1993). However, it is used here in the context of policy analysis, that is to say that the
‘PAA’ is analysed from the point of view of its impact on the policies analysed, and not
for its own sake (cf. Chapter 8).
In order to draw on the defining element of government control, research into public
actors also focuses on the administrative organisations to which they ‘belong’
institutionally. The aims and internal workings of these organisations explain, in part at
least, actors’ behaviour and, more widely, the development and implementation of
policies. The structures and procedural regulations of a public administration (for
example the federal offices and cantonal services in Switzerland, the different levels of
the administrative hierarchies in France, the law determining the functions of local
government in the UK) prevent an individual actor from acting in a wholly autonomous
way. The principles of democracy and the constitutional state imply that public actors are
subordinate to departments politically responsible for the actors’ ‘parent’ organisations.
Therefore, public actors can be ‘handicapped’ by the administrative context in which they
operate when a specific policy is implemented. They will then tend to try to break free of
their control and create a political and administrative coalition outside of their formal
organisation.
But in order to escape the confines of an institutional framework deemed to be too
narrow or restrictive, actors may also create new structures and organisations which will,
for example, be able to collaborate more closely with private actors involved in policy.
These collaborations can take various institutional forms, which we will discuss in detail
later (cf. ‘policy network’ and ‘public-private partnership’ concepts) (cf. Chapter 8).
3.3.3 Affected actors
The affected actors consist of target groups, end beneficiaries and negatively or positively
affected third parties.
The target groups are made up of people (individuals or legal entities) and organisations
whose behaviour is required to change. It may be the (in) direct cause of the collective
problem that a given policy aims to resolve, or it may be regarded as appropriate that it
should adopt the appropriate remedial action. Consequently, the target groups’ decisions
and activities are – or will be – the subject of concrete state intervention. The policy in
question will impose obligations on them or grant them rights. It is presumed that, as a
result of such measures, either that the target groups will alter their behaviour and or that
the collective problem can be resolved or at least mitigated. The notion that groups who
are the cause of the problem (polluters, providers of unsafe goods, drivers causing
accidents) should change is straightforward here, a word of explanation is required
however on the identification of target groups who are not in this category. For example,
schools may not themselves be deemed to blame for illiteracy or doctors to blame for ill
health, nevertheless the alteration of behaviour by these actors may reasonably be
expected to contribute to the reduction of these problems.
Even formally ‘public’ actors may need to be seen as the ‘targets’ of policy inasmuch as
the concern is with changing their behaviour. The literature on implementation to be
discussed in chapter 9 has been much concerned with situations in which super-ordinate
authorities are seeking behavioural change from public organisations. Whilst in some of
this literature the issues involved are about inter-governmental relations, where the
political-administrative authority relationship is a matter of dispute (this is particularly
true in federal contexts), in other cases it is intra-organisational relationships that are
given attention (where professionals claim autonomy or street-level bureaucrats resist
instructions). In many of these cases political processes can be involved very like those
where private organisations are the targets, with organisations of implementing agencies
or their employees participating in the policy process (even in the agenda setting and
policy formulation stages).
The end beneficiaries are the people (individuals or legal entities) and organisations of
these people who are directly affected by the collective problem and experience its
negative effects. Once the policy has been implemented effectively, these actors can
expect a (possible) improvement of their economic, social, professional and ecological
circumstances etc. End beneficiaries are actors who benefit, more or less indirectly and
according to the intended objectives of the policy concerned, from the target groups’
altered behaviour. In most cases, the individuals who make up this group of actors are far
more numerous and more difficult to mobilise and organize than target groups.
Finally, the third-party groups comprise all persons (individuals or legal entities) and the
organisations representing their interests who, without being directly targeted by the
policy, see their individual and/or collective situation altered permanently. This change
can turn out to be either positive or negative. In the former case, people are defined as
positively-affected third parties, sometimes unintentionally so, of the implementation of
the policy and the corresponding term applied to groups of people who are negatively
affected by the implementation of a policy, is negatively-affected third parties. These two
sub-groups of actors tend to either support or oppose the policy that is indirectly altering
their own situation and circumstances and, as a consequence of this, to form coalitions
either with the end beneficiaries (in the case of those who benefit), or with the target
groups of the public initiative that has been implemented (in the case of the negatively-
affected third parties).
