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 7: POLITICAL AGENDA SETTING
If, as defined in Chapter 2, public policies consist of a group of activities and decisions taken by
different actors with a view to resolving a problem that is politically defined as public in nature, it
is important first and foremost to look into the actual concept of ‘public problem’.
In this chapter, we discuss the processes whereby a social problem is identified and then
‘thematicised’ as a public problem as well as the different characteristics of agenda setting.
According to our analytical model, the political definition of the public problem constitutes, in
effect, the first product which the analyst must study when tackling the cycle of public
intervention in the context of an empirical study.
In reality, the processes involving the political definition – and redefinition – of public problems
have not been the subject of theoretical and empirical analyses that are comparable, in number at
least, with those carried out on the (subsequent) stages of policy programming, implementation
and evaluation. Obviously, the failure to take the stakes associated with the definition of the
problem tackled by a policy into account is a definite drawback when it comes to the analysis of
this policy, including studies carried out on its implementation:
Yet we know that the problem definition stage frames and generates virtually everything that
follows in the policy process, so our failures to examine problem definition sentences us to
operate through a glass darkly (DeLeon, 1994, p. 89).
As Anderson (1978, p. 20) and others demonstrate, the development of a policy cannot be
interpreted as the simple resolution of given problems defined on the basis of their intrinsic
characteristics. It also involves the constitution and definition of public problems. This is a
political designation process that influences or determines the actors involved in the policy (i.e.
the public authorities of the political-administrative arrangement, the target groups, the end
beneficiaries and third party groups), and the actual nature of the public actions carried out (i.e.
the modes of intervention selected in the political-administrative programme).
From this perspective and by way of introduction, we stress the need for a (re-)constructive
analysis of public problems (7.1). Then, in accordance with the structure proposed in Chapter 6,
we will move on to the general definition and operationalisation of the concept of the ‘public
problem’ (7.2). Finally, we identify the agenda-setting processes on the basis of the strategies of
the actors involved, their resources and the institutional rules which frame this first stage (7.3).
7.1 THE SOCIAL CONSTRUCTION OF PROBLEMS
In order to analyse a problem relevant to the public sphere and on this basis legitimise public
intervention, it is necessary to adopt a constructivist approach. In effect, it is reasonable to
assume that no objective fact constitutes a problem in itself (Dery, 1984, p. xi; Cobb and Elder,
1983, p.172). The – social and then political – definition of a problem always represents a
96
collective construction directly linked to the perceptions, representations, interests and values of
the actors concerned on an individual basis and/or as part of organised groups. Thus, all social
reality should be understood as a historical construction, situated in time and space. It always
depends on the constellation of the persons affected by the problem and/or those whose
behaviour may need to change to solve it.
It is important that the reader fully understands what we are saying here: it is never a question of
denying the objective conditions that constitute a problematic situation (for example, the
existence of an elevated level of carbon dioxide emissions which represents a threat to climate
stability, the permanent nature of the situation of a growing number of people seeking
employment whose material and psychological situation becomes precarious, the high frequency
of acts of violence in urban areas, the massive influx of war refugees at borders), but of stressing
that these established facts only represent one of the dimensions – even if it proves fundamental
in some cases – that constitute a social problem. Thus, the role of the policy analyst consists in
identifying the processes, actors and arguments by means of which these objective conditions are
perceived and are defined as problematic and requiring state intervention.
The adoption of a constructivist approach to the reality of social problems and policy has at least
three implications for the manner of analytical consideration – i.e. the reconstruction and
interpretation – of public actions (cf. Vlassopoulou, 1999, pp. 13-17) and, even more specifically,
the policy products.
1. The limits of the rationalistic approach: according to what is known as the
‘definitional’ approach (Spector and Kitsuse, 1987), it would not be possible to
compare public policies with precise action programmes. Such comparisons are,
however, quite common in the American research tradition. Action programmes
presuppose a prior and clear definition of the objectives of the public action in
question. If necessary, their objectives are externally defined (for example, on the
basis of objective criteria concerning the scope of the problem to be resolved), the
political challenge consisting solely in selecting the means that would enable their
realisation. This rationalistic vision of public policy, which was pushed to extremes by
the Planning, Programming, Budgeting System approach and other attempts to apply
scientific planning to policy processes, assumes that the problem to be resolved is
defined for once and for all and that the objectives of public actions are never
questioned. Without doubt, the action programmes concretise the intentions of the
legislator and the executive – possibly in the form that is most tangible for the analyst.
However, they may not be interpreted as independent of the policy of which they are a
part and, by extension, of a social situation that is collectively acknowledged as
problematic. In summary: if the analysis is limited to the examination of legislative
programmes (political-administrative programmes – cf. 8.1.), it is impossible to pose
or answer satisfactorily the question with regard to which social problem the state is
ultimately trying to resolve and why one problem as opposed to another is selected as
the basis of a policy.
Thus, for example, a ban on allowing public lighting to be left on after midnight may
constitute either a measure for protection against nocturnal air attacks (in the case of a
97
country at war), a measure to reduce energy spending or to reduce emissions which
represent a risk to the climate of our planet.
2. The limits of the sequential approach: the analysis of a policy on the basis of a
sequential model conceived in a rigid and strictly linear manner (agenda-setting >
programming > implementation > evaluation) tends to suggest that the definition of a
problem constitutes an individual and isolated stage in the policy process. If all public
policies are based first and foremost on the collective recognition and thematicisation
of a particular problem, it is necessary, nonetheless, to keep in mind that the
programming, implementation and evaluation stages are also based on the definition –
in the sense of a concretisation – or redefinition – in the sense of a modification – of
the public problem defined at the outset (Plein, 1994). The construction of a problem
is, therefore, an ongoing, non-linear and open process.
