Enlightening policy mix complexity: a

typology of actors’ choices

Giliberto Capano, University of Bologna, Italy
giliberto.capano@unibo.it

Andrea Lippi, University of Florence, Italy

lippi@unifi.it

Abstract

Policy instruments are a fundamental component of public policies. Policy instruments

are often a result of mediation within the policy design process, whenever political
actors reshape existing instruments in various ways without introducing any real
innovation. This apparently calls for imitation, layering and ambiguity in tool choice
selection, and raises the theoretical problem of the logic according to which policy
actors choose specific policy instruments. This paper focuses precisely on such a
problem. Policy makers are influenced when selecting policy mixes, by a decisional
context which is a consequence of the dominant governance mode, the shape of the
political-institutional context, and the conflict between stakeholders’ interests and
ideas. However, regardless of the decisional context, the actors’ options are limited to
four selection strategies: hybridization, stratification, contamination and persistence.

1st ICPP Conference Grenoble 2013

Panel: Policy Design : principles and processes
Chair: Michael Howlett, S. Fraser Univ.
Session 3: Policy design: theory and concepts
Friday 28 June, h 10.45-12.45
Chair: Per-Ola Aberg
Discussant : Elisa Chelle

1
 1. Introduction

Public policy studies offer different typologies of governance modes, each of

which tries to theoretically order the ways in which public policies are coordinated and
steered (see for all: Considine and Lewis 1999, Treib, Bahr and Falkner 2007; Howlett
2011; Capano 2011). What all of these classifications of governance modes have in
common is not only their adoption of the three fundamental, albeit differently arranged,
principles of social coordination (hierarchy, market and network), but also the tendency
to associate specific policy strategies including equivalent sets of policy instruments
with each of the designed governance modes. This fact refers to basic theoretical
problem that people select policy strategies and instruments whenever adopt a
governance mode.
From this point of view, for example, it is quite normal to associate authority and
law with hierarchic modes of governance, regulation with market modes, and
information with network modes. However, we know perfectly well that:
- governance modes are conceptual categories;
- policy makers very often have ambiguous ideas when choosing which
instruments to adopt;
- very often such instruments are not in keeping with the purported governance
mode adopted at that given moment;
- policy makers decide on a contingent, instrumentally incoherent, manner.
This is because policy design and implementation is mainly constituted by policy mixes.
This contribution deals with the policy mixes when policy-makers choose
strategies and instruments in governance modes, that is, with the fact that very often,
policy-makers construe real governance modes based on policy instruments that
theoretically incorporate different paradigms, ideas and toolkits. As certain studies of
policy instruments show, the idea of any coherence between policy instruments and a
specific ideational framework or goal, has been superseded (Gunningham and Sinclair
1999; Bressers and O'Toole 2004; Howlett 2005), for the simple reason that it is
empirically fallacious and inconsistent with the facts. This is very often justified as an
outcome of the ambiguous preferences of policy-makers. We, on the contrary, assume

2
that policy mixes are an intrinsic, inescapable dimension of the policy choices that
policy-makers pursue. It does not mean that policy-makers cannot learn from their
experience, and thus modify their ideas regarding the importance of certain tools
(Bennett and Howlett 1992; Radaelli 2009). What we assume is that in any case, tools
are chosen through strategies grounded on the context within which policy-makers are
obliged to act. This ‘contingency perspective’ means that studies of policy instruments
need to pay closer attention to the context surrounding policy-makers’ choices, and to
their constant tendency to design policy strategies by adopting different mixes of policy
instrument.
This paper aims to examine this very question by proposing a framework within
which to analyze the possible strategies adopted when choosing policy instruments,
which could be useful in analyzing and explaining policy dynamics and development in
relation both to governance shifts and policy instruments choice.
The paper is organized as follows: the first section presents the general problem of
instrument’s choice; in the third section the concept of policy mix is presented and
discussed; in the fourth section our approach to incoherent pluralism in instrument
choice is presented; in the fifth section a typology of actor’ choice on policy instrument
is presented and discussed.

2. Instruments and the choice thereof

There have been a great many studies of policy instruments in the policy
literature, due to the crucial role played by their adoption. Apart from the broad range of
books and articles published by scholars specifically interested in describing and
classifying the intrinsic nature of such instruments (see Hood, 1986; Salamon, 2002),
the central theoretical issue in policy instrument studies has increasingly become the
choice made by political actors. The leading question is: why do policy-makers choose
one tool rather than another when attempting to deal with, and resolve, a specific policy
problem?
Hence the importance given in policy instruments literature to the analysis of the
reasons why political actors choose given instruments, and of those factors which
influence such choice (Salamon, 1981).
In this regard, the majority of policy instrument studies have emphasized the

3
variable and situation-driven features of such choice. According to this school of
thought, over the course of last decades a considerable number of studies have
increasingly focused on the conditions, or the situational factors, leading to the
adoption of a certain policy instrument regardless of whether it is ideologically,
economically or technically suited to resolving that problem.
For instance, in the 1950s Dahl and Lindblom (1953) submitted that
governments can choose from a variety of different available techniques when called
upon to deal with policy problems, and they subdivided them into five categories. In
the 1980s, Doern and Phidd (1988) emphasized the importance of a government’s
propensity towards two extreme choices, that of private behaviour (self - regulation) and
that of public ownership, and ranging across a spectrum characterised by exhortation,
expenditure and regulation. The authors suggested that a good match between a given
instrument and a given policy does not exist, since all instruments are theoretically
substitutable and the only valid criterion is the government’s preference for one of them.
Finally, Linder and Peters (1989) classified a sample created on the basis of eight design
criteria.
All of the aforesaid theoretical considerations led to the belief that the choice of
an instrument depends significantly on a range of factors, and is not the automatic result
of any one factor such as economic performance, technical understanding/preferences or
ideological stance.
Consequently, scholars realized that actors’ choices were typically driven by
factors grounded in the political, economic and institutional context. Peters (2002)
summarized these factors by subdividing them into five categories labeled as the ‘five I
s’: ideas, individual, institutions, interests and international environment. All of them
are perceived as capable of channeling actors’ choices. Moreover, Peters emphasized the
situational, rather than the neutral, nature of instruments, by showing that the choice of
instrument is influenced by external factors, and thus the choice of the same instrument
could be reshaped in different decision-making contexts according to the specific
spatial or time framework. Accordingly, the choice of policy instruments has been
increasingly seen to be of a contextual (Howlett and Ramesh, 1993) or contingent
(Bressers and Klok 1988) nature.
As such, a choice is made because the government in question is always aware