This system for the classification of actors is illustrated by the following examples of
actors involved in a range of policies:
•The target groups of environment policy are polluters (industries, farmers,
households and public bodies) whose pollutant emissions need to be
reduced; the end beneficiaries are all those whose environment is affected
by the different sources of pollution in a given area (human beings, flora
and fauna); the positively-affected third parties are the industrialists who
develop new less polluting technologies (environment-friendly industries),
and the negatively-affected third parties are those who can no longer
market their polluting technologies and the consumers who end up paying
more for their products.
•According to the model prevailing in most European countries, the target
groups of agricultural policy are farmers producing subsidized agricultural
products, the end beneficiaries are the consumers who profit from the best
market prices. The positively-affected third parties are the food-processing
industries, while the negatively-affected third parties are the
environmentalists, who see the environment as being harmed by intensive
farming methods, those small farmers whose produce is not subsidised and
third countries who import products that compete unequally with their
home-grown produce as a result of this agricultural dumping.
•The target groups of policy to combat unemployment (at least in the micro-
economic sense as opposed to broad economy regulation measures) are the
companies which need to recruit staff (whose resistance or discriminatory
practices are in question); the end beneficiaries are the unemployed people
who are likely to obtain employment; the positively-affected third parties
are the job agencies who act as mediators on the job market; and the
negatively-affected third parties are those who see their incomes limited
by the increase in compulsory deductions designed to fund, partially at
least, the measures against unemployment. In this field of policy we may
particularly find individuals who may be seen as both targets and
beneficiaries: they are required to change their behaviour – undergoing
training, participating in job search activities – but may benefit if they
nevertheless acquire desired employment.
•The target groups of education policy are schools, colleges and
universities (with various statuses in terms of the public/ private
distinction), parents and employers; the end beneficiaries those who secure
more or better education; the positively affected third parties may include
some employers and parents; and the negatively affected third parties
those who have to pay the costs of these improvements without securing
any significant benefits and (perhaps) those whose access to education is
diminished as a consequence of efforts to supply more education to others
(as in the case of some forms of affirmative action, a topic explored in
similar terms to this discussion in an appendix to Stone’s Policy Paradox,
2002, pp. 384-414).
Obviously, it is not always easy to define the different categories of actors. In the context
of a single policy, disagreements can emerge (between actors) regarding the exact
definition of target groups and beneficiaries as they are interpreted from the various
actors’ standpoints, i.e. in terms of the hypothesis of causality that underlies the policy
initiative. The boundaries between these and the third-party groups who may, often
perhaps without being fully aware of it, pay the real costs or secure the benefits may be
obscure. It is worth noting here how public choice theorists have suggested that
governments will deliberately try to obscure how costs fall.
While the discussion above suggests that the distinction between public and private actors
is often a blurred one, the main issues about targeting and benefits can be seen in terms of
the likelihood that the actors in these categories will be private ones. Nevertheless these
actors participate equally in the constitution and structuring of the policy arena by virtue
of their own responsibility and because they are outside the scope of direct government
control.
3.4 THE ‘TRIANGLE OF ACTORS’ OF A PUBLIC POLICY
These different types of actors constitute what we call the ‘basic triangle’ of a policy. The
political-administrative authorities, target groups and end beneficiaries constitute the
three points of the triangle. Actors affected indirectly by policy (third parties constituting
either negatively-affected third parties or positively-affected third parties) are located at
the peripheries of two of these three poles (see Figure 5).
Insert figure 5 about here
To add concrete examples to this triangle, the analyst seeks to identify the ‘empirical
actors’ (in) directly concerned by the collective problem to be resolved, as well as the
hypotheses (often implicit) upon which the public initiative is based. In fact, here we
suggest that every policy can be interpreted as a theoretical construction whose
consistency and rationality must be questioned analytically, ‘… in the sense that it
implies an a priori representation of the measures implemented, of the actors’ behaviour,
of the sequence of measures undertaken and of the effects produced on society’’(Perret,
1997, p. 292). We describe this theoretical construction as a ‘model of causality’
(Knoepfel et al., 1998, p.74ff) or as a ‘theory of social change’ (Mény and Thoenig,
1989, p.140ff; Muller 1985, 1995). It comprises a causal hypothesis and an intervention
hypothesis, the analysis of which makes it easier to discern the links between the
different actors and the ways in which they are altered in the aftermath of public
intervention.