During a diachronic study of a policy cycle, at each stage in the cycle, the analyst
must pose the question as to the extent to which the content, actors and institutional
context of the policy being studied are associated with the initial definition of the
problem to be resolved or, conversely, are associated with its political reformulation
by the actors concerned (who, in some cases, were deliberately excluded from the
decision-making process at the outset of the public intervention). A sequential
approach to public policies – adopted in part here for its heuristic advantages – which
does not take this on-going process of (re-)definition into account risks obscuring not
only a fundamental challenge of all public action, but also one of the explanatory
factors of the eventual substantial and/or institutional changes to the policy along the
way. In effect, a redefinition of the problem to be resolved (for example due to the
revised interpretation of the initial objective data or the knowledge of new facts)
should translate into a change of policy and vice versa.
The evolution of air-pollution-prevention policy is one of the clearest examples of this
kind of dynamic (Weale, 1992). In the 1950s, this policy was aimed at the emission of
pollutants by households, industrial companies and emissions in urban areas. The
discovery of the phenomena of long-distance transportation of acid pollution
combined with the mandatory erection of high chimneys (to ensure better dispersion
of pollution and hence a low concentration of pollutants in the local air) noticeably
changed policy in this area to the extent that all sources of emission were considered
irrespective of their location – urban or non-urban. The dilution/diffusion principle
was then replaced by that of treatment at source which translated into the obligation to
install filters in industrial chimneys. In the 1980s, the same policy underwent a second
fundamental change due to the emergence of phenomenon of the greenhouse effect
and leading to the inclusion of mobile sources (cars which were formerly exempt from
environmental interventions) as target groups.
3. The limits of the sectoral approach: finally an approach centred solely on programmes
and their political-administrative structures tends to interpret policies in accordance
with a sectoral analysis framework. If such a public administration is in charge of the
programming and execution of a policy, the public problem at the basis of this
intervention comes under the responsibility of a particular sector, predetermined by
98
habit (for example, the logic of a profession and its individual members) and the
traditional area of competency of the administration responsible. By taking the
organisational affiliation of the administrative actors responsible for policy
programming and implementation as the only point of departure, the analyst neglects
the critical interpretation of the process of (re-)definition of a collective problem. In
effect, a social problem will generally not only concern a single sector of intervention,
but several (for example: atmospheric pollution may constitute a public health
problem as well as problems which are the concern of environmental protection,
transport, production and the consumption of fossil-fuel-based energies). Furthermore,
if the definition of a problem is interpreted as an evolutionary process, the
responsibility for its management within political-administrative structures may, with
time, shift from one sector to another. Not to take these institutional changes into
account risks limiting the interpretation of a collective problem to just one of its
dimensions and, hence, the failure to recognise the evolution of the public policy as
well as the problems of coordination that arise between the different public actors
responsible for its resolution (cf. internal co-ordination of a public policy and co-
ordination between several public policies).
7.2 THE POLITICAL DEFINITION OF THE PUBLIC PROBLEM (PRODUCT 1)
Having justified the necessity for a (re-)constructive approach, we must now define and
operationalise what we understand by the term ‘public problem’ so as to guide the analytical
(re-)construction of this first policy product.
7.2.1 The public problem: defining elements
Gusfield (1981) makes a clear distinction between ‘social problems’ and ‘public problems’,
noting that all social problems do not necessarily become public problems, i.e. the objects of
political controversy. Hence, public problems represent an extension of social problems to the
extent that, having emerged within civil society, they are debated within an emerging political-
administrative arena. In this sense, the definition of a public problem is essentially political in
nature. In other words, a problem is only public if it is already on the political agenda. At this
stage of the definitional process, public actors (for example, the administration, government,
parliament) recognise the need to consider a possible state solution to the identified problem.
Even more concretely, Garraud (1990, p. 20) identifies the three following conditions for
referring to a problem as public: (1) the constitution of a demand emanating from particular
social groups; (2) the development of a controversy or public debate; and (3) the existence of a
conflict between organised social groups and political authorities.
The assumption of the transfer of a problem from the social sphere to the public sphere on the
basis of a strict chronological model prompts the definition of public actors as orchestrators of the
agenda-setting process. Although this vision makes sense because it attributes a proactive role to
99
the public actors, it must, however, be stressed that the passage of a public problem is neither
linear nor inevitable.
On the one hand, as Vassopoulou (1999, pp. 19-20) notes, ‘a public problem may involve the
recuperation of a former social preoccupation as well as an original political construction’. The
specificity of a public problem consists, therefore, in the fact that it is placed under the
responsibility of the public authorities and not necessarily in the fact that the latter take up a
social problem that has already been clearly articulated.
On the other hand, several problematic social situations are never thematicised as public
problems requiring state intervention. Contrary to the pluralistic vision of democracy which
assumes that every actor may access the decision-making arena to thematicise a particular
problem, Bachrach and Baratz (1970, p. 6) assert that a specific form of public power consists
precisely in the possibility of keeping certain social problems off the public agenda. Described as
‘non-decisions’, these institutional blocks force the social actors concerned by a problem either to
find other points of access, or to manage the resolution of the problem themselves (using
corporative or private policies):
Non-decision making is a means by which demands for change in the existing allocation of
benefits and privileges in the community can be suffocated before they are even voiced; or
kept covert; or killed to gain access to the relevant decision-making arena (Bachrach and
Baratz 1970, p. 7).