4
of the complete range of instruments available for the resolving of each problem, and is
frequently led to choose a coercive tool initially in order to give citizens the impression
that it is being tough (Howlett, 1991). From this perspective, the use of policy
instruments is not unique, but flexible and readily-adaptable,, since their adoption is
shaped by contingent factors (see deBruijn and Hufen, 1992). Moreover, one of the
interesting conclusions arising from the debate is that instruments are all potentially
substitutable, and their substitutability is an effect of a contingency or a layering
(Howlett and Ramesh, 1993).
Hence, the most important contribution made by policy instrument literature is
the idea that choice is of a contingent nature, and depends on the context of the
decision-making process. The scholars «argued that instruments’ choice is
circumscribed by existing social, political and economic circumstances which may
constrain or encourage the use of particular instruments» (Howlett, 1991:4). In other
words, the choice of an instrument is often induced by external factors leading to
contingent strategies whereby the actors try to adjust their preferences to the legal
framework and to current restrictions. As Kassim and Le Galès (2010) have pointed out,
the choice of instruments rarely follows any a priori agreement or any clear view of the
problems at hand.
Furthermore, policy instruments often exist independently from the decisions
leading to their creation, and are far removed from the situation in which they are to be
adopted. The choice of instruments is more frequently a matter of matching (Le Gales
and Lascoumes, 2007) different aspects, such as ideas, problems, the legal framework,
and economic resources. The actors’ choices should be seen as the result of the
influence of these factors and the consequent mix bringing together policy design and
tools (Howlett, 2011).

3. What are policy mixes?

The concept of policy mix originated in the economic literature during the 1960s,
and was largely marginalized within the field of economics until the early 1990s (Flanagan,
Uyarra and Laranja 2011).
It has since been adopted in other policy fields, such as childcare policy

5
(Gunningham and Young 1997; Stroick and Jenson 1999), innovation policy (Smith 1994),
and environmental policy (Sorrel and Sijm 2003). Furthermore, all the literature on the
“good governance” promoted by the OECD and by other international organizations, also
tends to focus on the best mix of policy instruments adopted in pursuit of specific policy
goals (OECD 2007, 2010).
While the question of policy mix has become a more general topic in public policy
research, empirical research has clearly shown that the majority of policy strategies are
composed of policy mixes (Howlett 2005; Braun 2006; Howlett and Reyner 2007). From a
policy perspective, the question of policy mixes is double-sided. On the one hand, in fact,
those scholars more interested in policy design have to understand how the best policy
mixes are to be formulated, and thus they are interested in understanding all possible
interaction among different policy instruments (Del Rio and Howlett 2013). Thus once it
has been accepted that the policy-making norm consists in the pursuit of policy goals
through specific combinations of a variety of instruments, and very often within a multi-
level institutional context, the theoretical challenge is to circumscribe the potential
combinations and effects. On the other hand, for those more interested in explaining how
and why certain policy mixes are designed, implemented and changed, the research
question is a radically different one. In this case, scholars have to identify frameworks
capable of describing, understanding and explaining the dynamics of policy mixes.
Here we are interested precisely in the second approach to instruments mixes, and in
particular we are interested in understanding how policy actors choose policy instruments
and thus continuously contributes towards creating instrument mixes.
However, the question of what constitutes a policy mix remains? Specialist studies offer
different definitions of policy mix, some of which are extremely sophisticated (Del Rio and
Howlett 2013). In this present paper we have adopted a simpler way of defining policy mix:
a policy mix is a policy strategy using a combination of policy instruments belonging to
“different” instrument categories or pertaining to different policy paradigms/beliefs
systems/ideologies.
As far as concerns the types of policy instrument adopted, scholars generally group
policy instruments according to their underlying rationale. Usually there are no more than
three or four possible general types of policy instrument, as clearly exemplified by the
taxonomy suggested by Laura Hand (2012), who proposes three large categories of policy
instrument: inducements (including both incentives and sanctions); regulations; and
capacity/knowledge tools. According to our logic, each family of policy instruments

6
contains all those instruments characterized by a similar relationship with the given policy
goals. From this point of view then, a policy mix exists when instruments belonging to
different types/families of policy instrument are assembled in order to pursue specific
policy goals.
Regarding the paradigmatic/ideological distinction, it is quite clear that policy
instruments are very often linked to a specific policy paradigm. Policy instruments are very
often ideological. This means that we expect to find coherence between the policy
instruments adopted and the prevailing/hegemonic policy values/principles within a specific
policy field. However, very often this does not happen, and we may find substantial
incongruence and incoherence between policy instruments and policy principles.
So, policy mixes are characterized by the co-existence, within the same adopted policy-
toolkit, of instruments belonging to different “technical” and “ideological” families.
This phenomenon seems to be an inherent characteristic of policy-making, and as such
it constitutes a highly challenging topic. What we find particularly interesting are the
reasons why certain mixes represent the policy-making norm, and why policy-makers
continually choose their instruments in a mixed way.

