DANIELE CARAMANI (ED.

COMPARATIVE POLITICS

Oxford University Press

POLICY-MAKING

YANNIS KARAGIANNIS (*) AND CLAUDIO M. RADAELLI (**)

(*)
Marie Curie Fellow, Centre for Regulatory Governance, University of Exeter, UK
(**)
Professor of Political Science and Jean Monnet Chair; Director, Centre for Regulatory
Governance, University of Exeter, UK

MARCH 2007

Acknowledgements: we are grateful to Daniele Caramani for his guidance

and comments throughout this project. Gloria Regonini provided
suggestions on how to map out the broad area of public policy analysis by
commenting on an early draft. Bruce Doern, Burkhard Eberlein and Nikos
Zahariadis kindly sent us their comments. The usual disclaimer applies.

1
Policy-Making

Yannis Karagiannis

Claudio Radaelli

Introduction

Research traditions

Definitions and dimensions of policy

Substantive and theoretical issues in policy-making and policy change

Policy-making and the new institutionalisms

How to compare public policies

Conclusions

2
Reader’s guide

In this chapter, we introduce the reader to the comparative analysis of public policies. We
start from a short historical account of the origins of political-scientific endeavors to
explain the policy-making process. Older research traditions were non-comparative and
often blended empirical analysis with normative judgments. As the field developed, it
became more analytical, more “positive” (as opposed to “normative”) and more
comparative. We then look at the format of public policies: some policies are made
through observable actions, while others seem to exist despite the lack of specific policy-
making activities. Similarly, the range of actors involved in public policy-making
processes varies across time, space, and type of issues. We make the point that the
definition of public policy is a component of the analytical exercise, as policies are
conceptual constructs. Different research traditions explore policies with different
ontological and epistemological aims. We therefore introduce a continuum on which
some of the most important traditions can be situated. Next, we examine in more detail
different analytical approaches. Some scholars start with a known policy problem, such
as the comparative inefficiency of a policy. Others start with from a purely theoretical
issue, such as the role of expertise in the policy process, or the apparent longevity of
certain policies in the face of substantial environmental changes. To understand and
explain policies, one needs theoretical framework. Thus, we introduce what is currently
the dominant (and probably still the fastest-growing) approach to policy-making, new
institutionalism. What difference do institutions make in terms of the policy-making
process and the outputs of the political system? We show how different institutional
perspective address this question and conclude with an overview of where current
research traditions may be going in the foreseeable future.

3
1. Introduction

Public policy analysis (hereafter ‘PPA’) deals with the identification, description,
explanation, and assessment of public policies. This ambitious focus makes PPA
one of the broadest sub-disciplines in political science. Not only does it
incorporate elements from many other sub-fields (e.g. political parties, electoral
studies, implementation studies, international relations, etc), it also covers a large
number of distinguishable activities (e.g. agenda-setting, decision-making,
implementation, evaluation), and it does so in a population of policies that is
almost incomputable (e.g. policing the streets of ancient Rome, raising taxes in
18th century Pennsylvania, or regulating financial markets in modern day Tokyo).

The encompassing nature of PPA is both an advantage (since it allows us to

make sense of various different realities) and a disadvantage (since it obliges us
to find a commonly-shared language). This makes comparative PPA a quite
complex area of study. Should one be discouraged by such complexity? Our
view coincides with that of a famous economist: “complexity can and often does
serve as an inducement rather than as a deterrent.” (Williamson, 1985: 15). Get
yourself submerged in complexity, try to find a reasonable way to simplify,
understand, and explain it, and the rewards will be great.

Policy-making goes beyond the examination of the policy outputs produced by a

political system; rather, it is a framework through which several aspects of politics
can be analysed, including (a) how problems get on the political agenda; (b) how
interest groups, the bureaucracy, and the other branches of government interact;
(c) how collective choice, institutions, and democratic politics work … or do not
work.

Two important findings of comparative PPA stand out and should be highlighted
from the outset. Firstly, although across-country comparison is still a valuable

4
strategy, the classification of public policies according to criteria other than
“nationality” is more intellectually rewarding. These criteria refer to the kind of
strategic interaction that characterises the institutions shaping the policy-making
process, the role played by networks in that process, the identification of the
actors with technocratic or political modes of decision-making, the patterns of
implementation, the balance between ideational and material resources, and so
on. This shifts comparison from inter-national to inter-type: one does not any
longer compare “American” to “Indian”, but “type X” to “type Y”. Secondly, the
study of public policy has ushered in a new approach to the understanding of
fundamental aspects of democratic politics. Policy design affects citizens’
participation (Ingram and Schneider 1993). Policy analysts have shown how the
link between citizens’ preferences, manifestos, elections, government formation,
and policy choice does not follow the classic template of democratic theory –
thus re-focusing the latter on how public policies strengthen or weaken
democracy (Ingram and Rathgeb-Smith 1993).

2. Research traditions

Contemporary PPA took off in the USA after the end of World War II. At the
beginning, it was conceived as a separate discipline. The plan of the founding
fathers was to make public policy draw from neighbouring disciplines (political
science, economics, law, sociology, social psychology, etc), but encompass
them, not form part of them. In addition, this new discipline and comparative
politics were foes, rather than friends. Indeed, the initial efforts in policy analysis
were motivated by the desire to solve specific problems (and to evaluate specific
federal programmes) in the USA. At least in terms of international comparisons,
there was nothing “comparative” in this endeavour. Note also that early policy
analysts selected their theories and their models on the basis of the specific
problems that needed to be described and explained, not on the basis of

5
philosophical or ontological beliefs and traditions. Without a problem, they
thought, there is no policy inquiry.

Regarding Europe, PPA arrived in the late 1960s, and established itself
progressively through the 1970s and 1980s. The diffusion of policy analysis from
the USA to Europe was not a simple matter of exporting knowledge and
methods. Instead, each European country forged its own peculiar approach to
the policy sciences. For example, the UK tradition has been marked by a higher
dose of realism, as well as by the legacy of the pioneering work of J. Richardson
and G. Jordan on policy-making styles and pressure groups (Richardson 1982,
Richardson and Jordan 1979). In Germany, the path taken by pioneers such as
F. Scharpf (1973) was more influenced by a combination of traditional German
institutionalism and the strong influence of American political science (which, in
turn, had often been developed by German-speaking expatriates). In France, the
focus was still different, due to the innovative work of B. Jobert on the role of
ideational factors called “frames of reference” (Faure et al. 1995; Jobert and
Muller 1987), as well as to the influence of older sociological traditions. The
specific direction taken by PPA in each country may have been influenced as
much by academic traditions and innovation as by the constitutional, institutional
and political structures of each country. Whatever the national variations, the
common theme of PPA has always been its fairly inter-disciplinary character.