We propose – for all empirical analyses – to study the extent to which a problematic private
situation is perceived and then defined as revealing of the social sphere and then the political
arena. This kind of analysis aims, amongst other things, to identify the possible pitfalls (i.e. the
different types of non-decisions) in the definition process (cf. Figure 11).
Insert figure 11 about here
The sociological approaches concentrate primarily on the individual factors, the collective
conventions and norms that favour or, conversely, curtail the realisation that a private
problematic situation may concern the social sphere and, consequently, be defined as a social
rather than private problem (for a discussion of this topic using UK examples see Hulley and
Clarke,1991).
The more political-science-based approaches mainly analyse the articulation of a problem
referred to as social, the resulting demands for public intervention and the approaches to agenda
setting adopted by the different actors concerned, i.e. private and public, individual and
collective. As Garraud (1990, p. 17ff) notes, any analysis of agenda setting which takes this
diversity of actors into account will by necessity find itself at the intersection of the disciplines
dealing, in particular, with political participation, (new) social movements, the media and
decision-making processes.
Essentially, as suggested by Figure 11, we only speak of a ‘public problem’ if a situation is
judged politically as problematic and is the subject of political debate. Beyond this very general
definition, we present below some dimensions on the basis of which ‘the political definition of a
public problem’ may be described and analysed.
100
It should be noted, however, that it will sometimes be possible to observe a development running
contrary to that described in Figure 11, whereby a problem initially identified as public (for
example, the legal prohibition of cohabitation in Swiss family policy) is subsequently considered
as solely relevant in the private sphere and cannot, therefore, be the object of state intervention.
Certain sectors historically considered as relevant to the public sphere are being gradually
withdrawn from traditional policies, other forms of regulation by the state (e.g. competition
policy) and/or private instances. However the privatisation of public services may involve the
replacement of public provision by extensive public regulatory activities.
7.2.2 The operational analysis dimensions
Policy analysis may identify several constituent elements of a public policy (cf. Wildavsky, 1979;
Gusfield, 1984; Rochefort and Cobb, 1993; Peters, 1998). Let us stress once again that if every
problem can be qualified on the basis of the dimensions discussed here, this assessment does not
solely depend on the objective conditions of the situation deemed problematic, but also on the
evaluation and subjective weighting (i.e. of the political definition) of the actors concerned. Thus,
in evaluating the four constituent elements proposed below, the analysis is trying above all to
explain the perceptions of the nature of the problem of those directly concerned by it.
Note also that the agenda-setting stage of a public problem is also something of an ‘art’ in the
sense that for the promoters of a particular policy it involves the combining of substantive aspects
(for example, the promotion of out-patient health treatment) with the more institutional
dimensions (for example, the reduction of fixed personnel and infrastructure costs in hospitals,
altering the roles of various organisations or actors within the health care system).
It is also necessary to stress in advance that the operational dimensions discussed above intersect
in part; their simultaneous consideration must favour a general interpretation of a public problem
rather than partial interpretations.
1. The intensity of the problem: the intensity of a problem refers to the way in which the
consequences of the problem are estimated at individual and collective level. In this
context, the actors involved judge whether it is a serious problem worthy of consideration
in view of the negative effects caused (examples: financial costs and psychological
distress caused by an increase in unemployment; risk of a nuclear accident; negative
effects of tobacco, alcohol and illegal drug consumption on health etc.), or a ‘pseudo-
problem’ which is too insignificant, according to key actors, to galvanise public opinion
and/or immediate intervention by the public authorities (examples from the past:
difficulty of access to public places for handicapped persons; disappearance of rare plant
species etc.). Note that the degree of intensity acknowledged for a given problem will
vary from one actor to the next and at different times.
2. The perimeter (or audience) of the problem: the perimeter of a problem consists in the
extent (or scope) of its negative effects in relation to the different social groups affected
and their geographical location and the development of the problem over time. The
identification of this perimeter necessitates knowledge of the circle of persons and/or
101
regions that are affected by the negative effects of a particular problem. Obviously, the
social and spatial borders of a problem may develop quickly over time.
This dimension of a problem is closely linked to its public visibility. In effect, if the
social groups affected are restricted in number, located in peripheral regions and/or
belong to social strata that are not highly politicised, the chance that their situation will be
defined as a collective problem is limited. Schneider and Ingram (1997) demonstrate
specifically that the manner of definition of a social problem (and the policy intended to
remedy it) depends on the (positive or negative) social image and the (strong or weak)
public power of the groups that are either touched by its negative effects or responsible
for its emergence.
Thus, it is possible to make a distinction between problems that are perceived as ‘clearly
defined/concentrated’ (for example: water pollution, drug consumption, suburban
violence) and problems perceived as being ‘without boundaries/diffuse’ (for example: the
risk of an epidemic in connection with ‘bird flu’, the human and ecological costs of a
nuclear accident, the problems associated with unemployment), and between problems
whose perimeter is developing rapidly (for example: the problems associated with AIDS
or natural disasters) or slowly (for example: the economic situation of farmers and
unemployed people).
3. The newness of the problem: while certain problems are new on account of their
association with the recent evolution of our post-industrial societies, others are more
chronic in nature. The degree of newness of a problem is often considered as a
determining factor in its capacity to access the governmental agenda. Downs (1973), for
example, suggest that a new public problem will succeed more easily in mobilising public
opinion and, hence, in prompting private and public actors to intervene. However, after an
initial phase of expansion, the problem in question will gradually relinquish its
importance and weight in terms of the governmental agenda in favour of more recent
problems.
The problems perceived as ‘new/recent’ include, for example, bullying in schools, genetic
engineering and environmental problems. Problems perceived as ‘old/chronic’ include,
for example, the problems of illiteracy, public security, price stability and unemployment
among persons without professional qualifications.
