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 9: POLICY IMPLEMENTATION
The analysis of policy implementation involves the researcher and practitioner in what
often constitutes the most complex and rich part of a public policy process. In effect, this
phase of the ‘policy life cycle’ brings about a direct relationship between the public
actors of the political-administrative arrangement (PAA), the target groups, the end
beneficiaries and third-party groups (positively and negatively affected third parties). In
this analysis considerable attention is given to target groups. Where the distinctions
between target groups, beneficiary groups and third parties are less clear, or not
applicable to the argument the loose expression ‘affected groups’ is used to encompass
all groups.
The following sections deal, firstly, with the concept of ‘implementation’ (9.1) from the
perspective of classical theory on policy implementation and recent developments in this
area (9.2). From the policy analyst’s perspective what is involved at this stage is the study
of the interaction between these actors so as to reach an understanding of the in situ
functioning or failure of a policy.
Secondly, we define the products of this stage of the policy process from an operational
perspective, i.e. the action plans (9.3) and administrative products (9.4). By analysing
certain types of implementation processes, we then indicate the relationships between the
different types of actors involved that need to be subject to a relational analysis so as to
facilitate the identification of the factors behind the success and failure of a policy (9.5).
9.1 DEFINITION OF POLICY IMPLEMENTATION
It is necessary to start by providing a precise definition of the ‘implementation’ of a
policy as this term is used both widely and selectively in different countries and/or
schools of political and administrative science.
Many American authors use the term ‘implementation’ to designate all activities
involving the execution of a piece of legislation. This term, which is also adopted in
Germany by the Forschungsverbund Implementation politischer Programme (Research
Association for the Implementation of Political Programmes) (cf. Bohnert and Klitzsch,
1980), incorporates into the concept of ‘implementation’ all of the political-
administrative processes that follow on from parliamentary decision-making. Thus, this
definition includes48 the development of all the elements of political-administrative
programmes (PAPs) contained in governmental decrees which are not subject to separate
parliamentary decisions.
In contrast to the above-described approach, the definition of implementation that we use
does not include all of the post-parliamentary phase which extends to the end of the
decision-making process within the legislative assembly up to the concrete and individual
decisions taken by the administrative authorities. Instead, we define policy
implementation as the set of processes after the programming phase that are aimed at the
136
concrete realisation of the objectives of a public policy. Thus, our definition does not
cover the regulatory provision – by government and administrative authorities – of PAPs
and PAAs by means of decrees or simple governmental decisions. In our opinion, it is
necessary to make an analytical distinction between the ‘programming’ and
‘implementation’ stages of a public policy which both take place, in part at least, during
the post-parliamentary phase (or ‘after a bill becomes a law’ as the Americans say).
Despite the partly similar thrust of these two stages of the policy life cycle, the actors
actually often pursue different strategies during each (Kaufmann and Rosewitz, 1983, p.
32). In fact, the distinct phase of the execution ‘in the field’ of the legal and regulatory
norms that govern state intervention can be found in all public policies, irrespective of the
nature of a country’s institutional system.
Thus, our definition of implementation is very close to the notions of execution or
application that are generally used in the daily language of public administration.
However, implementation defined in this way does not only involve the production of all
kinds of juxtaposed concrete activities (in particular administrative acts); it equally
incorporates all of the planning processes necessary for the organisation of these
executive activities or action plans which define priorities – in time, in space and vis-à-
vis social groups – for the application of the PAP.49 In effect, having provided structures,
resources and coordination procedures between public and private actors, policy
effectiveness depends upon the establishment of an adequate mechanism of planned
production.
In summary, we define implementation as the set of decisions and activities:
•carried out by the public, para-state and private actors who belong to the PAA;
•framed by a PAP (either by a set of legislative and regulatory rules that are
more or less flexible and favourable to the interests of the different actors)
governing the specific institutional rules of the policy in question, and
•carried out with the intention of creating, influencing or controlling:
1. the constitution of a ‘policy network’ (as defined by Clivaz, 1998)
which structures the contacts between the formally competent
administration (PAA), other stakeholder administrative services, target
groups, end beneficiaries and third-party groups;
2. the conception of an ‘implementation strategy’, adopted within action
plans (product 4 as specified in the introduction to Part III) which take
into account updated analyses of the public problem to be resolved, in
particular its social (or functional), spatial (or geographic) and
temporal distribution. Increasingly, the production of these plans is
formally organised by means of codified planning exercises
(examples: plan for the elimination of waste in accordance with the
Law of 1992 in France, plans for the conservation of natural areas in
accordance with the Habitats Directive in France, ‘measurement plans’
137
to reduce atmospheric pollution in accordance with the Swiss Decree
on the Protection of the Air of 1985 in Switzerland); plans for health
and social services in the UK and
3. the (preparatory work for) concrete, general or individual decisions
and activities which are directly addressed at target groups (outputs:
products and/or administrative services; product 5 as specified in the
introduction to Part III), such as, for example, administrative operation
permits for industrial installations.
Based on this definition of ‘implementation’, which is applicable to all policies in all
political systems, the three types of activities identified above enable the analysis of the
substantive and institutional content of the implementation of a public action. This
approach also takes the two following points into account.
Firstly, the implementation process is only completed with the production of decisions
and activities that directly address those affected (outputs). Thus, it is always
characterised by interactions between public and private actors. These exchanges often
take the form of negotiations, with the interests and positions of social actors sometimes
being relayed by non-governmental organisations or sectoral pressure groups. Contrary to
the programming phase which may unfold (exceptionally) without any direct contribution
from (certain) civil society (groups), by definition, all of the stages of policy
implementation address the external actors of the political-administrative subsystem.
Therefore, the products and administrative services as themselves (outputs) serve as a
point of departure for the evaluation of the conformity of implementation with the
substantive content of the PAP.
Secondly, no PAP, even the most concrete and well developed, can replace or entirely
predetermine implementation by claiming to be ‘self-executing’. In explicitly limiting the
real import of the PAP to an essential but not exhaustive framing role, the definition
proposed here stresses that from the specific perspective of implementation, the
legislative and regulatory norms contained in the PAP present themselves as a more or
less relevant set of ‘rules of the game’ which distribute the positions and resources of the
different actors involved in the ‘implementation game’ (expression coined by Bardach,
1977). Even if these rules are clearly established, the game itself always remains to be
played. Furthermore, a number of implementation processes depend on the forms taken
by interactions between actors of a particular policy who are relatively independent
because they are organised on the basis of general institutional rules (applicable to all
public policies) and/or other concerns of a regional or local nature. The arrival of a new
piece of legislation may modify pre-existing policy networks or PAAs, for example by
introducing a new actor excluded by the past or giving this actor more political weight
than before. This is the case, in particular, with the introduction of a new actor into the
game through the attribution of the right of appeal to environmental protection
organisations or consumers in related policies. It should always be kept in mind that
policies that create completely new activities and policy networks from scratch are rare.
If not forging a new path for itself, a recent policy must take account of or integrate itself
into the existing organisational structures, procedures and relationships between pre-
138
existing actors (cf. notion of ‘path dependency’ proposed by institutional economics – cf.
Chapter 5.1).