PPA has considerably evolved since it was first developed in the post-WWII
USA. It has acquired an increasingly comparative dimension, and rather than
subsuming other disciplines, it has been subsumed under political science, while
building bridges towards all the neighbouring disciplines.

3. Definitions and dimensions of public policy

- What is a public policy?

6
The classic joke about academics applies (Q: “How many academics are needed
to fix a light-bulb?” – A: “What do you mean by a light-bulb?”). What is “public
policy”? And, how can it be analysed comparatively? Is “public policy” just
anything that is decided by the government, is it less than that, or is it more than
that? And, in attempting to compare public policies, does one not run the risk of
comparing the proverbial apples and oranges? Just like the question “What is
medicine?” leads to a description of the main problems and activities of medical
doctors, so the question “What is a public policy?” leads to a description of the
main problems and activities of public policy analysts.

Of course, most of us already have an intuitive understanding of the meaning of

“public policy”. Yet, we need a precise and explicit definition. According to one
conventional definition, public policy is “what governments do, why they do it, and
what difference it makes” (Dye 1976:1). This definition draws our attention
towards governmental action (e.g. taxation). Despite presenting the obvious
advantages of common sense and parsimony, however, this definition may be
too narrow. There are policies in which at least one of the major actors is an
international organisation (e.g. European Union competition policy, McGowan
and Wilks 1995). Similarly, there are policies where independent regulatory
agencies are in charge of complex tasks that effectively amount to public policy-
making (e.g. U.S. environmental policy). Thus, unless the term ‘government’ is
stretched to its limits, such public policies would unduly fall outside the scope of
our definition. And indeed, one important comparative public policy question
concerns precisely this: when is a policy decided by a government, when is it
decided by parliament, and when is it left to experts and/or independent
regulators? (e.g. Epstein and O´Halloran 1999).

- Does public policy-making need to be made?

7
Another point to consider is that policy can be effectively ‘made’ by …not doing
anything at all. For example, thorny issues such as migration and racism had
long been kept off the public agenda for many years (Bachrach and Baratz
1963). Or, doing nothing when a country invades another can be an eloquent
way to signal one’s foreign policy. Public policy, then, is as much ‘decision’ as it
is ‘non-decision’. It is precisely in situations like this that the comparative method
to PPA can help. Although it may appear paradoxical to start collecting empirical
data on a non-event, one can still compare a country where a decision was taken
with a country where it was not taken – and thus try to identify the crucial
difference between the two. Or, one can compare a situation within a country
where specific action was taken for a minor problem, while a much grander
problem was left unaddressed.

Further, politicians often want to be seen as doing something, even if they know
that there are no sufficient technological, financial, or political resources to
implement policies. This is the well-known case of symbolic policies, in which
there is no implementation. Yet, symbols play an important role in politics and
public policy is no exception. Symbolic policies can therefore either remain
permanently unimplemented, or they can be implemented by the private sector,
or again be implemented at a later stage. The lack of implementation does not
prevent the emergence of other political effects (such as swings in the popularity
of politicians, impacts on national identity, or influences on first- and second-
order beliefs). Comparative PPA can contribute by comparing cases where the
effects of symbolic politics differ.

- The time and timing of public policy-making

Another point to keep in mind here is that, more often than not, a policy is not an
individual decision. The obvious example is the present number of policy-relevant
actors (“stakeholders”). Yet, like in so many other human activities, time and

8
timing too are fundamental parameters in public policy-making. 1 Public policies
have a past, a present (which involves a whole series of activities), and a future.
For example, the preferences and the expectations of political actors depend on
time horizons, both past and future. The past helps actors learn – learn about the
content of policies, but also about the content of the politics that support (or
surround) such policies. Similarly, the future may be bright and long, dark and
short, or completely colour-less and uncertain. Again, that will significantly affect
the way policy-makers perceive their role, their allies, and the world on which
they act.

But time (timing) matters in the “present”, too. The sequence of policy-making
activities may have a very important effect on the content of policies. This can
best be explained by reference to the baseline model of policy-making, the “cycle
model”. Policy-making consists of a series of stages (see diagram 1 below): (1)
problem definition: politicians, interest groups, bureaucracies, and the media
identify a problem, and develop action-oriented views; (2) agenda-setting: the
relevant actors decide who shall take the lead on the emerging policy-making
process; (3) deliberation and policy formulation: the actors interact by mobilizing
their respective resources (knowledge, power, alliances, etc) and thus arrive at a
preliminary formulation of policy options; (4) decision-making: the “constitutional”
decision-maker makes a choice in favour of one option; (5) implementation: that
option is passed on to the relevant implementing authority, which thus gives the
policy its final shape; (6) assessment, and re-statement or termination: finally, the
policy is evaluated, and the policy is either terminated (a rather rare event
indeed) or renewed. Three points need to be made here:

(a) Like a tree that hides a forest, each stage can hide numerous sub-stages –
perhaps ironically, it can even create its own policy cycle. Crucially, there is
nothing automatic about these stages: nothing is essentially and purely a-

1
To be more precise, time (or timing) may be defined parametrically, in which case individual actors act in
exogenously defined tempos. But it may also become a strategic parameter, i.e. an object of purposeful
manipulation in view of securing one’s most preferred public policy outcomes.

9
 political. Hence, different public policies may exhibit different characteristics
over one or more stages.
(b) In the real world, those stages may not be as neatly ordered as in the
baseline model. The “present” may take different forms, depending on
whether stages appear sequentially, or whether they overlap. Some actors
may prefer clear sequences, while others may better be served by
“confusion”. And again, different policies may differ in that respect, too.
(c) The stages may be feeding back into one-another in a rather complex way.
For example, decision-making may feed back into problem definition,
especially when a stake-holder group estimates that its preferences were
unduly disregarded by decision-makers. In general, policy analysts may gain
significant insights by investigating instances of non-normal sequences and
by attempting to explain those by conducting comparative analysis.