Note that new problems and, hence, completely new public policies are rare. While such
situations were still common during the phases of the emergence and affirmation of the
welfare state, there is now no choice but to accept that most public intervention is more
likely to involve the correction or re-orientation of a previous and (partly) failed policy,
or the integration of several previously separate policies.
4. The urgency of the problem: a social problem may be perceived as more or less urgent
in nature. In extreme cases, often associated with a shock arising outside the political-
administrative system (for example: an accidental cause), we speak of crisis management
with the state providing an (almost) instantaneous response to the social problem which
has quickly become obvious to all. Such contexts should open a ‘window of opportunity’
102
(an expression used by Kingdon, 1984) for the ‘policy entrepreneurs’ who wish to
promote policies to address that particular problem.
In the context of problems perceived as urgent, the following are worthy of mention: bird
flu., foot and mouth disease, AIDS, the oil crises of 1973 and 1979, the nuclear accidents
at Three Mile Island in 1979 and Chernobyl in 1986. The problems that are perceived as
non-urgent include, for example, the progressive degradation of the landscape and the
ongoing increase in cardio-vascular diseases associated with (professional) stress and
health issues (nutrition, tobacco, alcohol).
These four dimensions are not claimed to be exhaustive in terms of the operationalisation and
comparison of different social and public problems. The analyst may also try to qualify the nature
of a problem on the basis of its political complexity (several parties involved ↔ few actors
involved), programmatic complexity (several identifiable causes ↔ a single cause), its capacity
to be expressed in monetary form (costs of X million ↔ barely tangible and non-quantifiable
effects) or its degree of interdependence with other public problems (isolated problem ↔
interlinked problems) (cf. Peters, 1998).
Thus, it is possible to dissociate multi-causal problems, such as air pollution related to motor
vehicles, industries and households, from mono-causal problems such as the absence of low
water flow in water courses as a result of the production of hydro-electric power (Switzerland) or
water shortages as a result of leaks in the water transmission system (UK).
Similarly, it is possible to identify problems that can be expressed in monetary terms, such as the
net loss of X billion Euros due to corporate practices with respect to public markets, and
problems whose financial costs are not at all tangible such as racist attitudes among certain
groups.
Finally, it is possible to identify an isolated problem, such as that of inefficient weather
forecasting, from interlinked problems, e.g. unemployment which depends on complex links
between macro-economic, monetary, fiscal, educational and social security policies.
Without going into these complementary dimensions in more detail here, it should be said that a
public problem does not necessarily evolve into a problem of a certain type on the basis of its
intrinsic characteristics.
As stated above, the political definition of a public problem results from a symbolic battle taking
place between rival groups in an at least partly established institutional context. The dimensions
of political agenda setting in respect of a situation judged collectively as problematic are,
however, complex and worthy of particular attention.
7.3 THE AGENDA-SETTING PROCESSES
The following paragraphs aim to identify the explanatory factors behind the process of agenda-
setting. Initially, we suggest that this definitional process may be interpreted as a power struggle
whose main challenge is the recognition or imposition of an initial ‘hypothesis of causality’
103
which (pre-) structures the development of the future public intervention. We then identify the
actor constellations and means of action (resources and institutional rules) associated with
different ideal-type processes.
7.3.1 From competing ‘causal stories’ towards a dominant ‘hypothesis of causality’
According to the constructivist approach, all social problems – and to an even greater extent all
public problems – are collective constructs. Thus, the definition of the problem which a public
policy seeks to resolve emerges in the interaction between the actors concerned by a particular
situation. This process of definition most often consists in a power struggle between groups of
actors than as a consensual process accepted by all in civil society. The control of the process of
definition of public problems and, hence, of the alternative means of their resolution represents a
fundamental political challenge (Weiss, 1989), or even the supreme instrument of power
(Schattschneider, 1960, p. 66). As Stone (1989) states, different social groups are pitted against
each other as each tries to impose its own definition of the problem. In other words, the actors
concerned clash with each other in their efforts to become the ‘owners’ or legitimate trustees of
the problem (Gusfield, 1984, pp. 10-11).
This debate, which is generally conflictual in nature, is expressed in terms of competing ‘causal
stories’ advanced by the different groups of actors:
Problem definition is a process of image making, where the images have to do fundamentally
with attributing cause, blame, and responsibility. Conditions, difficulties or issues thus do not
have inherent properties that make them more or less likely to be seen as problems or to be
expanded. Rather, political actors deliberately portray them in ways calculated to gain support
for their side. And political actors, in turn, do not simply accept causal models that are given
from science or popular culture or any other source. They compose stories that describe harms
and difficulties, attribute them to actions of other individuals or organizations, and thereby
claim the right to invoke government power to stop the harm (Stone, 1989, p. 282).
Thus a ‘causal story’ is indicated by social group whose situation is deemed problematic; the
group in question also imputes responsibility for this politically unacceptable situation to the
behaviour of another social group (Edelman, 1988, p. 17). In this context, to define a problem
means to identify the groups who suffer its negative effects as well as identifying the origins, i.e.
which will often involve designating those responsible whose behaviour gives rise to the problem
and who must consequently bear the costs of its resolution. The causal stories create or delineate
particular social groups by a process of symbolic designation. In the course of this process, the
different actors manipulate images and symbols in order to strengthen their arguments (cf. in
particular Edelman, 1964, 1988; Schneider and Ingram, 1997). This is particularly evident in the
case of the theories supported by the extreme Right which deliberately attributes the causes of
unemployment, violence etc. to a particular social group, i.e. immigrants.