9.2. ‘CLASSICAL’ IMPLEMENTATION THEORY AND RECENT DEVELOPMENTS
‘Classical’ implementation research starts with the observation of ‘implementation
deficits’: these deficits were traditionally demonstrated by comparing the products and
real effects of a policy (outputs and impacts or outcomes) and the substantive elements of
PAPs that describe the policy objectives and instruments. In this regard, the interest of the
researcher is close to that of the administrative-political actors (in particular at
federal/central level), who consider such deficits as an expression of a failure at the level
of implementation and not as raising questions about the intervention instruments and
objectives formulated – in the majority of cases – by these actors during policy
programming (PAP).
It should be noted, however, that nowadays the analytical concepts used in
implementation research have moved far beyond the mere adoption of the formerly
predominant doctrine of legislators and lawyers who considered implementation as a
purely executive or technical function. As we know from numerous empirical research
projects, this old conception overestimated the capacity of the law to influence and
control administration and its action with respect to the development of civil society. In
fact, it was actually the rejection of this over-determination by the law, a previously very
widely-held belief among politicians and civil servants and also among lawyers, which
gave rise to the impetus for research on implementation (Knoepfel, 1979, p. 23). From
the outset, empirical studies were used in an attempt to provide a more nuanced
explanation of the quality of policy products and effects. The five following dimensions
were generally adopted as explanatory factors in such studies:
•the ‘structure of programmes’ to be implemented (or PAP according to our
terminology): the notion of the structure of a programme is not limited to the
simple substantive content of policies. On the contrary, the first research
projects carried out in this context (Mayntz, 1980; Knoepfel and Weidner,
1982; Weidner and Knoepfel, 1983) endeavoured to characterise the different
possibilities for expressing the intention of the legislator (for example: degree
of precision, scope for manoeuvre given to federal/centralised or decentralised
administrative actors) and the different modes of intervention adopted (for
example: obligations or bans, financial incentives, information etc.) or the
institutional arrangements predetermined in this legislation (cf. 8.1 on the
PAP). A related UK debate has been concerned with the arguments for and
against rules and discretion in the structuring of legislation (Adler and Asquith
eds. 1981, Baldwin, 1995).
•the ‘implementation administrative system’50 (or PAA in our terminology): the
recognition of the characteristics of different political authorities and
administrative organisations responsible for policy implementation as a factor
that explains the quality of state intervention is without doubt the main
139
innovation achieved by the ‘classical’ research on implementation. Thus,
these approaches stressed the necessity of conceiving administrative
organisations as groups of actors participating in a network of interactions
incorporating a multitude of actors (‘interorganisational or intergovernmental
networks’). After that it was suggested that in many cases implementation
processes are propelled through the management of these policy networks.
Thanks, inter alia, to the contributions on the theory of collective action
(Olson, 1965) and the sociology of organisations (Crozier, 1963), these
analyses have not been limited to public actors but have included the
interactions between administrative agents and interested social groups (cf.
8.2 on the PAA).
•the economic, political and social weight of target groups: the capacity of a
subject of the administration to influence a state intervention varies according
to the composition of the socio-economic, political and cultural fabric of
which that subject is a part. Thus, this position may be dominant or marginal;
it may be supported by other social groups or, conversely, rejected by other
public and private actors. Furthermore, the target groups of the instruments of
a policy may be members of a pressure group whose cooperation is necessary
for the realisation of the policy in question and/or other policies considered as
more important. Of course, this situation strengthens their position with
respect to the intervention of the policy-executing public administration. From
another perspective, the weight of a subject of the administration varies in
accordance with the responsibility that public opinion and public actors
attribute to that subject in relation to the public problem to be resolved.
Finally, this position will also vary according to his/her location:
independently of the above-mentioned factors, the capacity for negotiation of
a target person will be strengthened if this person is located outside ‘high
pressure problem zones’ and, conversely, it will be reduced if they are situated
within a zone considered as highly problematical (for example, an industrial
enterprise will find it difficult to negotiate reductions with respect to a clean-
up order in the context of atmospheric pollution if it is located in a zone in
which pollution levels clearly exceed the admissible values).
•related issues about the power of target groups within the public policy
delivery system: there is an issue here about the extent the state is able to
influence the behaviour of its own functionaries. In this respect the literature
on implementation connects up with a wide literature about control within
organisations (see discussion in Hill, 2005: chapter 10). An important concern
within the implementation literature has been about the behaviour of staff at
the delivery end of policy systems. Furthermore the ‘street level bureaucracy’
literature (the seminal work here was by Lipsky, 1980 but the ideas here are
further explored in Hill, 2005: chapter 12 and in Hupe and Hill, 2006b)
suggests reasons why a measure of autonomy may – and even should – be
present at this level, particularly when complex professional services are
involved. In this sense the targets of policy changes may be changes of
behaviour within the system. The whole issue is made more complex,
140
particularly in the context of New Public Management, by the extent to which
tasks are delegated to organisations that are not directly managed by policy
makers. Hence, we may see issues about targeting, and about the strategies
used to resist targeting in similar terms as those where the target groups are
clearly private and external to the system, and issues about the ‘weight’ of
such target groups (note, for example, doctors) as having many similarities.
• The position of target groups vis-à-vis public activity is also influenced by
variables referred to as ‘situational’ which may extend or reduce their room
for manoeuvre during negotiations concerning implementation. Such variables
are constituted by external events, changes in the economic and social context,
disruptions of an economic nature etc. which occur independently of the will
of the actors involved. For example, the increases in the price of petrol have
strengthened the position of public actors implementing energy-saving and
rational use policies vis-à-vis the target groups which have shown a certain
reticence in regard to such processes. Similarly, in the context of synchronous
comparisons in particular, the position of actors will vary in different
implementation zones in accordance with variables referred to as ‘structural’.
In effect, research has demonstrated the existence of relatively constant
determinants with respect to the relative influence of actors, such as
institutional rules governing property relationships or constitutional political
rights (e.g. excluding foreign populations from the policy arena).
The presentation of these five issues affecting the quality of the services provided by
public policies and, in principle, adopted in the approach presented here in a modified
frame also demonstrates the comparative nature of the process adopted by classical policy
implementation research: in order to evaluate empirically the explanatory contribution of
each type of factor, it was necessary to examine the implementation processes carried out
by different administrative actors in several locations from the outset of these processes.
The choice of these ‘empirical sites’ took into account either the different quality of
services (outputs or variables to be explained) or differences with respect to four types of
explanatory variables.
In the 1980s, ‘classical’ research on implementation was re-examined and accused of
overestimating policy at the expense of ‘politics’, that is power struggles and the more
general characteristics of the development of post-industrial societies.51 Writers interested
in the sociology of law (Treiber, 1984) also formulated several criticisms, for example,
highlighting the danger of the blinkered state with respect to the – sometimes positive –
role played by implementation deficits (cf. on law Müller, 1971, p.53 and p. 98). Apart
from this external criticism, the teams carrying out research on implementation
themselves made two significant changes to their initial conception of the policy
implementation.
•From ‘top-down’ vision to ‘bottom-up’ approach: the majority of authors agree
with the criticism initially expressed by Hjern (1978), Hjern and Hull (1983) and
also Barrett and Fudge (1981) which claimed that the ‘top-down’ perspective
renders researchers blind with regard to the social or political processes that are
141
not foreseen by PAPs. The latter need to be analysed, and should then not be
considered as simple obstacles to effective implementation. In effect, the
supporters of a bottom-up approach consider these socio-political processes as the
expression of the fact that one and the same PAP may inspire hope and different
reactions and strategies based on the interests of the public and private actors
involved. It is not always a question of unwillingness when a local actor, who is
subject to all kinds of social, economic and political-administrative constraints,
uses or instrumentalises a programme in a way that was not planned by the
legislator.52 For example, it has been confirmed that Swiss civil protection policy
(subsidies for the construction of anti-nuclear shelters) has on occasion been used
to fulfil the aims of local authority sport, cultural etc. policy (construction of
multi-purpose gyms and entertainment centres).