Diagram 1: The baseline model of policy-making cycles

Problem Agenda-
Definition Setting Deliberation

Decisión-
Making

Evaluation Implementation

Termination

10
- Will a new government always make a difference?

This question leads to the effects of one of that fundamental characteristics of

modern democracies, elections. Can elections change the course of public
policy? Is one election enough? Can we safely say that all policy commitments in
a party manifesto will become public policy if only that party comes to power? Or,
is it safer to acknowledge that newly elected governments are bound by their
inheritance (e.g. past agreements), and by the preferences of other actors (e.g.
veto players)?

Often enough, the new incumbent will try to attract the attention of the media to
some high-profile changes. Yet most of the activities of the executive will be
concentrated on carrying out policies of the past (Rose 1990). Similarly, the
winners of an election may be constrained by the constitutional roles of other
powerful actors (Tsebelis 2002). The crucial point here is that public policies may
be more or less easy to change, independently of the (declared) intentions of a
new government.

When elections do have an impact, public policy is said to be more responsive. It

is more democratic. On the other hand, such volatile policies run the risk of being
assimilated to political footballs, which are kicked in the left and the right side of
the pitch when new politicians are elected. But, when does that happen, and
when does it not? Are all policies affected similarly by a change in government,
and if not, why? And, is kicking the ball, or the knowledge that the ball may be
kicked, good or bad? The “stubbornness” of certain policies is particularly
interesting for political science, and comparative institutional research illuminates
its causes. It follows that it might be wrong, at least for some purposes, to
approach public policies from the conceptual angle of an individual decision, or
with a short-sighted lens (Pierson 2004). Such “snap-shot” approaches are also
inadequate if we think that what goes on between time t 1 and t2 is as important as

11
the events at t1 and t2. For those interested in policy as process this is yet
another reason to go beyond individual decisions. Overall, it is better to consider
a whole course of action and, when relevant, non-decisions.

- Back to light-bulbs: How should we define a public policy?

A more satisfying definition of public policy should therefore take into account the
following factors: (a) public policy can be made and implemented either by public
or by private actors, or both; yet, some special role and responsibilities must be
reserved to public actors; (b) public policies may be made and implemented by
one, a few, or many different actors; yet, some centrally located actors must be
seen as more decisive that other (more or less causally relevant) actors; (c)
public policy must be distinguished from law; yet law may serve to support,
sustain and implement public policies; and (d) public policy is a series of actions
whose causes and effects span a more or less identifiable period of time; yet
such political actions are not by themselves public policy. In our view, it is also
important to note that public policy is not defined only at a particular moment of
time; rather, it arises out of governance processes, i.e. systematic attempts
carried out by a multiplicity of actors to tackle collective problems over a certain
period of time. Indeed, now that the whole of political science has somewhat
switched from a ‘government’ to a ‘governance’ agenda (Rhodes 1997), it would
be surprising if PPA were to remain the last bastion of the ‘government’
approach.

Coming to our definition of public policy (box 1), it is a course of deliberate public
action or inaction, which revolves around a collective problem. Note that this
definition does not mean to suggest that public policy actually ‘solves’ collective
problems – indeed, there is a whole literature on policy fiascos (Bovens and
T’Hart 1996, Moran 2003).

12
 BOX 1
LOOKING FOR PUBLIC POLICY
 We follow what is now mainstream PPA and define public policy as a course of
action or inaction revolving around a collective problem.
 But this is not as simple as it looks like. In fact, as Hugh Heclo shows: (1972: 84)
policy does not seem to be a “self-defining term”. It is an analytic category, the
contents of which are identified by the analyst rather than by the policy-makers
or pieces of legislation or administration (Parsons 1996).
 Muller and Surel (1998) explain the public policy is a process of social mediation
in which what is at stake is the social legitimacy of governance. In their views,
governments do not engage in public policy to solve or deal with problems, but
to produce social legitimacy. This introduces a normative dimension that is
difficult to handle with the classic tools of positive political science.
 Recent scholarship looks at policy analysis in different guises, including puzzle-
solving, critical listening, advice, instrument for democracy, and social critique
(Moran et al. 2006; chapters 5-9).
 Students facing this complexity may end up agreeing with Cunningham (1963,
quoted in several public policy textbooks) that ‘policy is rather like the elephant.
You recognize it when you see it but cannot easily define it’.
 The truth is that the identification of what is public policy is already a task in the
territory of policy analysis

To conclude with the dimensions of analysis, there is an important distinction to

be made between policy scholars interested in content (that is, knowledge about
the substance of policies), and policy scholars interested in form (that is, the
analytical structure of public choice). Those in the first group typically acquire a
very detailed knowledge of specific public policies, and their skills combine those
of historians, lawyers and applied economists (usually in that order). Those in the
second group typically acquire more abstract knowledge about the workings of
the political system in general, and their skills combine those of theory-oriented
economists and constitutional lawyers. Ideally, of course, scholars should be
aware of both literatures. In practice, however, one does observe this distinction,
which therefore leads to the question, Do the policy sciences look at what

13
collective decisions are, or do they make recommendations about how such
decisions should be made?

We can plot the various answers along a continuum (drawing on Regonini 2001,
see also Parsons 1996). At one pole we have studies explicitly informed by
values and ethical considerations of what is good democratic governance and
good governance. Prescriptive studies draw on managerial and engineering
notions of what policy analysis should be. Analysis is oriented towards advancing
recommendations and best practice and explaining how things should change
and for what aims. Evidence-based accounts acknowledge the scientific merit of
several methods of inquiry, including case studies, analytic narratives (Bates et
al. 1998), and small-n comparisons. General theories are looked with some
scepticism, unless they are historically and geographically situated – for
example, by clarifying that social-democratic parties may be able to control
inflation better than conservative parties only in some countries and only in some
periods. At the other pole of the continuum, positive policy analysis is based on
causal relationships drawn from new political economy and supported by
statistical analysis.

Fig. 1 – Types of policy analysis

prescriptive:: best descriptive: positive: causal

Normative-values practice , evidence based explanations,
management accounts forecasts

4. Substantive and theoretical issues in policy-making and policy change.

14
As already mentioned above, the lively debates on various aspects of policy-
making are continuously shaped and re-shaped by wider debates in political
science. But, this is a two-way street: theoretical policy analyses and empirical
results have also contributed to the development of major political science
programmes (such as the rational choice, historical, and sociological variants of
the neo-institutionalism). It follows that the answers given by policy analysts to
the questions raised here have varied, and that variance reflects changing
paradigms in political science, as well as feed-back loops.