The causal hypothesis provides a political response to the question as to who or what is
‘guilty’ or ‘objectively responsible’ (i.e. without subjective guilt) or able to make changes
to enable the collective problem to be resolved. Thus, the definition of the causal
hypothesis of a policy consists in designating the policy target groups and the end
beneficiaries. This attribution of responsibility is still determined by political value
judgements and by the way in which the problem is perceived. Furthermore, uncertainties
of a scientific nature with respect to the effective (objective) functioning of the
intervention sector greatly limit the possibility of correctly identifying the target groups at
the root of the problem.
Thus, for example, within economic theory the individual causes of unemployment
remain matters of dispute. The way in which ecosystems function and the development of
anthropogenic phenomena within ecosystems (the generation of ozone, the greenhouse
effect, global warming) are still subjects of scientific debate. Similarly, the external
observer is partially ignorant of the intricacies of heroin trafficking and heroin use, as
well as of the profile and behaviour of drug addicts.
It is, therefore, essential for political-administrative authorities to know and understand
the factors and effects of social change and the collective problems that stem from these,
if they wish to make some impact on them and alter them. In order to do this, the state
often depends directly on information that the private actors control and partially
produce, especially the target groups at or close to the origin of the social problem to be
resolved. This functional dependence of the state upon certain target groups sometimes
allows the latter to present information from their own perspective and, as a result, to
designate the behaviour of other social groups as being the cause of the public problem
(i.e. to formulate a competing causal hypothesis). For example, for a long time farmers
who pollute underground streams by spreading fertiliser were able to evade compliance
with government regulations by arguing that the resultant deterioration in the quality of
drinking water was due primarily to adverse effects on the environment produced by
industry and private households.
The ineffectiveness and adverse effects of certain policies often derive from false or
incomplete causal hypotheses. The causality model that was for a long time favoured by
agricultural policy was based on the effects of subsidies on farmers’ incomes, ignoring
both the effects of overproduction of agricultural produce and the impact of intensive
farming on the environment. The model used in the area of policy dealing with road
congestion assumed that the increase in the provision of facilities for drivers (the building
of extra roads and motorways) would make it possible to resolve the problem without
taking into account the effects that these new facilities would actually have in real terms
(a correlative increase in the number of cars on the roads leading to the same level of
congestion). Similarly, the causality model used in public transport policy was initially
based on the premise that increased services and lower costs would be enough to
persuade drivers to leave their cars at home and start using public transport in large
numbers. In the context of Aids prevention policy, the disease was first regarded as
exclusively affecting homosexuals and drug addicts before being treated as a general
public health problem.
As for the intervention hypothesis, it establishes how the collective problem requiring
resolution can be mitigated and, indeed, resolved by a policy. It defines the methods of
government action that will influence the decisions and activities of the designated target
groups so that these will be compatible with the political aims. Thus the state can compel
them to change their behaviour (for example through the imposition of obligations, bans,
enforcing compliance with requirements for permission-granting schemes), induce a
change of behaviour by positive or negative economic incentives (for example, taxation
schemes, tax relief, subsidies), or again suggest it through the manipulation of symbols
and information (for example, campaigns to heighten public awareness of an issue,
training programmes). The effectiveness of each method of government intervention with
regard to the resolution of the collective problem depends, among other factors, on the
practical relevance of the behavioural hypothesis that underlies it. This process of pre-
emptive evaluation of private actors’ capacity to react to government intervention is,
however, contingent on the social structure of the target groups (Schneider and Ingram,
1993). In anti-drug addiction policy, the initiatives favoured by private and public actors
vary significantly, depending on whether the actors perceive drug addicts as deviants and
criminals who should be subjected to police intervention and legal sanctions, or as people
who are ill and who should, therefore, be provided with medical help and social
rehabilitation. Thus, it is up to the state to anticipate the possible reactions of the relevant
target groups if it wishes to modify their behaviour with some degree of predictability.