In its attempt to clarify the behaviour and real motivation of these actors, the ‘bottom-up’
movement proposes that all research on policy implementation should start from the
system of basic actors (cf. Table 7). The analyst must, therefore, concentrate on the real
behaviour of the target groups of a public policy and on the real interaction between the
latter and the actors actually involved (PAA, end beneficiaries and third-party groups).
The definition of the implementation proposed here takes this criticism into account when
it speaks of the uniquely framing role of PAPs53.
Table 7 about here
•From the concept of ‘implementation’ towards that of ‘implementation game’:
following an analysis of different implementation processes which emphasises
their many facets, Bardach (1977) proposed a reorientation of research on the
execution of public policies. Anticipating the criticism levelled at the excessively
‘governmentalistic’ nature of the ‘top-down’ model, in his work entitled ‘The
Implementation Game’, he proposes the game metaphor to ‘attract the attention of
researchers to the players, to what the latter consider as stakes, to their strategies
and their tactics, their ways of entering into the game, the rules of the game
(which stipulate the conditions for winning) and the rules of ‘fair-play’ (which
stipulate the limits beyond which you enter the domain of fraud and illegality)’.
Bardach (Ibid.) also insists on ‘the characteristics of the interactions between the
players and on the degree of uncertainty with respect to the possible result of the
game (‘outcomes’)’ (Ibid., p. 56). This metaphorical concept should make it
possible, among other things, ‘to draw the attention of the researcher to those who
do not wish to participate in the game, to the motives for their abstention and to
those who demand that rules of the game be changed before they will enter it’
(Ibid., 56).
This conception of implementation indicates that policy actors are involved in a
competitive process not only with respect to the distribution of resources placed at the
disposal of those responsible for application processes, but also with respect to the actual
distribution of the power to define and redefine the rules of the game in the course of
these processes. The (public or private) actors who succeed in (re)defining the rules of
the game and who, as a result, is able to use a policy to serve their own interests possibly
142
in cooperation with other authors, are capable of exerting a significant influence on
policy implementation. The main challenge, therefore, consists in occupying such a
position as the person who defines, interprets and modifies the institutional rules of the
game (‘fixer of the game’).
Thus, this approach, which was adopted in part in our theoretical framework (‘indirect
games’ – Chapter 6.2), tends to relativise the ‘ideal-type’ image of implementation as a
processes based on the intention of implementing actors, the substantive content of a
PAP, to abandon the conceptualisation of programming and implementation as two
clearly distinct stages and to subject to re-examination the ‘legitimate supremacy’ of
public actors over private actors – both economic and social.
The definition of implementation proposed here only partly takes the logical
consequences of Bardach’s approach into account. Despite its unquestionably original
nature which draws the analyst’s attention to all of the actors (and not just the public
actors), this approach seemed too ‘American-centrist’ to fully encompass the realities of
certain European democracies. In effect, the latter’s traditions with respect to the
primordial role of the public function and content of substantive and institutional
decisions taken democratically at the level of policy programming (PAP and APA) are
very deeply rooted. In the United States, in contrast, the ‘rule-making’ process (i.e. the
production of rules by independent agencies in accordance with a procedure of legal
inspiration) is located between the ‘law-making’ (or classical ‘programming’) and
‘implementation’ (i.e. classical implementation). This last statement must not, however,
be considered as an invitation to strictly limit the analysis of implementation processes to
this last phase. On the contrary, it is increasingly accepted today that the understanding of
implementation processes also incorporates the analysis of the programming phase. From
this perspective, we discuss below the action plans (product 4) which constitute a direct
link between the PAP and the results of its application (outputs, product 5).
Although our approach does not limit itself to a simple analysis of policy implementation,
it does adopt a considerable part of the traditional concept of ‘implementation research’
while generalising the elements adopted at the level of the entire policy life cycle and
bringing them back to the three basic elements, i.e. actors, resources and institutions.
When applied to the analysis of the implementation stage of a policy, as a factor
explaining services, in our conceptualisation, the programmes of a certain ‘structure’ to
be implemented become the substantive elements of products number 2, 3 and 4 (PPA,
PAA and AP) which guide the direct game of the actors. The old ‘administrative
implementation system’ covers the institutional elements of these three products which
are likely to direct the indirect games of the actors participating in the production of
outputs. The third factor considered by the traditional approach as decisive for the results
of policy implementation, i.e. ‘the economic, political and social weight of target groups’
corresponds to the explanatory (intermediary) variable in our approach, i.e. in the – direct
or indirect – games of actors (target groups, end beneficiaries and third parties) and their
capacity to mobilise their resources in accordance with institutional rules specific to the
policy or institutional rules concerning administrative organisations, or general
institutional rules.
143
This conceptualisation does not neglect the variables referred to as situational or
structural because the latter become supplementary or increasingly scarce resources for
the actors (for example, political support more important for consumer protection
movements after the case of hormones in meat, loss of political support for the same
movements in periods of economic crisis due to general decline in consumption), or
institutional rules that are more or less favourable to the interests of the actors in question
(right of constitutional initiative open only to Swiss citizens, freedom to publish free
opinion limited to those with access to the media).
9.3 ACTION PLANS (PRODUCT 4)
We define action plans, which are not necessarily observable as distinct formal elements
in all policies, as the set of planning decisions considered as necessary for the
coordinated and targeted production of administrative services (outputs) and which, in
certain circumstances, may already be partially established by the PPA. Thus, action
plans define priorities for the production of concrete measures and for the allocation of
resources necessary to implement administrative decisions and activities. As an
intermediate stage in the policy execution process between the PAP and the – sometimes
unplanned – appearance of implementation acts, action plans are increasingly being
adopted in policy processes as real management instruments.
Thus, many federal laws in Switzerland or national laws in France, and financial
interventions by the European Union in the area of structural funds, necessitate
arrangements set up at infra-national levels such as the cantons, regions or departments
which they develop in advance for the granting of subsidies and establishment of
programmes or plans. This is the case in the area of economic development (development
of a regional development plan by the regions in the frame of European policy of support
for the regions), in the area of infrastructure (for example for the construction of national
and cantonal roads and universities) and also (or increasingly) in the area of the
environment (for example in the protection of the air, the efforts to combat noise
pollution, water management, waste disposal) and in social policy (to facilitate the co-
ordination of health and social care). This kind of planning approach has a longer
tradition in the area of spatial development (for example: structure plans). Finally, it
should be noted that the ‘service contracts (or mandates)’ developed as part of New
Public Management processes can be related to action plans in the sense that we define
them here.
The analysis of these planning exercises on the basis of dimensions adopted from legal
analysis (see Table 8 below) makes it possible to distinguish these action plans from
certain other intermediate implementation acts. According to our approach, the action
plan should be interpreted as an individual product which makes it possible to establish a
link between the general and abstract norms of the PAP and the individual and concrete
implementation acts. The same plan may be distinguished from the intermediate
implementation acts of certain policies (in particular those concerning land) which are
144
defined as general and concrete products (but sometimes have the same name, such as a
land use plan).