So, what do we “know” today? One must start with the first wave of studies that
defined the theoretical work in the policy sciences during the 1970s and early
1980s. Interestingly, most of these landmark studies were comparative, in the
inter-national sense of the term. Indeed, they were the first successful attempts
to get out of the one-country box of earlier studies and produce knowledge by
using the tools of comparative politics. The impact of classic political science
debates on the comparative method, in the wake of Lijphart´s and Sartori’s
seminal contributions was evident.

- Comparative PPA starting from a problem: research in social welfare policies

One successful combination of comparative methodology and policy sciences

arose out of the research programme centred on the welfare state. Why do some
welfare systems seem to be more efficient than others? And why are some more
developed than others? What is the problem with the inefficient systems? What
impedes the development of the least developed ones? As opposed to more
theory-driven research (see the next paragraph), the motivation for research in
this branch of comparative PPA was empirical. The comparative analysis of the
welfare state exposed the presence of different families of countries. Typologies
of the welfare state then evolved into explanatory models, based on what we now
call historical institutionalism.

15
Today, policy scholars working on the welfare state and similar issues are
engaged in major methodological discussions on macro-comparative historical
research and the role of time in politics and policy dynamics (e.g. Mahoney &
Rueschemeyer 2003). In terms of findings, the upshot of this kind of comparative
PPA research is that, since macro-historical sequences deeply affect the
evolution of policies, we should not expect to observe automatic policy
convergence. In other words, national policies still differ, because the national
histories of these policies still differ.

- Comparative PPA starting from a model

A second research programme arose out of a more theoretical interest in the

stages of the policy process (see diagram 1). The thrust of the stage model is to
examine public policy dynamically, by distinguishing between stages (see
diagram 1). Of course, real-world policies cannot be cut salami-slice (Sabatier
2007): there are feedback loops between one stage and the others - policies do
not progress ‘naturally’ from one stage to another. In addition, the arrows that
connect different stages do not seem to convey any specific meaning of
causality. In other words, it is not clear why one stage should be the cause of
another. These criticisms notwithstanding, this model has potential for the
analysis of new, emerging policies, such as some EU policies.

Not all problems make it to the political agenda. Indeed, given that all policy
actors have limited resources, it seems that problems do not compete only with
their own solution, but also with other problems. Thus, the stages model may
usefully lead us to develop a specific research question, such as, Which
problems will be processed by the political system? Which make it to the
shortlist, and why? (Notice the essentially comparative nature of the question.)

Similarly, the stages model has led some researchers to focus on the definition of
a problem. A ‘policy problem’ may reflect a pressing and general (or even

16
universal) public need, such as the current need to control industrial emissions.
But, it may also reflect only a political opportunity for public intervention, as
shown by the long history of non-natural public monopolies, or more recently by
the emergence of Chinese capitalism, where politicians have been very active
political entrepreneurs. Or, again, it can be a demand emanating from a small
segment of society, as in the case of demands for tourist information policies
emanating from local shop-owners, restaurateurs, etc. The media, to take
another example, may voice their concerns about dangerous errant dogs, and
politicians may wish to respond to this ‘demand’ (Hood and Lodge 2006). Finally,
policy problems may be generated by the civil servants themselves, in a
bureaucratic-demand fashion: an agency, for example, may want to expand its
jurisdiction and therefore formulate a problem that justifies this expansion
(Milward 1980; for the social construction of problems see Rochefort and Cobb
1994).

Studies of agenda setting show that there are several actors involved in this
process, but also that those actors with most agenda-setting power can use (or
abuse) that power to gain significant leverage in subsequent stages (Shepsle
1979). Similarly, actors with gate-keeping powers may be able to bias the
number or content of issues that occupy the policy-making system. The agenda-
setting and gate-keeping rules may differ from one policy to another, or from one
moment to another. Comparative PPA contributes to our fuller understanding of
the importance of such rules, not only in terms of policy content, but also in terms
of democratic governance. It follows that different policies within the same
country may exhibit different degrees of democracy.

- Looking at substance: Expertise, networks, and interest groups

If one just accepts to look beyond politicians, it immediately appears that the
world of public policy-making is populated by various other actors. The role of
such actors may be either explicitly planned by politicians (McCubbins &

17
Schwartz 1984; McCubbins, Noll & Weingast 1987; Huber & Shipan 2002 for a
different view), or it may it may be imposed by resourceful non-elected officials
(Carpenter 2001). Inside government, bureaucrats and politicians impose rules
on other bureaucrats, thus contributing to the phenomenon of ‘regulation inside
government’ (James 2000). Communities of experts with shared beliefs about
how problems and policies work have successfully pressed for specific agendas,
often at the international level (Haas 1992). Pressure groups in countries such as
Denmark and the Netherlands are involved in hearings, horizon scanning
exercises, and consultation bodies that effectively enable them to set the agenda
with the departments in a consensual fashion. In other countries they use other
pathways to make their voice heard in the policy formulation stage and in
government-industry relations (Richardson 2000, Wilks and Wright 1987). Non-
governmental organisations have been active in setting the agenda for anti-
smoking policies, health care provision, corporate social responsibility,
sustainable development and to a limited extent trade in national and
international contexts. Finally, there is evidence that lawyers, insurance
companies, auditors and management consultants are perceived by firms as
“regulators”, at least in some sectors such as the food industry (Hutter 2006).

- Windows of opportunity and punctuated equilibria

In Kingdon’s model (1995), actors, problems and solutions are creatively

combined by policy entrepreneurs exploiting so-called “windows of opportunities”.
Kingdon incorporates the fundamental uncertainty of policy processes in a model
of agenda-setting that explains how entrepreneurs combine “solutions looking for
problems”, “problems looking for solutions”, and “political attention looking for
popularity and election deadlines”. His model provides a sophisticated approach
to the multi-dimensionality of time. There are three types of time: the time of
politics (based around elections), the time of solutions (which emerge gradually,
first among experts and bureaucrats, and then in the top civil service and
ministerial advisors), and the time of problems (exposed by media via the well-

18
known attention cycle originally identified by Anthony Downs). Public policy
theorists have thus discussed questions such as the emergence of policy
windows, the arrival of policy entrepreneurs, the difference between business
and policy entrepreneurs, and the balance of ambiguity and predictability in
agenda setting (Zahariadis 1999, Natali 2005).