4. Policy mixes generated by contingent incoherent pluralism

The basic assumption of our reasoning is that any selection of policy instruments
is characterized by an intrinsic policy mix trend (Bressers and O’Toole 2005), and as
such should be considered a result of a miscellany of different ideas, interests and
technologies, and as institutionalized in certain specific, recurrent situations. Basically,
according to several studies on layering (Smith 1983; Mahoney and Thelen 2010; der
Heijden 2010), the choice of instruments represents a kind of institutional change in the
way policies are made, based on a process of layering which can be characterized, as we
shall see, by the different approaches adopted by diverse actors when choosing those
new instruments in addition to those already in use.
From this perspective, policy instruments are not entirely new, even if they seem to
be new, since different situations lead policy-makers to perceive them from different
points of view (Landry and Varone 2005). On the other hand, they do not last forever,
nor are they the same as they were years beforehand, as they tend to be adopted in

7
various different ways in order to meet changing contingent requirements (Bijker,
Hughes, and Pinch 1989). Something has changed, and yet this change does not affect
the instruments as such, but the context within which they are selected (Slembeck
1997).
In fact, we are accustomed to observing the introduction of new instruments in the
policy-making process, whereas in reality what we are looking at different situations
which known instruments are associated with, as the consequence of the changing
political situation. The overall effect is that of an ‘incoherent pluralism’, that is, a
recurrent situation in which the instruments are adopted by policy makers in a manner
that may not be in keeping with their original design and the technical and ideological
background of that design. The label 'incoherent pluralism' derives from consolidated
studies and practices, and is intended to evoke a seemingly contradictory situation. Both
‘incoherence’ and ‘pluralism’ derive from studies of decision-making processes, such as
Braybroke and Lindblom’s theory (1963), and Di Maggio and Powell's new
institutionalism (1983).
Focusing on policy instruments from the perspective of ‘incoherent pluralism’
means dealing with the governance problem from a different point of view which -
rather than starting from the top that is, by means of a specific form of governance that
works through a presumed set of specific, intrinsically coherent policy instruments -
addresses the analysis from the bottom up, that is, from the concrete choice of policy
instruments. Furthermore, this analytical approach may help us get a better
understanding of one of the shortcomings of existing governance studies, highlighted by
recent empirical research: namely, the fact that in the real world of policy making, pure
governance modes do not exist, while the various different types of real governance
resemble one another to a greater degree than one would imagine (Capano, Reyner,
Zito 2012; Tollefson, Zito, and Gale 2012)1.
The incoherent mix of policy tools is a consequence of the fact that policy-

1
We are not saying here that the various typologies of governance modes are not useful.
On the contrary, they represent a very useful way of ordering reality, and when considered as
paradigms they represent important points of reference both for scholars’ analyses, and as an
indication of policy-makers’ conceptual frames of reference.

8
makers’ choices are not made on the basis of a mechanic deduction from the actual
governance mode adopted within a specific policy field (although such represents a
powerful institutional factor), nor on the basis of the eventual prevailing policy
principles and beliefs, but as a response to specific contingences arising within such a
context. It would seem a rough argumentation to refer to the context as one of the most
relevant con-causal mechanism in producing incoherent pluralism and thus policy
mixes, but this claim is grounded on an empirical point of view in place of a theoretical
one: the context matters because real actors are asked to decide day-by-day in an
environment made by routines but also by unexpected events.
As a consequence, we assume here that the decisional context is the result of
three possible factors - which may also interact with one another - (Schicker 2001;
Peters 2002; Landry and Varone 2005; Hood 2007; Capano and Lippi 2010) - namely:
- the dominant governance modes (meaning the dominant policy values and ideas
together with the institutionalized policy instruments);
- the features of the given political-institutional context (the political colour of the
ruling government; party system dynamics; the economic situation)
- interaction and tension between different coalitions of political and policy actors
promoting different interests and ideas.

It is evident that the above-mentioned factors influence one another to varying

degrees. For example, the dominant governance mode may influence the workings of
the political-institutional context, while a change in the political-institutional context
may well affect the dominant policy values and thus impact on the existing governance
mode; even if policy-makers are used to deciding on a routine basis, any specific event
could convince them of the need to adopt a new policy instrument that does not pertain
to the dominant policy mode. The choice of instruments is made on the basis of the
specific “situation” in which policy-makers are called upon to act. The specific situation
is influenced by the specific configuration of structural and institutional factors
designing the contingent context within which policy-makers have to select policy
instruments.
However, the possible combinations of contextual factors, even when such are
reduced to just three, as proposed here, can give rise to numerous, unforeseeable types

9
of context. What is important here is that the three aforementioned factors should be
considered as external variables with respect to the actual decisional situation in which
policy actors are called upon to decide on policy instruments. According to other similar
frameworks developed in relation to public policy and institutional change - namely
the IAD and ACF frameworks (Ostrom 2011; Sabatier and Jenkins-Smith 1993) – these
kinds of external variable establish the confines of the decisional situation in which
actors are called upon to decide, even if they do not directly determine their choices.

5. A typology of actors’ strategies when choosing policy instruments

The key question concerns the strategy adopted when choosing policy instruments.
We suggest that the strategy lies somewhere between ‘entirely situation driven’ choices
and the mechanical, unsophisticated approach whereby instruments are the
epiphenomena of technical arguments. It is precisely in order to solve the dilemma of
the excess of possible decisional contexts, that we propose a typology of policy-makers’
selection strategies based on two specific aspects of policy-makers’ perception of policy
instruments, designed to clarify the complex adoption of such instruments. These
specific features consist of the policy makers’ perception of the legitimacy and
functionality of policy instruments.
The assumptions underlying both dimensions concern rational action in policy
making. We think that policy makers are rational beings, and as such they try to find
solid arguments in order to convince themselves that a policy instrument is suitable or
useful. Furthermore, we assume that the perceptions of policy actors are absolutely
ineludible in the public policy field, as well in political science (Peters 1986), as shown
by Wilson (1989) in his well-known typology of public policies, and by Schneider and
Ingram (1993) in their typology of policy targets.
The focus on legitimacy and functionality as fundamental aspects of actors’
instruments compared with others – such as effectiveness, efficiency, equity, partisan
support (Wallner 2008) - is based on the assumption that instruments should necessarily
have the support of the majority of policy-makers, and that this support is based also on
their congruence with prevailing social values. Both derive from the well-known

10
approach to tool selection referred to as ‘instrument substitutability’ (Howlett, 1991:4)
which in turn is the synthesis of the two previous models to be found in policy
instrument literature, namely the resources-based scheme (Hood, 1983), and the
contextual models (Linder & Peters, 1989). Functionality has been chosen because
actors’ choices are based on a “perceived” technical interpretation of the way policy
instruments work. Actors need to be convinced that a specific policy tool serves to
achieve a specific goal.
Seen from this perspective, we sincerely believe that the perception of the
legitimacy and functionality of policy instruments is an extremely powerful trigger.