Table 8 about here
What all of these planning activities, which are traditionally found in the context of
infrastructure and management policies, have in common is that they precede the
production of final implementation acts. Thus, they aim to define temporal sequences and
to answer the two following questions: How is the problem pressure (distributed) and – as
a result – to which target groups and end beneficiaries of the state action should priority
be given? What is the cost of implementation and, consequently, which political-
administrative authorities must dispose which resources? (cf. Flückiger, 1998).
This definition of priorities and lesser concerns may be undertaken from a functional
(according to the type of activities subject to state intervention), temporal (according to
the short, medium or long term), spatial (according to geographical zones) or social
(according to the boundaries of socio-economic groups) perspective. The definition of
priorities, in the case of regulatory policies in particular, is often associated with the
decision to accept partial implementation deficits among certain target groups for a
certain period. The nature of privileges – and conversely discriminations – which derive
from this may depend on political pressure, the accentuation of the problem in certain
sectors and regions or the provisions of the PAP itself (cf. for example the empirical
research on social discrimination specific to measures for the reduction of traffic in Swiss
towns carried out by Terribilini, 1995, 1999).
It should be noted here that action plans sometimes remain internal to the public
administration. They define certain priorities for action without being legally binding
with respect to the conduct of state intervention in civil society. In other words, action
plans do not create subjective legal rights. Even if the situation of a social group is
considered explicitly as a priority in an action plan (e.g. young and unqualified
unemployed persons, long-term unemployed), this social group may not take advantage
of a law to benefit from public services proposed by the PPA ahead of other social groups
(e.g. older and qualified unemployed persons). In this sense, action plans allow public
actors to manage the implementation of the policy in question better without, however,
legally connecting the plans to the private actors that the policy addresses.
Nevertheless, the tendency to introduce open planning activities to the target groups, end
beneficiaries and third parties can currently be observed, in particular in France and the
UK. Several public action intervention plans in the area of economic development,
environment, spatial development etc. are now subject to the mandatory consultation of
stakeholders in the framework of more or less open procedures. This trend is associated,
among other things, with increases in the complexity of the PAAs of the policies
concerned and the demand by target groups and end beneficiaries to be allowed to
participate in the definition of policy implementation priorities. In this case, the value of
these action plans sometimes lies in their ‘opposability’ and they can hence result in the
greater or lesser restriction of the production of implementation acts. Thus, the current
trend involving the formalisation of implementation activities through planning leads to
145
and/or originates from a desire to (re)define intervention priorities, the affected groups or
even the causality hypothesis of the policies in question, in particular on the part of the
infra-national levels. This demand is supported in the European Union by the increasing
mandatory establishment of formalised action plans (in case of policies linked to
structural funds) so as to compensate for its weaknesses in following up the
implementation of its policies.
Figure 17 provides a schematic view of the discriminatory logic of all action plans. It
shows, in particular, that an action plan defines priorities for implementation and, as a
result, cannot generally incorporate the entire field of application of the substantive
objectives of the PPA and will actually lead to predicted and legitimated implementation
deficits. Moreover, the concrete measures produced do not necessarily cover the entire
field of implementation as defined in the action plan. Thus, it is possible that contrary to
the action plan, a particular social group and/or geographical zone is not affected by the
policy measures (unpredicted deficits) during the defined temporal period. During the
empirical study of the boundaries of final implementation acts (outputs), the analyst
should identify whether eventual implementation deficits are the result of the targeted
discrimination of the action plan or its inadequate application.
Insert figure 17 about here
146
The action plans need to be based on sufficient knowledge of the temporal evolution of
implementation activities, thus making it possible to ensure that administrative resources
are used efficiently. This is the case if the measures selected and resources allocated
make a greater contribution to the resolution of the public problem while taking the
willingness of target groups where applicable to change their behaviour into account. The
economic consideration of the marginal utility of public action plays a decisive role here.
In effect, the question arises as to how the state may attain maximum efficiency with
respect to the allocation of resources for the resolution of a problem (in the sense of the
Pareto optimum or Hicks-Kaldor optimum) by investing the given means. Thus, it is in
many cases a question of identifying the target groups whose change in behaviour will
contribute in all likelihood to the resolution of the public problem. Secondly, it may be a
question of establishing which target groups display the greatest willingness to change
behaviour (for example: the sectors in which a new investment cycle unfolds in a way
that enables the target groups to support a change in their behaviour economically).
Thirdly, it is necessary to identify the potential target groups for which the administrative
costs of intervention are the lowest in relation to the expected effects. Thus, the planning
of actions also includes the consideration of their effectiveness (relationship between
effects and objectives) and their efficiency with respect to allocation (relationship
between effects and resources) (cf. Chapter 10 on policy evaluation criteria).
In practice, experience has shown that the designation of groups that are will be
discriminated against as opposed to privileged groups is of primary importance.
Observation of the execution of national policies by regional authorities have revealed
that these policies are often implemented in fits and starts. As a general rule, initial
efforts are concentrated on target groups whose change in behaviour makes it possible to
anticipate the maximum possible contribution to the resolution of the problem at the
lowest possible administrative cost. Thus, in environmental policy, for example, the
primary target is major industries while small and medium-sized enterprises are
temporarily ignored at the price of an implementation deficit. Similar priorities are also
defined in the framework of noise pollution protection policy (according to the size of the
population exposed to noise pollution) and public transport policy (depending on the size
of new populations linked up). In view of the importance of the sometimes very profound
discriminations made by actors in this phase of the policy process, it is important to
observe the extent to which these discriminations are made visible and politically debated
(note, for example, the way in which this element in the implementation of the Child
Support Act in the UK – in which, despite the political rhetoric but for administrative
reasons, absent parents who were most willing to pay were targeted – contributed to
opposition and ultimately the need for further policy change).
So as to facilitate and systematise the empirical study of an action plan, we recommend
that the analyst take the five operational elements explained below into account. By way
of introduction, we emphasize that these dimensions, which are applicable to all action
plans, are not exclusive but complementary. In fact, the substantive and institutional
content of this product can only be taken into account through their simultaneous
analysis.
147
a. Explicit or implicit action plans
As already indicated, an action plan is generally presented in the form of an internal
document which does not legally link the political-administrative authorities to the
affected groups. Irrespective of this principle, an action plan can vary in terms of its
degree of explicitness. The fact that it is declared ‘official’ and is familiar to all of the
(public and private) policy actors should positively influence its degree of realisation.
Conversely, it can be assumed that an implicit, or (quasi) secret, action plan will only
represent a minimal constraint with respect to the different administrative services
responsible for its co-ordinated and targeted application.
Nevertheless, it is interesting to note that nowadays it is possible to observe a very real
trend involving the formal organisation of these action plans as an element for the
creation of a consensus during the effective implementation of public policies
(production of outputs). The degree of formalisation of the planning activity is, therefore,
a relevant criterion for the analysis of these plans.
b. Open or closed action plans
Along with the degree of formalisation, the degree of openness of the planning activity
requires analysis. The establishment of a priority plan between different areas or social
groups may be subject to internal actor activity or, conversely, it may be open to all of the
public actors concerned, i.e. target groups, end beneficiaries and third parties. As
mentioned above, this opening up of the planning process in a way which facilitates
political debate is increasingly common.
c. Action plans that are more or less discriminatory in nature
By definition, an action plan may define priorities and hence be discriminatory in nature.