In order to ensure that such reactions can be forecast with a degree of predictability, the
public authorities generally participate in a (pre-parliamentary) consultation and
negotiation procedure with the parties concerned and/or adopt a participatory approach to
the implementation of the policy. These two strategies, whose aim is to legitimise and to
make state intervention acceptable to the target groups, end beneficiaries and third
parties, result in the ‘co-production’ of certain policies. At implementation level, this then
results in the fact that several tasks may be delegated to para-state or private
organisations; for example the management of milk quotas in France and in Switzerland,
the monitoring of the origins of bank funds (in Switzerland), the provision of
psychological and material support for the homeless in France and the supervision of the
conduct of doctors in the UK.
The suggestion that a policy is based on a – usually implicit, partial or indefinite – model
of causality (i.e. hypotheses of causality and intervention) derives from an instrumental
and rationalist interpretation of public intervention. This is naturally open to criticism. It
must be emphasised, however, that, even in cases where a policy has been adopted and
implemented for a reason other than the resolution of a collective problem (for example,
in order to affirm the power of the state in a symbolic way, for the purposes of electoral
competition, personal, organisational or institutional prestige, selective targeting of a
certain social class), the methods applied inevitably generate new framework conditions
for public and private actors, and the effects that derive from this can potentially affect
the course of social change.
Furthermore, as every policy plays a part in the definition of the relationships between
the three defined groups of actors, it is intrinsically redistributive in nature. This being the
case, it always brings about a change in the material and symbolic attributes enjoyed by
the different actors, by imposing costs (induced by changes in behaviour) on the target
groups addressed, and by granting privileges (linked to the improvement of their personal
situation) to the end beneficiaries of the public initiative. By objectifying these
redistributive effects between individuals and social groups, policy analysis also aims to
answer the traditional question put to political scientists: ‘Who gets what, when, how’
(Lasswell 1936). At a time when it is no longer simply a question of distributing the fruits
of growth, today, it would be necessary to add ‘from whom?’ to this question.
Up until now, we have argued that policies aim to resolve a collective problem, and,
therefore, to have an impact upon the process of social change and the way that this
evolves. In order to do this, the public actor must designate the target groups whose
activities (or passivity) may be one of the (in) direct causes of the situation that is judged
to be unacceptable from a political point of view. Having formulated this causal
hypothesis, the public actors must then apply – in accordance with the politically
conceived intervention hypothesis – certain initiatives and procedures that effectively
encourage the target groups to alter their behaviour. Consequently, a policy’s causality
model always constitutes a normative representation of the way in which society
functions and of the behaviour of private actors. Ultimately, its validity can only be tested
in the context of policy implementation and the subsequent evaluation of their effects.
This entire discussion presupposes, nevertheless, that a social problem has been defined
politically as a public one and that, consequently, a public intervention measure of a
redistributive nature is imperative. This precondition concerning the link between the
beneficiaries of public initiatives and political-administrative actors is not necessarily a
given in everyday reality (cf. Chapter 7 on the political definition of the public problem).
20 Note here that every group is always a social (and political) construct. Cf., with regard to this point, the example of
the typology of social groups developed by Schneider and Ingram (1997).
21 While, in conformity with contemporary practice in English, gendered pronouns have generally been avoided in
situations where the actor could be of either gender, there are situations like these where they have been kept because an
alternative formulation would substantially change the form of a translated quotation from French.
22 Just as a ‘non-decision’ expresses one of the forms of power (according to the thesis of Bachratz and Baratz,1963,
1970; cf. also Wollmann, 1980, p. 34), a passive attitude (‘non-action’) is one of the possible modes of behaviour that
policy actors may adopt. It should also be noted that some discussions of this topic, such as that by Lukes (1974) go
further to try to identify ‘interests’ that are entirely suppressed (see the discussion in Hill, 2005, pp. 33-5).
23 Article 5 of the Federal Law on Administrative Procedure of 20 December 1968 (LPA, RS 172.02).