Furthermore, Stone (1989) suggests that public debate generally remains deaf to any excessively
complex interpretation or explanation and that (strategic) actors try to put forward simple causal
stories. In the majority of cases, the groups affected by a problem try to render credible an
‘intentional’ cause for their unfavourable situation: the social problem results from a considered
104
human action whose consequences are predictable and desired (for example: ‘My respiratory
problems are due to the fact that industries knowingly produce excessive pollutant emissions and
this gives rise to a deterioration in air quality’). If an argument of this type proves indefensible,
they aim to identify the causes as arising ‘through negligence’: the social problem is still the
result of a considered human action, however in this case the consequences are unpredictable (for
example: ‘I contracted the AIDS virus through a blood transfusion because the precautions
deemed indispensable by current medical knowledge were not taken’). Conversely, the groups
that are publicly defined as responsible for a collective problem advance causes referred to as
‘mechanical’ or ‘accidental’ to extricate themselves from all responsibility; in this case, they
suggest that the unpredictable and undesired consequences are the product of external events or
non-guided human actions (for example: aeroplane accidents were caused by an unknown
problem of a technical nature, for which the operating company was not responsible, and it was
not caused by pilot error or insufficient security controls; exceptional dry weather caused water
shortage not inadequate water conservation and transmission management).
From an analytical point of view, all causal stories have an empirical-cognitive dimension
(sometimes including a solid scientific basis) and a moral or normative dimension.
Hisschemöller and Hoppe (1996) combined these two dimensions to develop a typology of public
problems classified according to their degree of structuring. Thus, they suggest that, to the extent
that the state wishes to effectively resolve it, every type of problem involves a particular type of
policy (process) (cf. Table 4).
Insert table 4 about here
This typology is useful both to distinguish the processes of definition behind different policies
and the formalising of the links between the nature of the problem dealt with and the strategy
developed for state action and intervention. Nevertheless, its heuristic range is limited in the
sense that, as stated in Stone’s theoretical propositions, it focuses almost exclusively on causal
histories in its efforts to understand the process of (re-)definition of a social problem.
In referring us to the basic triangle of a policy (cf. Chapter 3.4), the concept of the ‘causal story’
proposed by Stone (1989) displays major similarities with the concept of the ‘hypothesis of
causality’ which we proposed for the identification of the target groups (required to solve the
problem and often its cause) and end beneficiaries (which suffer the negative effects of the
problem) of a policy. In this sense, what we qualify as a hypothesis of causality (and which is
found – at least implicitly – formalised in a political-administrative programme) is the causal
history that ultimately asserts itself as being the most plausible, based on the knowledge available
on the constitutive conditions of the social problem, and/or the most politically desirable one,
based on the interests and values of the actors involved in the definitional process. In terms of
empirical analysis, the (initial and often temporary) phase of political definition of a public
problem may be considered as complete when a causal hypothesis generates political consensus
or at least unites the majority of the actors concerned.
In the following paragraphs, we discuss some agenda-setting models for public policy. These
propositions identify different actors who mobilise and combine several types of resources, such
as the initiators or the ‘owners’, and engage in the debate on a particular public problem.
105
7.3.2 Process: actors, resources and institutions mobilised
The study of political agenda-setting looks into the factors which influence the fact that one
social problem attracts the attention of the actors concerned (and thus becomes a public problem),
while another is not subject to public debate or state intervention. In concrete terms, this involves
the definition of the actors and processes involved in the agenda setting. Up to now, no general
theory has made it possible to explain the constitution and definition of the policy agenda and the
ways that the ‘initiators’ of the debate on a social problem access it. On the contrary, a number of
partial models are proposed in the literature each of which describes a particular process within
the overall agenda-setting context. Without making any claim to their exhaustiveness, we present
below five ideal types of ‘agenda setting’.
a. ‘Thematicisation through media coverage’
The supporters of the ‘media coverage’ model highlight the decisive role of the media (in
particular the press, radio and television and the Internet) and polling institutes (for example
regular polls on problems identified as priorities according to the general public) in the
identification of a public problem. The media directly influence public opinion by placing the
emphasis on one or other social event, particularly in a crisis situation. This then prompts public
actors and the political parties to reappropriate the current topics of debate and to launch a
political debate and hence raise their profile among the public (McCombs and Shaw, 1972;
Gormley, 1975; Walker, 1977; Lambeth, 1978; Cook et al., 1983; Scheberle, 1994).
Political scandals in France and in the UK which have raised the question of the public financing
of political parties and the cases of corruption which led to the regulation of public markets may
be quoted in this context. Likewise, a policy can originate from the featuring of a pollution
catastrophe on the front page of the newspapers (for example: various shipwrecks leading to
severe oil spillages).
Information and communication infrastructure are, without doubt, the main resources mobilised
by (private) actors. Furthermore, access to and mobilisation of these two resources express
themselves as determining factors behind the actual structures and functioning of the media that
are active within a public body. Moreover, ‘time’ as a resource also appears to play an important
role here, because the time chosen to reveal information about a social problem to the public may
prove decisive for its inclusion in the political agenda, above all if a problem is presented as new
and urgent.
At the same time, it may be assumed that few institutional rules limit or promote the action and
the power of definition of the media: the latter can generally take advantage of the freedom of
expression guaranteed at constitutional level.
b. ‘Mobilisation’ or ‘exterior initiative’
Assuming that the policy agenda is constituted in response to clearly articulated social demands,
a number of authors suggest that the activities of pressure groups and/or (new) social movements
are a determining factor. As defenders of often general and long term social interests (for
106
example, environmental protection, the right to work, anti-racism), these organised actors try to
attract public opinion and the attention of political-administrative actors to the social problem to
be resolved using both institutional (for example: the popular initiatives and referendums in
Switzerland) and extra-institutional (for example: various demonstrations) means. In view of the
way that western democracies currently function, this model appears very plausible. It is
frequently adopted when the mobilisation of groups is associated with extensive public visibility
(Cobb and Elder, 1983; Cobb et al., 1976; Baumgartner and Jones, 1993). For example, the
strikes that are organised by public service trade unions and the occupation of nuclear sites by
ecologist ‘commandos’ like Greenpeace correspond to this phenomenon.