Combining image (the social perception of policy problems) and venue (the
choice of institutional venues), Baumgartner & Jones (1993) have launched their
own original approach to policy agendas, focussing on the explanation of both
stability over long periods of time, and change. Recently, Baumgartner, Green-
Pedersen & Jones (2006) have revisited studies of agenda-setting with an
explicit comparative orientation, showing how their ideas of punctuated
equilibrium and other models can be applied across nations (see John 2006 for
an assessment of the technical components of the policy agenda project
originated by Baumgartner & Jones).

- A closer look at the world between agendas and decisions

Some comparative PPA experts elect to look more carefully at what happens
once the agenda has been set, i.e. at the moment when political system “takes”
decisions. As a result, we now have a detailed “map” of what happens at this
stage. Comparative questions can use that map to examine substantive policy
issues (e.g. where the analyst has reasons to believe that decision-making
practices differ in terms of their quality), or to refine, systematize, and theorize on
this crucial stage of the policy process.

Decision-making usually involves pre-legislative scrutiny of proposals using

specific tools, such as regulatory impact assessment (“RIA” has recently become
very popular across the OECD and beyond). Proposals for legislation are then
processed by parliaments following procedures that differ from country to
country. However, policy decisions do not necessarily imply a law-making,

19
parliamentary mode of governance. Federal executive agencies in the USA and
independent regulators in other countries receive delegated powers from their
parliaments. They can take decisions and implement directly, subject to various
forms of executive control on the quality of the evidence produced to support
regulatory decisions (control is often based on regulatory impact assessment).
There are also forms of decision-making that are based on soft law, voluntary
agreement, codes of conduct, self-regulation and other new modes of
governance. Therefore, in domains such as environmental, energy, housing,
financial supervision, money laundering there is often a regime based on a mix of
laws, codes of conduct, international accords, and standards designed and
enforced by technical bodies.

- “Poly-lithic” views of policy-making, rationality, and learning

Actors often operate in policy networks. Instead of holding on to a monolithic

view of “the” decision-maker, policy scholars prefer to study different networks
characterised by resource interdependence (Carpenter 2001; Richardson 1982,
Rhodes 1997, Marin and Mayntz 1991, for criticisms see Dowding 2001).
Drawing on network and learning theories, Sabatier´s “advocacy coalition
framework” (ACF – see Sabatier 2007; chapter 7) explains policy change in
terms of the competition between a small number of advocacy coalitions.
Advocacy coalitions include a multitude of actors: politicians, bureaucrats, policy
experts, interest groups, and the media. What keeps these actors together is a
set of relatively stable beliefs and a sustained effort at coordination. Policies are
normally dominated by one advocacy coalition, but this is challenged by one or
more alternative advocacy coalitions. The balance of power between the
incumbent advocacy coalition and the other(s) can change if there is an external
shock, such as an election, or policy learning across coalitions (usually facilitated
by appropriate fora of discussion). The model accounts for a number of important
factors in the policy process, including ideational politics, learning, and the
exogenous elements that disrupt the balance within policy areas.

20
Turning to the resources at play in decision-making, a good deal of scholarly
work has focused on the role of rationality. Charles Lindblom (1965) and Hugh
Heclo (1974) have shown that policy actors do not take decisions on the basis of
a full information set. Rather, they operate in a context of generalised, sometimes
radical, uncertainty, limited rationality and learning. The idea that politics should
be analysed on the basis of imperfect rationality and learning was an innovation
that went beyond the limits of the policy sciences, and paved the way for a
reconsideration of the whole substance of politics. Since then, studies on policy
learning have proliferated (May 1992, Bennett and Howlett 1992). Authors such
as Lindblom have made explicit the linkages between limited rationality, policy
analysis, and democratic theory, and have thus presented a model of the policy
process that is also an approach to understanding the intelligence of democracy.
Democracy is similar to the market in that there is a plurality of actors with limited
information and high resource interdependence. Instead of establishing optimal
solutions on the basis of calculation, actors use social interaction to mutually
adjust and discover equilibrium points (Lindblom 1965; 1975).

One limitation of this approach is the difficulty of finding studies that control for
both learning and lack of learning. The resulting empirical work can be as vague
and imprecise as the theoretical concept of learning. Another limitation is that it is
not clear whether we can measure policy learning on individuals (policy-makers
may say they have learned to justify choices that were dictated to them) and/or
on institutions (since organizations do not have cognitive capabilities of their
own). Last but not least, learning is often un-necessarily coupled with an
optimistic view of History. Yet, politicians may learn from their experience and
then produce disasters. There is as much policy enlightenment as there is
“endarkement”. Nowadays, the learning research programme has met the major
debate in political science on the role of ideas, interests, and institutions.
Ideational analysis of various guises (Hay 2002, Jacobsen 1995, Majone 1989,

21
Radaelli 1995) is where we find today most of the learning-based approaches to
PPA.

- Implementation: will the original decision carry the day?

After decisions, the next stage of the policy process is implementation. This is the
stage where, arguably, bureaucrats and interest groups are more important than
politicians. Policy-making does not come to an end once a policy is set out or
approved. The traditional demarcation between policy and administration has
obfuscated the issue of implementation. The traditional view is that the
administrator’s job is to carry out policy formulated by decision-makers. This may
well be consistent with democratic theory (politicians decide and the public
administration executes), but it is also quite unrealistic in practice.

To assume perfect implementation is highly un-realistic. Negotiations and

adaptation prevail over the goals set by the legislator. Redistribution, distribution,
and regulation are not perfect contracts. Rules are enforced in real-world
settings, where inspectors and firms interact and specify the meanings of what
are often abstract and vague norms. Courts intervene with their jurisprudence
and steer policy in one direction or another. Goals have to be clearly defined and
understood, resources made available, and the system able to communicate
effectively to control those individuals and organizations involved in the
performance of tasks.

- Evaluation, and termination or renewal

As discussed earlier, the last stage of the policy-making process is evaluation.