The scope of this discrimination may be large or small and relate to temporal,
geographical and social dimensions. Thus, it may clearly establish implementation
deadlines or reference periods for the production of implementation acts (for example
annual plans versus ones that run over several years), deadlines which ensure a certain
predictability (or pseudo ‘acquired rights’, for example in the case of subsidy policies) to
the affected groups defined as being in a position of priority. Similarly, the borders of
geographical areas designated as priority areas may be more or less clear or fluid, more or
less stable or fluctuating. Finally, if the discrimination operates on the basis of the socio-
economic characteristics of the affected individuals, the criteria for eligibility may also
prove very strict or, conversely, leave a significant margin for assessment by the
implementing actors. For example, plans for the allocation of subsidies to disadvantaged
regions are generally highly discriminatory while plans to provide aid to companies are
far less so.
d. Action plans with extensive or limited (re)structuring of the PAA
An action plan primarily aims to concretise the scope of the substantive content of the
PAP of a given policy. To do this, it should also define precisely - as central actors - the
actors that participate in the implementation PAA. If the application of the legislative and
148
regulatory programme involves the de facto collaboration between different
administrative organisations, para-state organisations, private actors and various public
levels (central, regional and local), it is generally necessary to identify the actors who
must coordinate their actions for the purpose of targeting implementation, i.e. to
constitute a special policy network (cf. 9.1, first stage of implementation). Viewed from
this institutional perspective, an action plan may be interpreted as more or less selective
depending on whether it structures the interactions between the responsible actors (for
example in the form of a ‘project organisation’) – beyond the elements already defined in
the initial PAP and PPA – or, conversely, depending on whether it does not define any
specific new organisational and procedural rule for the implementation PAA.
e. Action plans involving a more or less clear allocation of resources
Finally, the development of action plans is guided by strategic consideration of the
impact and efficiency of the policy in question. Theoretically, it is a question of investing
the available or newly allocated administrative resources (for example, a supplementary
budget and additional personnel) so the public problem is resolved to the maximum
possible extent. This implies that action plans clearly establish which resources are at the
disposal of which actors for the purposes of policy implementation. This kind of resource
allocation processes is not so much an exercise in the optimisation of costs as a process of
political arbitration between the interests of the different actors (for example, a struggle
for resources between administrative services, decision-making levels etc.). Thus, the
analyst should examine whether a particular action plan formally links the use of a
particular resource by a given administration to the implementation of a given measure
among a given priority group or whether, conversely, no explicit link is established
between the use of available resources and the fields of action designated as priority
areas. Obviously, the action plan is considerably less effective in the latter case.
For example, in Switzerland, the planning of anti-noise measures, sections of national
roads or cantonal bypass roads (which are subsidised by the Swiss Confederation)
specifies priorities and makes financial resources available in accordance with the
provisions of the medium-term financial planning. Similarly, in France, the intervention
programmes of the water agencies associate temporal, spatial and target-group priorities
with specific financial sums. On the other hand, many action plans contain no link
between the planned measures and the allocation of finances (for example in Switzerland:
the cantonal measurement plans to counteract atmospheric pollution merely contain a list
of measures and fail to make any statement on the financial resources necessary for their
implementation or the sums reserved for this purpose).
In conclusion, it should be noted that an action plan is only a real policy management
instrument if it is explicitly formulated, makes clear discriminations within a given
temporal framework, structures the tasks and competencies assigned to actors in the PAA
and links the allocated administrative resources (in a supplementary way) to specific
decisions and activities. Needless to say, such action plans promote the coherent and
targeted management of policy implementation. On the other hand, they tend to make
visible to all actors the (temporary) discriminations actually being made. As a result, their
formulation presupposes that the responsible political-administrative actors legitimise the
targeting of the public action and this involves acceptance of the high political costs
149
while taking into account the imperatives of consensus. In fact, the phase of the
development of action plans rarely takes place within the legislative assembly, but at the
level of the executive authorities, allowing them greater scope for manoeuvre (for
example, if the action plan remains implicit, a referendum or political control are out of
the question).
We would finally like to note that in many European countries, administrative
reorganisation has led to the issuing of so-called ‘service contracts’ or mandates by
parliaments, governments and even ministers. These measures, which are concluded
between different administrative bodies or between public, para-public and private (or
privatised) actors may be considered as action plans to the extent that they operate
choices with regard to the desired services (and, more rarely, in practice, undesired ones).
They allocate resources, in particular finance, in the form of budgets to implementing
actors and stipulate the decision-making mechanisms for the population and the
evaluation of services (cf. Mastronardi, 1997; Knoepfel and Varone, 1999 for
Switzerland; Gaudin, 1996 and Warin, 1993 for France; cf. also 12.2.2). In the UK this is
particularly in evidence in the field of the contracting out of services by local
government, backed up by the requirement of local authorities to secure ‘best value’ not
only in terms of low cost services but also customer satisfaction. Criticisms have been
expressed, however, with respect to the lack of indicators concerning the effect of the
stipulated services.
Another response to the issues explored here is the setting up of monitoring arrangements
to deal with situations in which general action plans impose broad and complex
implementation responsibilities and it may be considered that it is outcomes rather than
outputs that really matter – lowered levels of pollution, decreased illiteracy, raised health
standards. A characteristic of modern implementation control in the UK has been the
extensive use of inspectorates and of the Audit Commission to audit overall performance
rather than to ensure exact rule compliance. This has become a key element in central-
local government relationships (Power, 1997; Pollitt, 2003).
9.4 IMPLEMENTATION ACTS (PRODUCT 5)
We define the formal implementation acts (outputs54) of a policy as the set of end
products of the political-administrative processes which, as part of the scope of its
implementation, are individually aimed at the members of the defined affected groups.
The final acts comprise administrative products directly aimed at the affected groups by
the administration and the other (private and para-state) bodies responsible for the
execution of public tasks. These products comprise all kinds of decisions or
administrative acts55 (for example, conditional authorisations, individual bans, approvals,
etc.), the granting of financial resources (for example subsidies, fiscal exonerations), acts
involving the collection of money (for example, indirect taxes, levies, fines), police
intervention, direct services (for example, health checks, financial checks, training or
treatment services), advisory activities and organisational measures. In reality, these
formal acts are accompanied by a large number of informal acts (see below).
150
The implementation acts are characterised by the fact that they create an individual
relationship (in particular in the case of formal acts, even those of a legal nature) between
the persons that constitute those affected and the competent public bodies responsible for
policy implementation. Procedural administrative law then grants the affected groups
rights or obligations of participation and cooperation. It is important to remember that,
given their limited resources, administrations rarely succeed in (punctually and/or
simultaneously) serving or sanctioning all of these groups in relation to the policy in
question. In such cases, one refers to temporal and/or regional implementation deficits
which arise – at least in part – from the adopted action plans (cf. 9.3.).
The implementation acts may be understood at different cumulative levels. The lawyers
look into some of the individual activities mentioned above and then examine whether
and to what extent these take up the provisions of the PAP and the action plans
(examination of legal conformity). The policy analyst normally relates to more widely
encompassing units composed of several outputs; the term ‘product groups’ has been
used to designate this phenomenon. Experience has shown that these product groups are
not easy to identify; they generally have to be reconstructed by the analyst which
involves an autonomous and ambitious definition task. It is possible to apply different
models of categorisation for this purpose. Despite these difficulties, as a general rule, the
conscious definition of implementation acts or products/product groups is particularly
relevant in the context of the reorganisation of administrations. In effect, it helps the
administrative services to focus their structures and procedures more directly at the
quality of the products and the requirements for their production (organisation of the
administration on the basis of products rather than laws –> cf. reorganisation of
administrations by ‘service project’). Furthermore, the identification of these sets of
output products may contribute to the reinforcement of the collective identity (‘corporate
identity’) of the administrative service or para-state body concerned. Finally, these
definitions make it possible to quantify and justify products better vis à vis political
instances. Thus, they make it possible to improve the management of policies in
accordance with the needs of their own products (Schedler 1995, pp. 13, 47 and 127).