According to this model, the (private) actors primarily make use of the resources of ‘political
support’ and ‘organisation’, with financial and personnel resources naturally acting as
indispensable supports to the latter. What is often involved is the setting up of a (new)
organisation capable of asserting the interests and values of its members (for example,
establishment of a neighbourhood watch group, the creation of a coalition of charity
organisations). After this the actors try to thematicise a new problem by opposing an existing
policy or a current (infrastructure) project. Unlikely to have access to the resource ‘law’, that is
not enjoying the right to be heard during the consultation and co-decision procedures, these actors
concentrate on the creation, exploitation and combination of other resources such as
‘information’ and ‘time’ so as to delay or block a specific project.
From an institutional point of view, (new) social movements frequently avail of legal institutions
and/or direct democracy, in particular in the case of the popular initiative in Switzerland, to
thematicise (at a national level) a problem that was previously obscured (or perceived at local
level only). Although these institutional procedures are on the whole very expensive and
uncertain in outcome, they may represent the only means available to individuals and groups to
assert themselves as full actors in a defined sector. Finally, it should be noted that several private
actors free themselves from all institutional constraints by adopting strategies which although
illegal they considered morally justified; for example: call for civil disobedience motivated by
moral values in the case of the non-denunciation of illegal immigrants in France, the occupation
of a nuclear site or opposition to NATO military intervention in a foreign conflict). The actors
sometimes avail of the resource ‘violence’ in such situations.
Compared with the ‘mobilisation’ or ‘external initiative’ models, let us note finally that the
analyst may clearly identify in the case of some policies the existence at local level of ‘the
laboratories of emergence’ for a particular problem. The situation is subsequently thematicised at
regional and then (inter)national level. In contrast to this ‘bottom-up process’, ‘top-down’
processes also exist which introduce collective problems initially discussed at international level
to the (infra-)national agenda. This is particularly true of European Union directives, one of
whose impacts is to introduce public problems which were not hitherto acknowledged to the
national or local agenda (for example, the taking into account of air pollution caused by motor
vehicles as a result of the enactment of the ‘ozone’ directive in 1992). Furthermore, the
harmonisation of the political agendas at the level of all Member States is one of the major
impacts of the European Union (cf. in particular Mény, Muller and Quermone, 1985; Larrue
2000, p. 49ff., Larrue and Vlassopoulou, p. 1999).
107
In the British literature on environmental regulation this external influence is linked with another
influence (external to the original policy agenda if not to the country): privatisation. For example,
the co-incidence of the establishment of a privatised water industry requiring specific regulation
by government agencies with European directives on water quality together put water pollution
issues on the policy agenda and provided opportunities for pressure groups concerned with the
quality of drinking and sea bathing water (Maloney and Richardson, 1994; Jordan, 1998).
c. ‘Policy supply’ or ‘electoral competition’
Inspired by ‘public choice’ theory, the policy supply model assumes that political parties do not
just respond in a reactive way to social demands that have been articulated already. They may
also take the initiative. Hence they may define and formulate public problems with a view to
expanding their electoral base through the addition of the beneficiaries of the proposed new
policies. Thus, in this context, the policy agenda is constituted on the basis of the topics selected
by the main competing parties in their programmes and during campaigns. Various sub-variants
of this model are proposed according to which the confrontation between the parties is instead
expressed in terms of an ideological dimension (a situation of direct competition according to
Downs, 1957 and Odershook, 1986), or in terms of the selective declaration of certain topics, for
which one party has greater credibility among the population than another (the situation of
indirect competition according to Budge and Farlie, 1983; Klingeman et al., 1994). This is
particularly true in the case of the problem of immigration which is thematicised by extreme
right-wing parties in most European countries, the highlighting of the problem of unemployment
by the parties of the left and of environmental problems by the green parties.
It should be noted that these theories were developed to account for agenda-setting in democratic
regimes of the Westminster type, in which parties develop clear programmes of legislation which
they can realise with the support of a parliamentary majority (Hofferbert and Budge, 1992; Pétry,
1995).
Here, the resources mobilised by political parties generally encompass the resources
‘information’ (declarations of political programmes), ‘organisation’ (party apparatuses) and
‘political support’ (political majorities in power or the achievement of a governing coalition). As
the analysis of the financing of political parties and electoral and/or referendum campaigns show,
the resource ‘money’ also plays a decisive role in the capacities of the respective parties to make
their presence felt on the political scene and to lead the debate on a given social problem with
continuity.
Amongst the institutional rules that the political parties can use so as to assert their ideologies and
stances, the following should be noted in particular: the constitutional recognition of parties, the
instruments of direct democracy (popular initiative and referendums), the electoral rules
(majority versus proportional systems) and the informal rules associated, for example with the
Swiss system of concordance (for example: it will be easier for a problem thematicised by a
government party to be taken into account than a problem placed on the agenda by a non-
government party). Furthermore, it should be noted that the federalist rules or laws governing
decentralisation partly influence the internal organisation of political parties (for example: the
divergence between the slogans of political parties at national level in Switzerland and those of
their cantonal sections).