Has the policy succeeded? What has it achieved? Evaluations and appraisals
take different forms, including statistical studies, ex-post impact appraisals, value
for money studies, surveys of populations targeted by public programmes, and
performance indicators. Evaluation is supposed to provide learning points for the

22
decision-makers. A feedback loop can then link evaluation and re-formulation of
policies. The reality is more complicated. Evaluation studies are often
commissioned to show to the media that something is being done, but their
conclusions are not taken into account. “Governments – Weiss explained - do
not usually have institutionalised channels and procedures to connect evaluation
findings to the arenas in which decisions are being reached” (Weiss, 1999: 479)
Politicians have several sources of information, including evidence provided
directly by pressure groups that fund their campaigns. Why should they be
interested in social scientific evaluations instead of listening to their core
constituencies for support? In addition, it is often difficult to establish un-
ambiguously benchmarks for success. And the theories of knowledge implied by
evaluation studies can be extremely complex, as they have to specify under what
type of knowledge one can ascertain that there is success or failure (Shadish,
Cook, Leviton 1991). When there are multiple decision-makers operating under
conditions of uncertainty, a model based on the assumption that evaluation is
simply information that feeds back into decisions and policy re-formulation is
inadequate. Recent approaches to appraisals have thus made endogenous the
political and institutional context in which evaluation takes place (Owens, Rayner
and Bina 2004).

The stage model can finally be used to clarify concepts such as policy
convergence. One common argument is that globalisation has reduced the policy
autonomy of governments and produced convergence across countries. The
stage model has shown that convergence can occur at the level of having a
common discourse (“convergence of talk”, following Brunnson 1989), or can be
deeper and involve similar decisions across countries. Very rarely have political
scientists find convergence at the level of implementation and final policy
outcomes and impacts (Knill 2005 for a review). This is yet another reason why
policy studies have not detected uniformity and convergence as a result of
globalisation – it would be odd to find that shared ideas lead to the adoption of
the same decisions, and that these decisions are implemented similarly by

23
different political parties operating in specific institutional settings (on how the
institutional dimension refracts and limits processes of convergence see Swank
and Steinmo 2002; Radaelli 2005).

- The classification of policies

Up to this point, we may have given the impression that comparative PPA focuses more
on inter-national or inter-temporal comparisons. Although we have hinted to the
possibility of inter-policy comparisons, we here focus explicitly on such research. That
inter-policy research programme is largely due to the pivotal typologizing work Theodore
Lowi, according to whom “policies determine politics” (1972: 299, see mainly Lowi
1964). Lowi´s innovation was to look at public policies as arenas of power, rather than as
mere outputs of the political system. His principal idea was that policy objectives affect
politics by conditioning actors’ focus and behaviour. In other words, the characteristics of
public policies format the logic of political action. But then, how do policies differ?
According to Lowi, policies can be, and actually are, meaningfully distinguished
according to the likelihood and applicability of public coercion. That criterion yields four
types: re-distributive, distributive, regulatory, and constitutive policies. (box 2)

24
 Box 2: Lowi’s four types of policies
1. Redistributive Policies (e.g. taxation, welfare, labour market)
- GOAL: Re-allocate wealth
- MECHANISM: Coercion
- CHARACTERISTIC: Obvious relation between costs and benefits
- POLITICS: Conflictual, polarized, ideological
2. Distributive Policies (e.g. education, research, tax reductions)
- GOAL: Fund socially valuable activities
- MECHANISM: Incentives
- CHARACTERISTIC: Collective public provision
- POLITICS: Consensual
3. Regulatory Policies (e.g. competition, food safety, environment)
- GOAL: Correct perceived (potential) market failures
- MECHANISM: Evidence, argument, and coercion
- CHARACTERISTIC: Legal norms
- POLITICS: Changing coalitions, as costs and benefits are (re-)defined
4. Constitutive Policies (e.g. constitutional policy, meta-policies)
- GOAL: Make rules about rules
- MECHANISM: Co-ordination, governance
- CHARACTERISTIC: Institutions and norms
- POLITICS: Varying (according to the degree of identification of a problem)

It is vital to stress that these are ideal-types. Hence, one would expect to find in a
given real-world policy (say, agriculture) elements of regulation and re-
distribution of resources. Similarly, although Lowi started by expressing his
dissatisfaction with the two classical dichotomies of regulation v. non-regulation,
and coercion v. non-coercion, it is probably beyond doubt that regulation has
become increasingly important in modern politics. Indeed, given that
governments and institutional organisations have increasingly less faith in
Keynesian policies, and prefer to intervene by creating rules rather than by
spending money or by using taxes to change income distribution across social
classes, regulation has reappeared at the forefront of policy-making choices.

25
5. Policy-making and the new institutionalisms

Since the 1980s, new research programmes have emerged. The most important
common characteristic of these new research programmes is the emphasis they
all put on institutions. Yet, “institutions” is a rather vague concept. It can cover
rules (as in “the bureau has an obligation to consult with all interested parties and
to take their views into consideration”), or organizations (as in “the institutions
involved include the administrative agency and interest groups X and Y”), or
both. Within each category, too, there is a certain degree of ambiguity.
Institutional rules may refer only to formal rules (as in “A has the power to
overrule the decision of B”), or only to informal ones (as in “all actors agree that
M should not intervene in the day-to-day business of A”), or to both. Finally,
institutions can be identified as explanatory variables (as in “administrative
procedures regulate the access of interest groups to the agency”), or as
dependent variables (as in “legislators´ lack of interest in the matter has forced
the Prime Minister to create the National Agency for the Protection of Public
Policy Analysts”), or both. Hence the new research programmes are united by
their common emphasis on institutions, and disunited by the different definitions,
treatments, and emphases they put on them. Some research programmes are
purely “institutionalist”, while others retain insights from different traditions.

The first well-established institutional approach is associated with the

organization-sociological institutional agenda of J. March and J. Olsen, in which
institutions are defined as both organizations and (formal and informal) rules.
Although March and Olsen´s initial contribution to decision-theory was non-
institutionalist, they subsequently broadened their model to include institutions
(March and Olsen 1989). Their initial model, which was termed “garbage can”,
described situations in which actors, problems, decision-settings, and solutions
were randomly combined. The only source of stability in the erratic flows of
decisions was time (a decision-making opportunity). In contrast, March and
Olsen´s new institutionalist model construes actors as provided by (and with) a

26
logic of behaviour; problems are edited and filtered by institutions; and decision-
making rules provide stability. Hence, the new model can provide an explanation
of political inertia by endogeneising institutions.