5.1. Legitimacy: policy instruments should be recognized as fair and appropriate

Legitimacy is a crucial political fact affecting policy makers (Beetham 1991;

Suchman 1995). It is strictly related to the political context within which any decision is
made. In the words of Salamon (2002:24): «tool choice can affect the overall sense of
legitimacy that government enjoys in the eyes of citizens». It means that policy makers
are influenced by decisional conditions, which in turn arise within a policy or
institutional setting. Really, legitimacy is a contextual dimension to understand the
actor’s strategies.
In practice, policy makers can hardly select a policy tool merely on the basis of
their own preferences, technical background or personal values. More precisely, policy
makers cannot consider such instruments simply as ‘means’, but are also obliged to take
account of ‘symbols’, ‘opinions’, ‘coalitions’, and ‘trust’. Thus we can assume that the
tool choice implies a degree of mediation between existing cognitive conditions
rendering the instruments’ selection acceptable to the public, and their own preferences
based upon technical arguments or ideological assumptions. The policy instrument has
to be accepted by the public, and has to be perceived as coherent, that is, in keeping
with the policy problem and also acceptable with respect to the values of that particular
policy sector (Schon and Rein 1994; Bakvis and Skogstadt 2002; March and Olsen
2006).
Such reasoning is in line with the Weberian conception of legitimacy as a
precondition for the acceptance of power (and of the instruments of power, as policy

11
instruments intrinsically are). Accordingly, policy instruments have to be justified for
the same reason that political and policy actors require legitimization. Bemelmans Videc
et al. (1998:8) offered a clear synthesis of this concept when they stated that the
legitimacy of a policy instrument is particularly crucial for democratic life, because, it
«may refer to the degree to which government choices are perceived as “just” and
“lawful” in the eyes of the involved actors».
A policy instrument can be legitimized in different ways. We focus here on two
categories of legitimacy: internal legitimacy and external legitimacy (Lippi, 2012).
Internal legitimacy is endogenous in relation to the policy setting, because the
policy makers provide values, arguments and rhetoric deriving from that specific policy
field. For example, internal legitimacy is provided by aspects rooted in the practice, the
legal framework or the moral background of a policy sector (so, from this point of view,
internal legitimacy is given by the institutionalized set of policy instruments and by the
policy legacy). For this reason, such legitimacy is already taken for granted by those
(policy-takers as well as policy-makers) within a given policy arena. For instance,
subsidies are a consolidated tool in social policies. A tool choice of subsidies for social
policies is internally legitimized by the insiders. Nobody would be surprised if a policy
maker were to choose subsidies as the best way of governing a given social policy. The
insiders will deem this choice as being efficient, and fitting to the problem, even if such
choice may not be the best solution from a rational point of view in that particular
situation. These “insiders” can consequently validate the instrument as acceptable
because they perceive it as familiar, coherent or potentially capable of helping improve
policy performance.
The internal legitimacy principles implies that the ‘insiders’ are the fundamental
source of legitimation of the adoption of new instruments (i.e., the introduction of
tickets or tariffs for social services) not pertaining to the legacy of the policy field in
question.
On the other hand, external legitimacy is exogenous in relation to the policy
setting. In this case, the legitimation of instruments originates from a different policy
sector/political context. The social environment where the tool originated (e.g. the
market, the EU or the field of social voluntarism) has now become fashionable and
appealing, and invested with positive meaning. The tool is validated as outstanding,

12
regardless of any real transferability to, or subsequent success within, the new context.
External legitimation means that a policy tool is perceived as fitting sectoral policy
needs due to the high reputation its source has. Thus the instruments are transferred
from their original technical, cultural and moral background, and are perceived from a
new perspective within the new policy sector. For instance, the logic of the market is
historically extraneous to public administrations, but over the course of more than
twenty-five years the majority of European governments have adopted a range of tools
with which to govern public sector policy, by borrowing these tools from financial
policy or development policies explicitly directed towards the empowerment of markets
and firms. Tools such as contracts or quasi-market competition were originally
extraneous to all public sectors, but are now perceived as suitable and fitting, even if
they are not directly or immediately workable.

5.2. Functionality: instruments are a function of available resources and capacities

Functionality derives from the resource based scheme approach to the

instruments. It is crucial, since it concerns the instrumentality of a tool, that is its
coherence and effectiveness respect to the goal to be reached. The technical nature of an
instrument is presumed to influence the choice thereof (Hood and Margetts, 2007:3)
although it is mediated by the policy-makers perception, which is obviously based on
their competences and knowledge. From this point of view the functionality of the
policy instruments is never fixed or objective but always subjected, and thus,
constructed by the policy-makers’ perceptions. Functionality leads to the available
resources (namely capability, knowledge, technological and financial arrangements, use
of authority) that are able to influence the actor’s strategy as a legacy. Functionality is
the pre-determined setting where the instrument will be adopted. According to Hood,
the most part of instruments’ selection is a learning experience through trial and errors
looking at tha available functionalities.
Hence, we assume that the policy-makers’, as well of the public opinion,
perception of an instrument’s way of working (e.g. the ‘instrumentality’ according to