108
d. ‘Internal anticipation’
The model known as ‘internal anticipation’ (Garraud, 1990) awards a crucial importance to the
administrative actors and public authorities during the constitution of the policy agenda.
According to this model, these actors – who are already involved in the implementation of
existing policies – would be the best placed to identify the gaps between current state actions and
social problems that remain unresolved. Kingdon (1984, 1995) identifies what he calls ‘policy
entrepreneurs inside as well as outside the administrative system, perhaps a particular feature of
the fragmented system in the United States but by no means absent elsewhere. In the UK think-
tanks, often with links to the political parties and therefore partly ‘inside’ and partly ‘outside’ the
system, have played a key role in this process (Denham and Garnett, 2004). Evaluation reports on
the effects of a particular policy often constitute essential information support for this kind of
anticipation of immanent problems or those that already exist but have not yet been resolved. As
a result, the administrative actors propose changes to old policies and/or new intervention
strategies on their own authority. According to the supporters of this model, the internal
dynamics of the political-administrative sub-system are strengthened in particular in situations in
which a social problem is ‘badly’ articulated by civil society. In such cases, the public actors
replace private actors so as to appropriate and (re-)define the public problem to be resolved. In
doing this they also secure for themselves a new legitimacy as a useful organisation as well as
support for or extension of their area of competence and their resources (personnel, money,
knowledge etc.).
The launch of an awareness-raising campaign targeted at adolescents by the public health
authority on the topic of tobacco and alcohol addiction and the ban on advertising of these
products is an example of the launch of a new policy (sometimes in the absence of a legislative
mandate), similarly the problem of road safety is regularly raised in the media by the public
authorities.
In fact, all of the resources theoretically at the disposal of political-administrative actors may be
activated and combined in the case of ‘internal anticipation’. More specifically, however, it
seems that the targeted presentation of information (on a social situation that is deemed
problematic) and the privileged position of public actors in initiating a legislative or regulatory
process can prove as determining factors in explaining why a given problem ultimately reaches
the policy agenda.
Likewise, all of the institutional rules that define decision-making procedures potentially act as
procedural supports for the activities of public authorities. Thus, these range from the possibility
of establishing a commission of external experts to analyse the data concerning a particular social
problem to organising an (informal) consultation (pre-) procedure for certain concerned actors
etc. It should be noted here that the political-administrative actors are perfectly versed in the
institutional nuts and bolts of the public sector – particularly the informal ones – a fact that is not
necessarily true of certain private actors and more so if they are not organised in a pressure
group. Thus, the former potentially enjoy greater room for manoeuvre to assert their own
interests (and public policies) or those they represent.
109
e. ‘Silent corporatist action’
While the ‘mobilisation’ model concentrates on actions involving a high level of public visibility
on the part of pressure groups and (new) social movements, the ‘silent corporatist action’ model
analyses the more discreet role of interest groups in the context of the policy agenda. Defending
their own (often particular and short term) interests, these actors, who are very well organised and
benefit from networks of influence, seek direct access to the decision-making arena while
deliberately avoiding the thematicisation through media coverage and politicisation of the
policies they would like to either maintain, introduce or avoid at all costs. This kind of model is
based both on the neo-corporatist theories and on empirical studies (Baumgartner and Jones,
1993) which demonstrate that certain administrations and political authorities maintain
‘clientelistic’ relationships with various private or para-state actors. Cases involving the
agricultural sector, the oil industry, the construction sector and public works are the best known
in this context.
If the relationships between certain private and public actors are characterised as clientelistic, it
may be assumed that the resources ‘organisation’, ‘consensus’ and ‘political support’ are
essential to the maintenance of this corporatist equilibrium. In effect, an administrative authority
will be interested in negotiating bilaterally with a particular pressure group insofar as the latter
offers – in exchange for refraining from thematicising the problem to be resolved – its action
network (or a implementing para-state administration) and its support during the adoption of the
policy. In fact, the majority of resources mobilised will be mobilised to prevent the information
on the problem to be resolved from becoming (too) public or prevent the adoption of
(excessively) rigid legal solutions etc.
At institutional level, it is the informal rules that will logically influence the nature and content of
negotiations between public authorities and private pressure groups. From a more extreme
perspective, it may even be assumed that the actors involved will direct all their activities and
decisions towards the development of informal conventions or tacit agreements. This is done
expressly to prevent the other equally concerned actors from pursuing the option of approaching
formal institutions to access the policy in question (cf. 3.3.1.). This may be observed in relation
to issues like sustainable development and food policy, where companies make commitments in
efforts to pre-empt government interventions (Cahill, 2002).
7.3.3 Comparison criteria
None of the five models briefly discussed above are sufficiently differentiated and complex to
explain on their own the process of political agenda-setting associated with a particular social
problem. However, by combining these models, the analyst may find it easier to identify the
intermediaries to which the social groups (target groups, end beneficiaries and third parties) and
public actors may or should appeal to form coalitions at the stage of the political definition of
public problems. To facilitate this kind of comparison, Table 5 presents an overview of some of
the variables relevant for empirical studies on agenda setting. It also presents, in the form of
hypotheses, the main resources and institutional rules which actors mobilise to promote their
causes.
110
Table 5 about here
When we consider the different variables presented in the above table, it should appear obvious
that a detailed study of the process of the political agenda setting of a public problem must
comprise at least three stages:
1. An analysis of the actors involved: who is the initiator, the ‘owner’ and/or
legitimate possessor of the problem? What kind of causal history does the latter
propose? What are the competing causal histories proposed by the other actors?
2. An analysis of the thematicisation processes: which resources and which
institutional rules are used by whom and how successfully in order to access the
decision-making arena?