Other institutional explanations of policy-making borrow variably from economics

and sociology to build “historical institutionalist” accounts. This work is usually
associated with authors such as I. Katznelson, J. Mahoney, P. Pierson, K.
Thelen, and S. Steinmo. One of the core ideas of historical institutionalist
accounts is that the evolution of public policies is “path dependent”. Path
dependence refers to a situation where an initially contingent event places the
policy on historical tracks which become increasingly difficult to change. The
theory becomes interesting where the track chosen randomly at the time of
origination was not the most efficient one, but then produces reinforcing effects
(“increasing returns”), which make it increasingly efficient as time passes
(Pierson 2004). The most famous example in economics is the adoption of thee
QWERTY keyboard for personal computers. Analogies have been drawn in
public policies, especially in the areas of trade and social welfare policies (see
references in Katznelson 1998, and Thelen 1999).

Neo-institutionalist analysis has made us aware that politics is sticky and inertial.
In turn, this means that we need to explain stability, and that we have to avoid
functionalist ideas according to which changes in the political system and/or
public policies are mere reflections of “deeper” economic and/or social changes.
On the other hand, real-world policies have changed dramatically in more than
one occasion. For example, taxation was once used to steer markets and
redistribute income; it is now a market-conforming policy (Steinmo 2003).
Similarly, environmental policy has moved from end-of-pipe solutions to the logic
of incentives. Monetary policy and competition policy have moved from being
instruments in the hands of politicians to being delegated to central banks and
independent competition authorities respectively. Thus, the explanation of both

27
stability and change has re-defined the key research questions for a large
number of policy analysts.

Other authors have looked at the constitutional level of policy-making and

institutions by drawing on economics. This programme is usually referred to as
the rational choice approach. Nevertheless, it is important to acknowledge that it
should more accurately be described as a set of approaches, sharing key
assumptions (methodological individualism, utility maximisation under complete
or incomplete information, and exogenous preferences) and the idea of using
theory to advance logically complete and internally consistent hypotheses that
may then be empirically tested.

On these bases, some rational choice approaches to public policy treat

institutions mainly as the independent variable, while some others treat them
mainly as the dependent variable. Among the former, this is the case of studies
of the delegation of policy-making powers from the legislature to the
bureaucracy, where scholars attempt to identify the institutional determinants of
delegation (e.g. Epstein & O´Halloran 1999, Franchino 2007), or the effects of
the institutions emerging from delegation on the evolution of the corresponding
policies (e.g. Moe 1985, Pollack 2003). Such studies are usually heavily
influenced by the economic theory of transaction costs, as developed by Oliver
Williamson (1996). Other rational choice approaches treat institutions mainly as
the dependent variable. This is the case of studies that focus on the problems
created by the existence of different preferences and asymmetric information
between the legislature and the executive, and thus on the incentives offered by
the former to the latter. Such studies borrow heavily from the economic model of
principal-agent (Laffont & Martimort 2002, Ross 1973). It remains true, however,
that this approach has been developed mainly by economists, who appear to be
more willing to make the strong necessary assumptions for pure principal-agent
analysis. These assumptions are: (a) all actors are rational utility-maximizers who
enter contractual relationships that are supported by a costless judicial system;

28
(b) the executive agent holds valuable information to which the legislative
principal cannot gain direct access; (c) the legislative principal offers a take-it-or-
leave-it contract to the executive agent, so that the latter is just a passive actor
who does not negotiate with the principal; and (d) the passive executive agent
engages in voluntary trade with the legislative principal: the agent has a credible
outside option that the legislative principal has to consider and “eliminate” by
offering more attractive options. Overall, the transaction costs is less restrictive in
terms of assumptions, for it explicitly allows for bounded rationality and
negotiations, imperfect and costly judicial systems, and ex post haggling costs
between the legislature and the executive.

Looking at the future, we can expect the current research programmes to remain
popular and even more integrated with mainstream debates in political sciences.
There are also options for radically new programmes. One option is to go back to
Harold Lasswell and his idea of democratic policy sciences, and combine
deliberative democracy with PPA (Fischer 1992, 2003; Haajer and Wagennaar
2003). This option puts policy analysis squarely on the normative side, in some
cases with post-empirical and post-modern tones (Dobuzinkis 1992). Literary
analysis, drama, anthropology and “continental philosophy” are the disciplines
that inspire some of the authors working in this direction (Amin and Palan 2001).
Other scholars are looking at evolutionary theory for inspiration about where to
go next (John 2003, Steinmo 2003) In turn, there is both a game-theoretical
variant and an historical institutionalist take on evolution and public policies.

6. How to compare public policies

Let us now turn to the crucial issue of why, and how to compare public policies.
Many of these arguments made in the previous chapter on this Volume apply
equally well to public policy. Students who wish to adopt the comparative method
in PPA are usually tempted by the idea of comparing the same policy in two or

29
more countries. As noted above, this idea has been very influential, and
important studies have compared the social welfare, trade, environmental,
monetary and budgetary policies of various Western European and/or North
American countries. This approach can lead to very good results when it is very
carefully thought through and very precise, but also to failure when it is based on
too general a definition of the research question, unjustified selection on the
dependent variable, and/or negative degrees of freedom.

It may be useful to start with two examples of problematic comparisons. First,

consider a student who wishes to analyse British monetary policy, and decides to
add France and adopt the comparative method. However, “monetary policy” is
too vague a topic (does the student intend to describe and explain agenda-
setting, decision-making, implementation, or effects, of monetary policy?). Britain
and France may differ on too many uncontrolled variables (such as the level of
their international commitments, the degree of independence of their central
banks, the structure and political influence of their financial sectors, the traditional
role of parliament in public policy-making, etc). Unless the student is able to
explain why France represented the most interesting case for comparison, and
how this comparison allowed him to resolve the problem of negative degrees of
freedom, the project will most likely fail.

Second, consider another student willing to work on environmental policy. Unlike

the first student, this student is aware of the problem of negative degrees of
freedom, as well as of importance of framing the research question in a precise
way. She may therefore choose to provisionally entitle her research “Emission
standards applied to industry in six technologically and economically developing
countries.” Although her chances of success may be greater than those of her
colleague, this student will still have to justify her choice of the six developing
countries. Unless she can show either (a) that these countries represent cases
with considerable variance of her key independent variables, or (b) that this
choice respect the “possibility principle” (Mahoney and Goertz 2004), or (c) that

30
the cases that she compares are not the six countries themselves, but numerous
instances of emission standards-setting within each country, her research may
only succeed by chance.