13
Salamon 2002) is a necessary prerequisite for tool choice.
The underlying assumption here is that every policy instrument is assumed, by
actors, to have a specific way of working and a specific field of application. This
creates a specific contingent situation, because decision makers are constrained by that
shape (which they have also designed with their perceptions) and that use when
choosing their tools. The functional pattern is precisely the framework of functions
included in the shape and field of interaction implicitly specified by an instrument.
Furthermore, it should be pointed out that as Le Galès and Lascoumes (2007)
have shown, the technical nature of an instrument is considered as a prerequisite for
action. From this perspective, policy instrument has to be primarily analyzed as an
artifact, rather than just a set of meanings, values or expected behaviors. This means
that a given tool is a 'concentrate of power' in its own 'material shape'. This happens
because an artifact is the effect of different resources, such the above mentioned ones.
Hence the material shape of a tool is not neutral (Van Nispel and Ringeling 1998), but
represents a dense structure of power relations among a certain number of actors, and
depicts the functional pattern of that power. This power structure is intrinsically
embedded in the material shape of the tool, and will influence the policy arena and its
stakeholders, whenever they call for, or are subject to, said tool (Peters 2002; Linder and
Peters 1989). So the perception of the functional pattern of a policy instrument is
double-sided, that is, it is both technical and political. This means that actors build their
beliefs regarding the ways policy instruments work, by paying attention not only to the
instruments’ capacity to solve policy problems, but also to their potential impact on the
distribution of power with the policy field
So we assume that the workings of policy instruments follow two dimensions:
the differentiated functionality and the integrated one.
The differentiated functionality dimension refers to the power of an instrument to
specialize a policy arena, being perceived as a unique tool of governance. It makes the
use of the policy instrument totally different from that of others. Differentiation means
individualization and distinction. A differentiated instrument is perceived by the policy
makers as being original, that is, a sample, a pattern, a best approach to be subsequently
imitated. The differentiated pattern may be the result of different legal or cognitive
factors. As regards the cognitive sphere, this stands for the creation of symbols and

14
codes or languages that set the instrument apart, so that everyone can talk about it by
using a specific set of meanings. As regards the legal factors, this refers to specific
procedures or to a characteristic regulation affecting the instrument, so that everyone
can recognize it within its legal framework. The differentiation category is very
selective and constraining. It strongly influences all those policy makers adopting it, and
requires the reshaping of relations between the various components of the policy tool
set. For instance, fiscal tools are technically shaped in a unique way, so that they are
reputed to be suitable for taxation purposes only. They are strongly differentiated, and
can also influence the behavior of those employing them. If you decide to adopt a fiscal
instrument, you have to take the cognitive and legal implications of its adoption into
account. The shape of the taxation instrument (including procedures, organizational
offices, regulation etc.) may also influence the decision makers adopting that
instrument. As a consequence, the differentiation category is conservative and selective.
It is oriented towards the persistence and institutionalization of an instrument
that becomes so taken-for-granted that it is no longer disputable. It strongly influences
all those policy makers adopting it, but excludes all other policy makers: a clear borders
divides insiders and outsiders.
The integrated functionality refers to the capacity of a policy instrument to
include and cover an increasing number of people, policy problems and situations, in
order that it may be generally considered as fitting due to its ability to encompass a
broad range of problems both in the same policy field and in different ones. For
example, in many European countries the partnership between the public and private
spheres can be seen in a broad range of policy sectors beyond that of public utilities (its
original sector of adoption). The public-private partnership (as well as project financing)
is now perceived as an umbrella term covering a lot of uses, policy problems and
strategies, including town-planning, social and environmental policies. It has become a
generic tool, and its own regulations and technicalities are broadly-defined. Thus, the
public-private partnership can be employed in a flexible way, and is capable of
achieving a consensus among a wide range of people. It does not strictly compel
anyone, but creates a weak framework within which a great many people can converge
(Hodge & Greve, 2007).
For this reason, this kind of functional pattern is prone to integration, it covers

15
several different problems and solutions without the need for any specific, exclusive
rules. This integration capacity that the instrument in question plays an ‘umbrella role’
and emerges as a fashionable solution to several problems. Integration stands for
depersonalization, meaning that a given instrument ‘loses its original creator’ (person
or agency) and its potentially different institutional logic, in order to become a
generalized tool.
While the differentiated functionality represents a constraint, that is, the tool in
question can only be used for a specific purpose and in a specific way, the integrated
pattern implies that the instrument in question will cover a lot of situations, people and
policy problems, and provide a generic, comprehensive, consensual approach to the
adjustment of goals, means and vested interests.

5.3. A typology of strategies affecting the choice of policy instruments

We would argue that tool choice is strongly based on a limited number of selection
strategies which policy-makers can adopt in contingent decisional context influenced by
the interaction of the factors listed above in paragraph 4 which create the conditions for
a situation of incoherent pluralism (policy mix). These strategies are presented in Figure
1, which provides a typology based on the abovementioned dychotomization of
legitimacy and functionality.

Fig. 1. Four types of strategies affecting tool choice

Legitimacy
Internal External
Differentiated Persistence Hybridization
Functionality
Integrated Contamination Stratification

16
 As Fig.1 shows, both the legitimacy and the functionality point to four policy
mix types of strategies: (i) persistence, (ii) contamination, (iii) hybridization and (iv)
stratification.

(i) Persistence is the strategy whereby policy makers are strongly legitimized by a
policy legacy, and by the past in general, to consolidate former choices. The shape
of the instrument encourages policy makers to continue with its unconditional
adoption, since its material profile is perceived so clearly and specifically that
nobody could possibly criticize its suitability. Obviously, this case fails to tell us
whether the instrument is really, and constantly, effective or highly promising. Tool
choice is strongly legitimized by path dependence and by considerable
specialization. Tool choice seems to be congruent (and consequently efficient and
effective), and there is no new policy mix, but simply the maintenance of the set of
tools deemed to be effective also in the light of the new contingency. This strategy
involves a conservative perception of tool choice: people are convinced that the
current mix is performing well so there is no need to change it by adding new tools.

(ii) Contamination2 is the specific strategy whereby policy makers decide, according to
their prevailing policy values and to the insiders within the policy field, to adopt
new tools in an unspecific way. These tools may not be well suited to the given
situation, but they are reputed to take account of the political situation to a greater
degree, through the creation of a broader consensus. Contamination is the tool
choice strategy whereby insiders try to defend their monopoly position (which
means the prevailing governance mode) within the policy field by showing that
they are capable of changing their instruments’ preferences in order to deal with
perceived policy ineffectiveness. However, this strategy implies a change in the
policy sector, because in one way or another people (policy-makers first and

2
Of course, we use the term ‘contamination’ not in the medical sense but in that of Greek and
Latin manuscript tradition. In this sense, contaminatio is a technique of writing whereby a single
manuscript contains readings originating from different sources or different traditions. The final result
may be highly innovative rather than a simple collage.