3. An analysis of the substantive content: what is the influence of the composition of
the actors involved and the agenda setting strategies adopted on the way in which the
(social, geographical and temporal perimeter of the ) public problem is ultimately
defined?
The simultaneous consideration of these substantive and institutional dimensions will facilitate
the general comprehension and interpretation of the process of the political definition of the
problem which ensures the passage of a so-called social problem to a so-called public problem.
7.4 THE DYNAMICS OF THE POLITICAL AGENDA: COMPETITION AND CHANGE
Up to now, we have discussed the concepts ‘social problem’, ‘public problem’ and ‘agenda
setting’ as though every problem would in itself constitute an independent entity. This postulate,
which is implicitly found in most of the case studies dealing with a single public problem (for
example, on AIDS: Rogers et al, 1991; on global warming: Trumbo, 1995; on economic policy:
Kleinnijenhuis and Rietberg, 1995) should not obscure the fact that several social problems are
always engaged in (in) direct competition when they aim to access the political agenda.
In effect, in view of the limited resources of the state apparatus and intermediary actors (for
example: the political parties, social movements, the pressure groups), the political agenda cannot
process all of the problems articulated by civil society at the same time and with the same
intensity. Thus, there is competition between social problems, and some end up being relegated to
the queue for inclusion on the list of priorities for public action (for example: environmental
questions are often seen as a luxury dealt with during a period of economic expansion), while
others are completely eliminated from the democratic debate (cf. concept of ‘non-decisions’). In
the framework of an analysis of the agenda setting process for a particular problem, the
researcher should bear in mind not only the factors relating to the problem studied, but also the
more general conditions relating to the other social problems simultaneously expressed by civil
society, in order to explain the access or non-access to the political/policy arena of the problem
(Crenson, 1971; Cobb et al., 1976; Hilgartner and Bosk, 1988).
111
Besides taking into account all of the public problems that constitute the political agenda, it is,
finally, also necessary to highlight the importance of a diachronic interpretation of the agenda-
setting process. Several different hypotheses have been advanced in this context too.
Baumgartner and Jones (1993) suggest that public problems and the policies developed to remedy
them do not always evolve in an incremental way. Short periods of radical change alternate with
long periods of marginal adaptations or the status quo. This has led to the development of a
model described as ‘punctuated equilibrium’ which is summarised by Vlassopoulou (1999, p. 29)
as follows:
Every actor system (policy venue) constituted around a problem and/or challenge
represents a particular perception of this problem (policy image). As a result, the diffusion
of an alternative vision by a new actor becomes a major element of instability: in asserting
itself, it changes, not only the definition of the problem, but also the composition of the
actors' system. Thus, each actor configuration reflects a particular definition of the
challenge, which suggests that they should not be considered independently: a change of
definition is supposed to produce a change of the system of actors and vice versa. In this
case, the combination of two elements should be capable of explaining both the stability
and rapid change of policy.
According to this approach, by correlating the definition of a public problem and the actors
mobilised around, the analyst should be able to explain a change in the policy agenda. Such an
approach has been applied to the process of ecologisation of agricultural policy in France and
Europe (cf. Larrue 2000).
Rose and Davies (1994) present a rather different hypothesis. According to them, once in power
all governments only have very limited room for manoeuvre when dealing with new public
problems. Having inherited programmes initiated by previous governments, a new political
majority mainly allocates public funds to the policies which have already been launched. This
phenomenon of inertia is even more accentuated by the routines and strategic games of political-
administrative actors who try to preserve their prerogatives. Thus, the new programmes primarily
constitute an extra layer that is added to the programmes that are already ‘sedimented’. As
Lascoumes (1994, p. 334) also suggests, no public problem is developed on virgin territory. Any
change in the definition of a problem and, to a greater extent, of a public policy is undertaken
with direct reference to the existing situation which influences the systems of thought, the actor
constellations involved and the action strategies. Thus, according to this second approach,
political inheritance is a determining factor.
By way of example here, we can quote the sedimentation of local authority agriculture policy
which has actually led to an accumulation of subsidies granted to farmers (starting with
production subsidies and ranging to hectare subsidies and subsidies for ecological practices).
A third approach, owing much to increasing global influences on policy deriving not only from
actual global initiatives but also from the increasing awareness on the part of national
government of what others are doing, stresses the role of ‘policy transfer’. Policy makers adopt
solutions to policy problems from other countries (Dolowitz and Marsh, 2000, Dolowitz et al.,
2002). International organisations (OECD, the World Bank etc) play roles in this process and the
literature suggests that wholesale and uncritical policy borrowing will often be problematical.
112
More usually the transfer process involves considerable adaptation under the influence of the
other factors examined in this chapter.
In conclusion, all policies are formed ‘step by step’ and over time, starting with multiple
retrospective actions and collective learning processes. During the initial development of a public
response to a problem to be resolved, the causal model often proves to be mono-causal and
partial. Due to lacunae in these initial causal and intervention hypotheses, the programming of the
policy (or the products PAP and PAA) often remains inadequate or at least incomplete. The
evaluation of the effects actually arising from its implementation allows the actors involved to
learn some lessons and readjust their objectives (i.e. to politically redefine the public problem).
During this new cycle of the policy, it is thus confirmed that the causal model fills out as also do
the institutional elements which flow – in part already – from the sedimentation of successively
applied rules in the course of the preceding cycle for the attribution of a given resource and given
position to a given actor. Thus, all empirical analyses of the political definition of public
problems should clearly identify the cycle in which the policy being studied is located. The
degree of substantive and institutional differentiation of the initial policy product to be explained
depends to a decisive degree on this identification.
113