The public bodies play the main role in the formulation of formal implementation acts
(outputs) and the fulfilment of the results expected by the policy depends to a large extent
on them. The executing bodies can have a significant influence on the concrete
application of a policy, depending on the use they make of the margin for interpretation
and action that the policy allows them. Thus, it is not surprising that these acts attract the
interest of monitoring bodies and that numerous laws require the presentation of activity
reports that make it possible to identify the formal implementation acts in both qualitative
and quantitative terms. This obligation to inform facilitates the monitoring of policy
execution. The reports can act as a basis for the actual monitoring of execution activities.
The forms this reporting takes may strongly influence the possibility of benchmarking
between administrative units in different regions and, hence, restrict the room for
manoeuvre (Knoepfel, 1997b).
The activities of executing bodies also play a key role during the implementation of
regulations that address private actors directly and only require subsidiary intervention on
the part of the public bodies (for example: monitoring of the respect of a norm and
151
application of sanctions in cases of abuse). In effect, the level of respect for a norm
depends essentially on the level of control exerted and eventual sanctions applied to
ensure that it is complied with (for example: police control of the observation of speed
limits and fines in the case of contravention).
It is generally necessary to collect a large volume of data in the course of the analysis of
implementation acts. Depending on the nature of the question being posed, it is usually
necessary, for example, to obtain information on the existence, quantity and quality,
spatial, temporal and affected group distribution, substantive and institutional content of
all administrative acts that are produced during the period being studied. The
implementation acts of other public policies must also be documented insofar as they
may have an influence on behaviour of the affected groups and the evolution of the
problem to be resolved. In cases involving a high frequency in the production of
implementation acts, it is recommended that they be represented in graphical form (for
example, spatial and/or temporal profiles of outputs based on the extent to which the
empirical activities are multi-faceted). In order to proceed with the development of
implementation output profiles for a specific public policy, the analyst may take the
following six dimensions into account:
a. The perimeter of financial implementation acts (=existence)
This dimension establishes a direct link between the priorities defined in the action plan
and the acts actually produced. In this context, the analyst should establish a
correspondence between the temporal and socio-spatial distribution of observable acts (in
the form of profiles of all of the outputs produced over time and in space) and the
intentions expressed in the action plan. By carrying out a comparison of this nature, the
analyst will be able to identify the eventual implementation deficits already planned –
and legitimised – at planning level (complete implementation even if it does not cover all
areas or there is objective problem pressure) from those resulting from incomplete
implementation. In the light of the original top-down use of ‘deficit’, implying fault on
the part of implementers, this usage draws a useful distinction between intended and
unintended deficits. There is also a possibility lying between these two, that is differential
implementation associated with the need for the exercise of discretion at street-level as
highlighted in the work of Lipsky (1980) noted above. In addition to this direct
comparison of the perimeter of the planned acts (in the action plan) and the real acts, the
analyst may put the latter into perspective with the area initially covered by the norms of
the PAP and thus discover implementation situations with gaps or distorted perimeters.
In other words, the analyst may identify missing acts with the help of a comparison of the
profile created for actual implementation with that required under the PAP and the action
plan.
For example, in Fribourg (Switzerland), traffic reduction measures were initially
concentrated in the affluent neighbourhoods, despite the fact that these neighbourhoods
were not necessarily exposed to the highest level of disturbance caused by traffic (noise,
pollution, accidents) (Terribilini, 1995); this constitutes a distorted output from the level
of the action plan. In contrast, a consistent output exists in the case of the taxation of all
152
natural persons and legal entities within the territory of a given public body. Moreover,
cases involving the renunciation of public intervention due to political pressure or lack of
resources (personnel, finance etc.) are examples of ‘holey’ profiles in which specific
target groups are omitted. Other examples of this include the dropping of police checks
during the carnival period and the failure to pursue the producers of so-called ‘soft’
drugs.
b. Outputs with a greater or lesser level of institutional content
Just like the PAP or the PA, beyond its substantive aspect (How is the problem to be
resolved?), the formal implementation act (output) may also contain institutional
elements that can be greater or lesser in scope (Which actors are supposed to contribute
to the resolution of the problem? Which rules of the game do they adopt? Which
resources do they use?). In effect, in order to resolve a collective problem sustainably, a
growing number of policies are being provided with control and monitoring mechanisms
which require the establishment of a network of actors for this monitoring process. Thus,
it is possible to observe the establishment of multi-party monitoring commissions,
monitoring bodies with strategic committees composed of a multitude of actors
representing the public services, target groups and end beneficiaries and the
establishment of communications networks for the interpretation of measurement data
associated with the effects of a service (for example: verification of the data forecasted
and recorded during an environmental impact study on a major infrastructure project
when the infrastructure in question is up and running).
Again and in accordance with the central research hypothesis proposed in this book, these
institutional elements will pre-structure the subsequent phase of the policy in question,
i.e. the eventual formulation of an evaluative statement on the impacts and effects of the
policy (cf. Chapter 10). For the most part, it is the actors who are fearful of having lost
out in terms of the direct game concerning the substantive content of the implementation
act who will be interested in obtaining – through an indirect game – a key position within
a particular monitoring structure by intervening with regard to the institutional content of
the output in question. For example, one of the institutional responses to the LULU
phenomenon (‘locally unwanted land uses’) associated with waste in France was the
establishment of a local information and surveillance commission for each waste
processing installation (Commission locale d’information et de surveillance – CLIS)
which is open to the entire implementation PAA, the local populations and environment
associations.
c. Formal or informal acts
Once the outputs actually produced have been recorded with the help of profiles,
empirical studies can examine the degree of legal formalisation of these administrative
activities. This operational dimension is important because it affects the production costs
(higher for formal outputs), the forms of negotiation and the institutional content of an act
which influences its follow-up. In effect, an administrative decision often includes an
appeal clause which opens the final (legal) phase of its production and its follow-up to
some or all of the affected groups. It should be noted that for the reasons cited, in many
153
cases, the administration prefers informal acts of implementation (for example,
information provided by telephone that is not formally noted and cannot be the object of
an appeal given in response to a request as to whether an application for a subsidy,
planning permission etc. has any chance of succeeding in a formal procedure) to heavy-
handed formal implementation acts.
d. Intermediate or final acts
In the above paragraphs, we considered final acts as one of the two products of the
implementation stage. However, certain acts are only intermediate in form (acts
undertaken within the administration that do not yet address the affected groups: general
and concrete acts – for example, draft spatial development plans – requiring further
individualisation to render a policy operational). These intermediate acts are sometimes
important in helping the analyst to understand the nature of the final acts which directly
address the individuals who compose the target groups, however, it is important not to
mistake them for final acts.
e. Coherence of the content of a policy’s final acts
The groups affected by a public policy are often subject to several public interventions
simultaneously having a combined influence their behaviour. Thus, it makes sense to
analyse the extent to which the different acts produced within one and the same policy
appear coherent and capable of mutually reinforcing each other. In effect, a lack of
coherence strongly reduces the substantive effect of a policy.