17
 foremost) are forced to adjust their preferences to the requirements of the new
constraining tools. This adjustment produces a new blend of policy tools which can
be the basis for incremental policy change. The tool choice is not necessarily
congruous, but it may be endorsed as being innovative, and thus efficient and
effective.

(iii) Hybridization3 is that strategy whereby policy makers are pressurized, by the need
for external legitimation, into innovating within a given policy sector, through the
adoption of a new tool that is highly constrictive, since it is well formed and highly
specialized. This implies the inclusion of policy principles belonging policy
paradigm/governance modes different to the existing one, possibly also of a
conflicting nature, within the existing set of policy tools, which forces insiders to
reframe that set. This tool choice entails a loss of congruity (due to the inclusion of
a highly constraining tool belonging to a different policy field, or political context,
or policy paradigm), although it also enables a wider range of people and situations
to be included, in the name of renewed efficiency or effectiveness. Ultimately,
hybridization means the establishment of a new policy mix in which different
institutional logics (governance principles) and ways of working are combined in
the pursuit of a new balance. Thus hybridization can potentially constitute the
premise for significant policy change.

(iv) Finally, stratification is a strategy whereby policy makers introduce instruments in

a generic way, such that they are readily accepted in other fields due to the need for
external legitimation. The degree of congruity in this case is extremely low, and it
does not necessarily entail any policy change since the policy-makers perceive the
new instrument to be suitable (from a political point of view) for the policy in
question, and its integrated nature means there is no pressure to render it congruous
with the existing set of policy tools. Policy makers simply have to adopt it, and
thus juxtapose it with previously-existing instruments, without having to search for

3
We use the term ‘hybridization’ in the biological sense, that is, meaning the interbreeding of
individuals from genetically distinct populations. This concept is also used in organizational theory
(Minkoff 2002; Hargrave and Van del Ven 2006) and in public administration studies (Christensen and
Laegreid 2001). Here, generally speaking, hybridization means the combination of different institutional
logics.

18
 any apparent congruity or integration. The choice of instrument is actually
legitimized by the appealing nature of the new instrument, and by the fact that it
encompasses a number of different situations, people and solutions, thus achieving
a broad consensus, although its generic nature means that it is not really of a
constraining character. The logic of stratification implies that the new instrument is
added, but that there are no real relations with existing instruments, and thus this
choice does not necessarily produce any real change.

What emerges this typology is that although context driven, the choice of policy
mixes is path dependent. It works all by confirming the existing instrument toolkit or by
layering to it new instruments. What is interesting then is thanks to this typology it is
possible to distinguishing different types of instrument layering. Contamination means
to add a new policy instrument in a conscious way which policy actors try to integrate
in the existing toolkit also by recalibrating it. It is a kind of layering by learning.
Stratification, on the opposite, is a choice through which policy actors simply add a new
instrument, because an external pressure, but without any attention to integrate it in the
existing toolkit. It is a kind of mechanic layering. Hybridization is a very different type
of layering because the added instrument is bearer of a different policy paradigm of
governance principle and this creates a mix which is qualitative different respect to the
pre-existing and could be something of substantial new. It is a kind of innovative
layering.
We assume that the four types of instrument choice strategies described above are
those by which policy-makers choose the policy instruments to be adopted according to
the specific contingences they are asked to deal with. However, this approach to policy
instruments selection emerges under the shadow of efficiency and effectiveness:
although the choice of policy instruments is very often unrelated to any pursuit of
efficiency and effectiveness (since the latter tend to disappear in the face of the
legitimacy issue and of contextualized functionality), these dimensions are ever-present
in the selection process seen from a cognitive perspective.
In practice, the effects of those policy instruments adopted in the past are perceived
as negative, and this leads to further debate aimed at changing such instruments. Such
debate, concerning legitimacy and functionality, tends to focus on the problem of

19
improving policy effectiveness and efficiency. These terms are clearly used in a
rhetorical manner, although they remain the unavoidable point of reference of tool
selection, adoption and implementation. While it may be true that policy instruments are
not neutral tools, but are indeed infused with values, are the subject of political
wrangling, and are capable of leading a life of their own, they nevertheless exist under
the constant shadow of efficiency and effectiveness requirements, which policy-makers
can always challenge in order to dispute the fairness and utility of existing policy
instruments.

6. Conclusions

In this paper we have tried to venture into the complex world of policy mixes, by
assuming their necessity, and thus the need for more sensitive conceptual tools with
which to analyze them. Our main assumption is that the world of policy instrument
choice is one of incoherent pluralism, that is, a world in which policy-actors can choose
from among different instruments, but do not search for instruments that are in keeping
with their chosen goals; despite being strongly committed to making choices
constrained by demands for policy efficiency and effectiveness, what they actually
choose is the more politically-useful tool. Their choice, then, should be contextualized
and interpreted as being contingency-driven. Starting from this theoretical assumption,
we have proposed a specific typology of strategies affecting the choice of policy tools
which could be helpful in ordering and analyzing the chaotic world of policy mixes.
The results seem interesting from a theoretical point of view, even though they
need to be empirically tested. From an empirical point of view, policy-making processes
need to be reconstructed in order to gain a better understanding of the dynamics of
policy-actors. First of all to understand if policy-makers, in the formulation phase really
decide as we proposed, secondly to thoroughly examine the real effects of
implementation of the new policy mixes, in order to identify any possible feed-back
effect on the policy-makers and on the actual interpretation by the implementers of the
new policy-mixes chosen (this could be very important in terms of understanding of
how the internal incoherence of policy mixes can be used by implementers to reinvent

20
the policy mixes themselves).
Finally, from the theoretical point of view our typological framework requires us
to start thinking about the gap between the literature on governance modes and that
regarding policy instruments. In fact, if the reality of policy instrumentation is
characterized by incoherent pluralism, policy mixes and contingencies, then it is very
difficult to understand how this reality can be related to the coherence between
governance modes and the corresponding policy tools, as claimed by governance
studies.

References

Bakvis, Herman, and Grace Skogstad. 2002. “Canadian Federalism: Performance,

Effectiveness, and Legitimacy.” In Canadian Federalism: Performance,
Effectiveness, and Legitimacy. ed. Herman Bakvis and Grace Skogstad.
Oxford, England: Oxford University Press, 2–23.
Beetham, David. 1991. The Legitimation of Power. London: Macmillan.
Bemelmans-Videc, Marie Louise, Rist Ray C., and Evert Vedung. 1998. Carrots,
Sticks and Sermons. Policy Instruments and Their Evaluation. New
Brunswick: Transaction.
Bennett, Colin, and Michael Howlett. 1992. “The Lessons of Learning: Reconciling
Theories of Policy Learning and Policy Change.” Policy Sciences 25(3),
275–294.
Bijker, Wiebe, Hughes Thomas, and Pinch, Trevor (eds.). 1989. The social
construction of Technological system, Cambridge, Mass., MIT Press.
Braybroke, David, and Charles E. Lindblom. 1963. Policy Evaluation ad a Social
Process. Glemcoe: Free Press.
Bressers, Hans, and Pieter-Jan Klok. 1988. ‘Fundamentals for a Theory of
PolicyInstruments.’ International Journal of Social Economics 15 (3/4): 22-
41.
Bressers, Hans. T.A., and Laurence J. O’Toole. 2005. “Instrument Selection and
Implementation in a Networked Context.” In Designing Government: From
Instruments to Governance, eds. P. Eliades, M. M. Hill & M. Howlett.

21
 Montreal & Kingston: McGill-Queen’s University Press, 132-153.
Braun, Dietmar. 2006. “The Mix of Policy Rationales in Science and Technology
Policy”. Melbourne Journal of Politics, 31: 8-35
de Bruijn, Johans A., and Hans A.M. Hufen. 1998. ‘The Traditional Approach to
PolicyInstruments.’ In Public Policy Instruments : Evaluating the Tools of
PublicAdministration, eds., B. G. Peters and F. K. M. V. Nispen, 11-32.
New York:Edward Elgar.
Capano, Giliberto, and Andrea Lippi. 2010. “Gli strumenti di governo stanno
cambiando? Aspetti teorici e problemi empirici.” Rivista Italiana di
Politiche Pubbliche 11 (2): 5–30.
Capano, Giliberto. 2011a. “Government Continues to Do Its Job. A Comparative
Study of Governance Shifts in the Higher Education Sector.” Public
Administration 89 (4): 1622–164.
Capano, Giliberto, Jeremy Reyner, and Anthony Zito. 2012. “Governance from the
Bottom up: Complexity and Divergence in Comparative Perspective.”
Public Administration 90 (1): 56–73.
Christensen, Tom, and Per Laegreid. 2001. New Public Management: The
Transformation of Ideas and Practice. Aldershot: Ashgate.
Considine, Mark, and Jenny M. Lewis. 1999. “Governance at Ground Level: The
Frontline Bureaucrat in the Age of Markets and Networks”. Public
Administration Review 59 (6): 467–480.
Considine, Mark, and Jenny M. Lewis. 2003. “Bureaucracy, Network or Enterprise?
Comparing Models of Governance in Australia, Britain, the Netherlands,
and New Zealand.” Public Administration Review 63(2): 131–140.
Del Rio, Pablo and Michael. 2013. Beyond the “Tinbergen Rule” in Policy Design:
Matching Tools and Goals in Policy Portfolios”. Working Paper no.
LKYSPP13-01. Lee Kuan Yew School of Public Policy WP Series.
der Heijden, Jeroen. 2010. “Institutional Layering: A Review of the Use of the
Concept”. Politics 31(1): 9-18.
Dahl, Robert, and Charles, Lindblom. 1953. Politics, economics, and welfare:
planning and politico-economic systems resolved into basic social
processes. New York: Harper.
DiMaggio, Paul J,. and Powell Walter. 1983. “The Iron Cage Revisited: Institutional
Isomorphism and Collective Rationality in Organizational Fields.” American

22
 Sociological Review 48 (2): 147–60.
Doern, G. Bruce, and Richard W. Phidd. 1988. Canadian Public Policy: Ideas,
Structure,Process . Toronto, ON: Nelson.
Flanagan Kieron, Elvira Uyarra, Manuel Laranja. 2011 "Reconceptualising the
‘policy mix’ for innovation." Research Policy 40(5) : 702-713
Gunningham, Neil, and Mike D. Young. 1997. ‘Toward Optimal Environmental
Policy:The Case of Biodiversity Conservation.’ Ecology Law Quarterly 24:
243-298.
Gunningham, Neil, and Darren Sinclair. 1999. “Regulatory Pluralism: Designing
Policy Mixes for Environmental Protection.” Law and Policy 21 (1): 49–76.
Hand Laura 2012. “Public Policy Design and Assumptions About Human Behavior”.
Paper presented the 2012 Western Political Science Association’s Annual
Conference.
Hargrave, J. Timothy, and Andrew Van de Ven. 2006. “A Collective Action Model of
Institutional Innovation.” Academy of Management Review, 31 (4): 864–88.
Hodge, G.A. e C. Greve (2007), “Public-Private Partnerships: an International
Performance Review”. Public Administration Review, 67, 3, pp. 545-558.
Hood, Christopher, 1986. The Tools of Government . Chatham: Chatham House
Publishers.
Hood, Christopher. 2007. “Intellectual Obsolescence and Intellectual Makeovers:
Reflections on the Tools of Government After Two Decades.” Governance
20 (1): 127–144.
Hood, Christopher, and Helen Z. Margetts. 2007. The Tools of Government in the
Digital Age. London: Palgrave.
Howlett, Michael. 1991. ‘Policy Instruments, Policy Styles, and Policy
Implementation: National Approaches to Theories of Instrument Choice.’
Policy Studies Journal 19 (2): 1-21.

Howlett, Michael. 2005. “What is a Policy Instrument? Policy Tools, Policy Mixes
and Policy-Implementation Styles.” In Designing Government. From
Instruments to Governance, eds. Pearl Eliadis, Margaret M. Hill and
Michael Howlett. Montreal & Kingston: McGill-Queen’s University Press,
311-50.
Howlett, Michael. 2011. Designing Public Policies: Principles and Instruments.