An elevated level of coherence exists, for example, when the obligation to clean up an
industry is accompanied by financial support for the clean-up process, when mandatory
health insurance is accompanied by the subsidy of premiums for low-income households
or the elimination of locations where drugs are taken openly is accompanied by the
establishment of injection centres and programmes for the prescription of methadone or
heroin under medical supervision. Conversely, the imposition of a speed limit on the
motorways in the absence of the monitoring of violations or the imposition of mandatory
sorting of waste without providing individual or collective systems for the collection of
sorted waste are indicative of weak coherence.
f. Degree of co-ordination with the acts produced by other policies
The question of the coherence of implementation acts is more obvious when these acts
involve several policies. Here too, inadequate co-ordination during the implementation
phase, e.g. as a result of an excessively closed PAA or action plan that does not
sufficiently structure activity, is unlikely to succeed in bringing about the desired change
in the behaviour of target groups.
For example, the granting of a subsidy for the updating of energy systems in buildings
(energy policy) during a period of economic recession (economic policy based on a
programme for the revival of the building sector) is indicative of a significant degree of
external co-ordination, while the parallel payment of subsidies for agricultural production
(which encourage the use of natural and/or chemical fertilisers) and introduction of
154
measures to combat water pollution (caused by the spreading of agricultural fertilisers) is
indicative of a weak level of co-ordination.
The collection of exhaustive data on implementation acts presupposes a clear and existing
definition of those affected. In accordance with the principles of the rule of law, the
affected groups are normally defined in the PAPs as both persons (natural and legal) to
whom the authorities may grant rights or impose obligations in order to attain policy
objectives, or whose behaviour should be influenced by other administrative activities.
When analysing implementation profiles, it must always be kept in mind that policies not
‘only’ produce implementation acts, they also produce effects. Thus, analyses that are
limited to formal or informal acts (outputs) and do not take impacts and effects into
account are incomplete and run the risk of confusing the services and effects of a policy.
For this reason the following chapter will basically deal with the impacts and results
(outcomes) of a policy which are normally the object of an evaluation (Chapter 10). And
in anticipation of this fourth stage of the policy life cycle, in the following paragraphs we
discuss the interaction between public actors, target groups, end beneficiaries and third-
party groups during policy implementation.
9.5 PROCESS: THE MOBILISED ACTORS, RESOURCES AND INSTITUTIONS
For obvious reasons the policy analyst must take a detailed look at the situation of target
groups because, ultimately, the success or failure of a solution proposed for the resolution
of a public problem often depends on (the change in) target-group behaviour. The
methodological approach adopted to identify the interactions between the actors of the
APA, the target groups, end beneficiaries and third-party groups is that of so-called
rational analysis. This approach tries to locate each type of actor in the social
environment and to identify the political, social and economic forces to which the actor is
exposed and, moreover, the intensity of the relationships that develop between the
different actors on the basis of their interests and respective resources. Thus, what is
involved here is the examination of the interdependencies between actors and the possible
coalitions that form between some of them. Figure 18 presents in very schematic form the
main relationships between the actors involved in the implementation of public policy.
In the following paragraphs, we review the main relationships and coalitions that can
exist or be formed between the actors (based on the numbering indicated in brackets in
Figure 18), and examine the potential influence of the latter on the products and effects of
the policy in question.
Insert figure 18 about here
155
1) Bilateral relationships between the administration and target groups
The close relationship between policy-implementing administrations and the target
groups arises from the simple fact that the implementation acts are aimed at various
natural persons and legal entities whose behaviour is intended to be modified by the
policy measures in question. In democratic societies legislation governing administrative
procedures generally recognises the principle of the right to be heard.56 Based on this, in
practice, numerous processes have been developed for exchange and interaction between
the political-administrative authorities and the target groups (exchange of information
and resources, negotiations, barter). Thus, the target groups tend to emerge as the
privileged negotiating partners of the public actors of a policy. Supported by numerous
special rules, this right to be heard has become an important asset at the disposal of target
groups in the course of the implementation game.
It should be noted that up to the late 1960s57, the axis between the administration and the
target groups formed the essential (and almost exclusive) bridge between implementing
actors and social actors. The exclusivity of this exchange was often reinforced through
professional confidentiality which guaranteed the ‘cosy relationship’ between the
administration and its ‘subjects’. This formula conceals the fact that in many policies
these ‘subjects’ only represent a small proportion of the (potentially) concerned citizens.
Thus, the affected group of spatial development is constituted by property owners, that of
economic policy by companies likely to benefit from public investments etc.
This exclusive relationship was re-examined from the early 1970s when the participation
movement started knocking at the doors of the closed chambers where negotiations
between the administration and target groups were being held. The battle was fought in
the context of several specific policy areas (e.g. spatial development policy,
environmental protection policy, consumer protection policy) and it was increasingly
156
successful in having a legislative impact. ‘Third parties’ (as the legal expression goes)
entered the political-administrative scene: parliaments and the courts started to grant
others affected the status of a ‘party’ in administrative procedures as well as a right of
appeal in litigation procedures. This opening resulted in the development of a triangular
negotiation structure which can be referred to as the ‘basic triangle’ between the
administration, target groups and end beneficiaries (cf. Chapter 3.4). This triangular arena
quickly replaced the old bilateral target group/administration relationship and
increasingly infiltrated all policy implementation structures and processes. The
emergence of this triangular structure significantly altered the position of target groups
vis-à-vis public actors: it was actually weakened, in most cases because the other ‘parties’
often represented interests contrary to those of the target groups.
(2) The power relationships within the target groups: competition and self-monitoring
One of the fundamental limits of the systemic or neo-corporatist neo-Marxist approaches
to policy analysis stems from the fact that they often conceive the target groups as a
homogenous group whose members are motivated by identical interests. However, social
and political reality tends to prove the opposite: in effect power relationships of a
competitive nature tend to exist between the target groups of the majority of public
policies (for example, between the different companies targeted by work safety policies,
social policy, economic policy or environmental protection policy). If it is true that at the
policy programming level these groups join forces to form relatively homogenous
associations and interest groups, this solidarity is often replaced by a self-monitoring or
competition mechanism at implementation level. In legal terms, this self-monitoring,
which is motivated by (economic) rivalry between the different sub groups of the target
groups, has taken the form inter alia of appeals lodged by competitors against decisions
targeted at a specific company and, more recently, in the right of appeal in the context of
matters of public acquisition (in accordance with the rules of the World Trade
Organisation). In cases of this nature, the competitor may cite violation of the principle of
the equality of treatment in claiming that the conditions granted to another company are
more advantageous than those imposed on it.
Thus, the position of natural persons or legal entities belonging to the target group will
equally depend on the competitive relationships in the different implementation arenas
(cf. 9.2 on classical implementation theory). Monopolistic or oligopolistic structures will
be more favourable for them than a nuclear structure. Furthermore, the degree of internal
rivalry affects the position of a target group in its entirety vis à vis the administration
which, in the case of a monopolistic structure, can hardly expect the intervention of these
self-monitoring mechanisms.
(3) Conflictual relationships between target groups and end beneficiaries: from state
arbitration of conflicts to contractual solutions
In the majority of policies, the target groups are not identical to the end beneficiaries
(note: the exceptions to this discussed at the end of section 3.3.3.). The end beneficiaries
are defined on the basis of the public problem which the policy proposes to resolve
because it is they who ultimately benefit from the proposed solution.
157
Thus, the objective of the provisions regarding the classification, size or design of a
building, defined in the form of conditions pertaining to planning permission, is not to
cause trouble for the owner, but to protect the building’s neighbours and to contribute to
the harmonious development of a village or town in the interests of all of its inhabitants.
Similarly, the issuing of an order to a company to clean up an installation in the interest
of work safety or environmental protection primarily aims to protect the workers or those
living near the site of the installation in question, or, in the context of certain ‘social’
policies such as illegal drugs policy, in addition to the target group (i.e. illegal drug users)
the end beneficiaries of the measures introduced are those who live next to or frequent
areas where such drugs are used, etc. In schematic terms, the typical constellation of the
interests at play takes the form of a relationship of indirect proportionality: the more
severe the obligations imposed on the target groups, the greater the protection afforded to
the beneficiaries. Thus, insofar as a policy is programmed and then executed, it falls to
the state and its actors to arbitrate the conflicts of interest between these two types of
actors. This kind of interpretation highlights the redistributive effects (between social
groups) of all public policies (Knoepfel, 1986).
In numerous cases, the strengthening of the position of end beneficiaries as described
above has resulted in the establishment of contractual solutions: the resolution of certain
conflicts is directly negotiated between the end beneficiaries and the target groups, even
sometimes in cases in which a formal intervention on the part of the public administration
is lacking (Weidner, 1997). In fact, the establishment of compensation mechanisms
between these two types of actors are becoming increasingly common; these mechanisms
are rooted in contracts: those located in the vicinity of nuclear power stations receive
danger money, the inconveniences caused by waste treatment installations are
compensated by lower prices being charged to the inhabitants of the relevant local
authority etc. (Kissling-Näf et al. 1998).
(4) Support of injured parties for target groups and of positively affected third parties for
end beneficiaries
As stated in Chapter 3 on policy actors, apart from the target groups where policies aim
to alter activities and the end beneficiaries, in a relational analysis it is also necessary to
take into account the actors indirectly affected by the implementation of a policy.
Without being the primary target of the state intervention, the latter may see their
(economic) situation change, either positively (i.e. positively affected third parties) or
negatively (i.e. negatively affected third parties). Depending on the evolution of their new
situation and their capacity to organise themselves, the indirectly affected actors will
become involved in policy programming and/or implementation by forming coalitions
with other social groups who support or oppose the state intervention. Obviously, the
‘natural’ coalitions form between the target groups and negatively affected third parties,
on the one hand, and the end beneficiaries and positively affected third parties on the
other. It should be stressed here that if the indirectly affected actors are very powerful
from an economic or political perspective (depending on their number, financial
resources, political support etc.), they may prove more influential than the actors directly
targeted by the policy during its conception and execution.
158
For example, if the target groups of environment policy are polluters, the end
beneficiaries are all those whose environment is affected by the many sources of
pollution in a region, while the positively affected third parties are those who develop
new less polluting technologies that are easier to market among target groups (cf. the
sustained development of eco-business – Benninghoff et al., 1997) and the negatively
affected third parties are producers who can no longer market their old polluting
technologies.
As a result of the entry into the game of groups negatively affected by the acts imposed
on target groups, the position of the latter vis à vis the administration or the end
beneficiaries may also be noticeably strengthened. This constellation arises quite
frequently in the area of fiscal measures imposed on producers or traders with the
intention that the latter should incorporate them into their prices at the expense of
consumers (opposing coalition of traders and consumers). Similar situations are also
familiar with license withdrawal. In such cases, the mobilisation of solidarity between
target groups and a more or less wide circle of negatively affected persons is quite
common. Such situations sometimes give rise to the formation of very unusual coalitions
(for example, between traders and consumers, electricity companies and ecologists during
the liberalisation of the electricity market etc.). The same thinking applies to the
coalitions that arise between the end beneficiaries of a policy and positively affected third
parties.
(5) Support of positively affected third parties and opposition of negatively affected third
parties for public actors
The support or opposition of positively and negatively affected third parties is not
restricted to social actors. In reality, the latter often make direct contact with the public
actors of the PAA and with other political-administrative actors to whom they enjoy
greater access and of whom they expect vigorous intervention among the actors of the
PAA.
Eco-business companies identified as the positively affected third parties of rigorous
environmental standardisation intensify their research activities and invite the state to
declare all of their anti-pollution products as ‘state of the art’ and their use mandatory by
all polluting companies in a certain sector etc.
(6) Opportunities and restrictions prescribed by the situation and structural variables
As already demonstrated by classical implementation analysis (Chapter 9.2), the position
of target groups may be perceptibly modified by economic events (or situation variables)
which unfold rapidly and completely beyond the control of the public or private actors
involved in the policy implementation process. Thus, in Switzerland the political
administrative authority may more readily grant planning permission in a village that has
been hit by disaster (e.g. flooding, avalanche, fire) and will allow all kinds of exceptions
to pass that would never be accepted in normal times. In contrast, events such as
technological disasters often result in the tightening of the controls on high-risk
159
operations, even if the disasters in question occur in other countries (e.g. ‘the Chernobyl
effect’ on the nuclear sector in energy policy; cf. Czada, 1991).
Despite the actors’ game and the vagaries of situation variables, the position of a target
group within a given policy process remains strongly (pre)determined by its political,
economic and social power which, in turn, is defined by the structural makeup of civil
society in general rather than a single public policy. Thus, for example, large enterprises
may be able to frustrate planning controls, particularly when they can threaten that those
controls will inhibit the provision of employment opportunities in the area (See for
example Blowers’ study of the brick industry in an English county, 1984). Conversely,
an alert raised by the refugee community about the emission of toxic gas from the only
factory in the village whose vice-director is the mayor is unlikely to mobilise negotiations
between the members of the famous ‘basic triangle’ described above, even in the case of
an obvious contravention of environmental law. While policies and the actors’ game
undoubtedly aim to resolve public problems it would be mistaken to think that they can
eliminate the causes fully and definitively. It should be noted that these situation and
structural variables affect not only the position of private actors, but also those of the
public actors in the PAA.
160
48Using the term ‘meso-implementation’, cf. Dahme et al., (1980, p. 158).
49Bardach (1977, p. 57) defines ‘implementation’ as a ‘process of assembling the elements required to produce a
particular programmatic outcome’.
50‘Implementation machine’, Bardach (1977), ‘Implementationsstruktur’ which translates as ‘implementation structure’,
i.e. all of the administrative units responsible for implementation (Mayntz 1980).
51Cf. for example the papers presented to the congress of the Association of Political Science of the Federal Republic of
Germany in October 1984 which were devoted to this debate.
52As suggested by the initial work of Sabatier and Mazmanian (1979) which bears the significant title of ‘The conditions
of effective implementation: A guide to accomplishing policy objectives’.
53 See also the exploration of this issue in Hill and Hupe (2002).
54 The terms ‘outputs’, ‘end product’, ‘administrative implementation activities’ and – depending on the circumstances –
‘administrative services’ are used synonymously in this book.
55 For Switzerland, in the sense of Article 5 of the Federal Law on Administrative Procedure of 20 December 1968 (RS
172 021). In the case of France, administrative acts are involved here.
56 Cf. for Switzerland Article 29 of the Federal Law on Administrative Procedure of 20 December 1968 (RS 172.021):
‘The parties have the right to be heard.’
57 This assumption of the 1960s as a turning point may vary from society to society.