23
 London: Routledge.
Howlett, Michael and M. Ramesh, 1993. “Patterns of Policy Instrument Choice: Policy
Styles, Policy Learning and the Privatization Experience”, Policy Studies Review,
12(1-2) pp. 1-22.
Howlett, Michael; and Jeremy Rayner, J. 2007. “Design Principles for Policy Mixes:
Cohesion and Coherence in New Governance Arrangements”. Policy and
Society, 26 (4), pp. 1-18.
Kassim Husseim, and Patrick Le Gales. 2010. “Exploring Governance in a Multi-Level
Polity: A Policy Instruments Approach”. West European Politics 33 (1): 1–21.
Landry, Réjean, and Frédéric Varone. 2005. “The Choice of Policy Instruments”. In
Designing Government. From Instruments to Governance, eds. Pearl Eliadis,
Margaret M. Hill and Michael Howlett. Montreal & Kingston: McGill-Queen’s
University Press, 106-131.
Le Galès, Patrick, and Pierre Lascoumes. 2007. “Introduction: Understanding Public
Policy through Its Instruments. From the Nature of Instruments to the
Sociology of Public Policy Instrumentation.” Governance 20 (1): 1–21.
Linder, Stephen H., and Guy B. Peters. 1989. “Instruments of Government:
Perceptions and Contexts.” Journal of Public Policy 9 (1): 1–33.

Lippi, A. 2012. La politica degli intellettuali. Catania, Bonanno.

Mahoney, James and Kathleen Thelen. 2010. ‘A Theory of Gradual Institutional

Change’ in J. Mahoney and K. Thelen (eds.), Explaining Institutional Change:
Ambiguity, Agency, and Power, New York: Cambridge University Press, pp. 1–
37.
March, James G., and Johan P. Olsen. 2006. “The Logic of Appropriateness.” In
Handbook of Public Policy, eds. Martin Rein, Michael Moran, and Robert E.
Goodin. Oxford: Oxford University Press, 689–708.
Minkoff, Debra C. 2002. “The Emergence of Hybrid Organizational Forms:
Combining Identity–Based Service Provision and Political Action.” Non-
profit and Voluntary Sector Quarterly 31 (3): 377–401.
Oecd. 2007. Instruments Mixes for Environmental Policy. Paris: Oecd.

24
 Oecd. 2010. OECD Science, Technology and Industry Outlook 2010. Paris:Oecd.

Ostrom, Elinor. 2011 “Background on the Institutional Analysis and Development

Framework”, Policy Studies Journal, 39(1): 7-27.
Peters, B. Guy. 1986. American Public Policy: Promise and Performance. 2nd ed.,
Chatham, NJ: Chatham House Publishers, Inc.
Peters, B. Guy. 2002. “The Politics of Instruments.” In The Tools of Government. A
Guide to the New Governance, ed. Lester M. Salamon. New York: Oxford
University Press, ???–???.
Radaelli, Claudio. 2009. “Measuring Policy Learning: Regulatory Impact Assessment
in Europe.” Journal of European Public Policy 16 (8): 1145–1164.
Sabatier, Paul, and Hank C. Jenkins-Smith. 1993. Policy Change and Learning. An
Advocacy Coalition Approach. Boulder: Westview Press.
Salamon Lester M. 1981. “Rethinking Public Management: Third Party Government
and the ChangeForms of Government Action”. Public Policy 29(3), pp. 255-275.
Salamon, Lester M. 2002. The Tools of Government. A Guide to the New Governance.
New York: Oxford University Press.
Schickler, Eric. 2001. Disjointed Pluralism: Institutional Innovation and the
Development of the US Congress. Princeton, NJ: Princeton University Press.
Schneider Anne and Helen Ingram. 1993. “Social Construction of Target Populations:
Implications for Policy e Politics”. American Political Science Review 87(2)
pp.334-347.
Schön, Donald A., and Martin Rein. 1994. Frame Reflection: Toward the Resolution of
Intractable Policy Controversies. New York: Basic Books.
Slembeck, Tilman. 1997. “The Formation of Economic Policy: A Cognitive-
Evolutionary Approach to Policy-Making.” Constitutional Political Economy 8
(3): 225–254.
Smith, BruceL.R. 1983) ‘Changing Public–Private Sector Relations: A Look at the
United States’. Annals of the American Academy of Political and Social Science
466(1), pp. 149–164.
Smith, Keith. 1994. New directions in research and technology policy: identifying the key
issues. STEP Report R-01, STEP Group, Oslo.
http://www.nifu.no/files/2012/11/STEPrapport1 1994.pdf.

Sorrel, Steven, and Jos Sijm. 2003. “Carbon trading in the policy mix”. Oxford Review
of Economic Policy 19 (3), 420–437.

25
Stroick, Shaorn and Jane Jenson. 1999. What is the Best Policy Mix for Canada’s
Young Children? Study no. F-09. Canadian Policy Research Network, Ottawa.
Suchman, Mark C. 1995. “Managing legitimacy: Strategic and institutional
approaches”. Academy of Management Review. 20(3), 571-610.
Tollefson, Chris, Anthony R. Zito, and Fred Gale. 2012. “Symposium Overview:
Conceptualizing New Governance Arrangements.” Public Administration 90 (1):
3–18.
Treib, Oliver, Holger, Bahr, and Gerda Falkner. 2007. “Modes of Governance: Towards
a Conceptual Clarification.” Journal of European Public Policy 14 (1): 1–20.
van Nispen, Frans K.M. and Arthur B. Ringeling. 1998. “On Instruments and
Instrumentality: A Critical Assessment.” In Public Policy Instruments. Evaluating
the Tools of Public Administration, eds. B. Guy Peters and Frans K.M. van
Nispen. Cheltenham: Edward Elgar, 204–217.
Wallner, Jennifer. 2008. “Legitimacy and Public Policy: Seeing BeyondEffectiveness,
Efficiency, and Performance”. Policy Studies Journal 36(3): 421-44.

Wilson, James Q. 1974. Political Organizations. Princeton: Princeton University Press.

26