These examples serve to emphasize the following two points. First, when
comparing public policies, analysts should always compare something more
specific than just “policies”. This is because, unlike other areas of political
science, public policy is too general and lacks a natural focus. Second, when
selecting cases for the purposes of comparing aspects of public policies, analysts
should always consider that the default condition should be comparison between
the values of the same independent variables, and that the second-best option in
small n research is selection on the dependent variable provided that the
possibility principle is respected.

For these reasons, comparative strategies other than those based on

international comparisons are often preferred. Most scholars will either engage in
the comparison of (i) two or more policies, (ii) that face a similar issue, (iii) within
the same political system (e.g. implementation problems in taxation, agricultural,
and environmental policies in Argentina), or of one policy issue in one country
across a long period of time (e.g. the regulation of interest group activities in
agricultural policy-making in the USA, from 1933 to 2001). In the former case,
what variation is on the characteristics of different public policies, and hence the
comparison is between public policies. In the latter case, the variation is on
underlying political and/or legal and/or economic conditions, and hence the
comparison is between these conditions. More complex designs and additional
assumptions are possible, but the underlying logic remains the same: policies are
not compared just because they happen to operate in different national systems,
but because they happen to represent a variation of a key variable.

7. Conclusions

31
Public policy is a term that covers numerous distinguishable (but not necessarily
distinguished) activities, such as problem definition, agenda-setting, deliberation,
decision-making, implementation, and assessment. Public policies come in all
kinds, colours and shapes, ranging from huge spending programmes such as
unemployment benefits or defence, to minute regulatory activities such as the
regulation of taxi licenses and fares in a small city. Furthermore, public policy
activities may be, and most commonly are, influenced by a whole set of
technological, social, economic, political, legal, and historical factors – all of
which may in turn be influenced by public policies. Given this complexity, it is not
surprising to find that public policy analysts have raised and analysed an
extraordinary number of theoretical, empirical and methodological questions.

More than a sub-discipline, then, PPA might be better defined as an alternative

lens to look at politics - whether one is interested in parties and elections, interest
groups and lobbying, implementation and bureaucracies, or public action and its
political and/or economic effects. Given this ambition, scholars of public policy
should be aware of the fundamental role played by research design and research
methods in their discipline. One way of doing this is to draw on the comparative
method. This solution is increasingly popular, and some of the cutting-edge
works in political science are now written by comparative public policy scholars.
Yet, even this solution carries some risks. Public policy analysts should not adopt
the comparative method for its own sake, but because it carries some clearly-
defined advantages.

References

Amin, A. and R. Palan (2001), ‘Towards a non-rationalist international political

economy’ Review of International Political Economy 8(4) Winter: 559-577.

Bachrach P.S. and M. Baratz (1963), ‘Decisions and non-decisions: an analytical

framework’, American Political Science Review, 57:641-651.

32
Bardach, E. (2006), ‘Policy Dynamics’, In M. Moran, M. Rein, and R.E. Goodin,
The Oxford Handbook of Public Policy. Oxford: Oxford University Press, 336-
366.

Bates, R.H., A. Greif, M. Levi, J-L. Rosenthal, and B.R. Weingast (1998), Analytic
Narratives, Princeton: Princeton University Press.

Baumgarnter, F.R., and B.D. Jones (1993) Agendas and Instability in American
Politics, Chicago: University of Chicago Press.

Bennett, C.J. and M. Howlett (1992) ‘The lessons of learning: reconciling theories
of policy learning and policy change’, Policy Sciences, 25: 275-294.

Bovens, M. and P. T’Hart (1996) (Eds.) Understanding Policy Fiascoes, New

Brunswick: Transaction Publishers.

Brunsson, N. (1989) The Organization of Hypocrisy. Talk, Decisions and Actions

in Organizations, John Wiley and Sons: Chirchester and New York.

Carpenter, D.P. (2001) The Forging of Bureaucratic Autonomy: Reputations,

Networks, and Policy Innovation in Executive Agencies, 1862-1928. Princeton:
Princeton University Press.

Dente, B. (ed.) (1990) Le Politiche Pubbliche in Italia, Bologna, Il Mulino.

Dobuzinskis, L. (1992) ‘Modernist and postmodernist metaphors of the policy

process: control and stability vs. chaos and reflective understanding’, Policy
Sciences, 25(4): 355-380.

Dowding, K. (2001) ‘There Must Be End to Confusion: Policy Networks,

Intellectual Fatigue, and the Need for Political Science Methods Courses in
British Universities’ Political Studies 49(1): 89-105.

Dye, T.R. (1976) What Governments Do, Why they do it, What Difference it
Makes, Tuscaloosa: University of Alabama Press

Epstein D. and S. O´Halloran (1999) Delegating Powers: A Transaction Cost

Politics Approach to Policy Making Under Separate Powers, Cambridge:
Cambridge University Press.

Faure, A. et al. (1995) La construction du sens dans les politiques publiques ,

Paris: L’Harmattan.

Fischer, F. (1992) ‘Reconstructing policy analysis: A postpositivist perspective’

Policy Sciences, 25(3), 333-339.

Fischer, F. (2003) Reframing Public Policy, Oxford: Oxford University Press.

Franchino, F. (2007) The Powers of the Union: Delegation in the EU, Cambridge:
Cambridge University Press.

33
Haas, P.M. (1992) ‘Introduction: Epistemic Communities and International Policy
Coordination.’ International Organization 46(1), 1-35.

Hajer, M.A. and H. Wagenaar (2003) (Eds.) Deliberative Policy Analysis:

Understanding Governance in the Network Society, Cambridge: Cambridge
University Press.

Hay, C. (2002) Political Analysis, Hampshire: Palgrave.

Heclo, H. (1972) ‘Review article: policy analysis’, British Journal of Political

Science, 2: 83-108.

------ (1974) Modern Social Politics in Britain and Sweden, New Haven: Yale
University Press.

Hood, C. and M. Lodge (2006) ‘Pavlovian Innovation, Pet Solutions and

Economizing on Rationality? Politicians and Dangerous Dogs’, in J. Black, M.
Lodge and M. Thatcher (Eds.) Regulatory Innovations: A Comparative Analysis,
Cheltenham, Edward Elgar.

Huber, J.D. and C.R. Shipan (2002) Deliberate Discretion? The Institutional
Foundations of Bureaucratic Autonomy, Cambridge: Cambridge University Press.

Hutter, B. (2006) The role of non-state actors in regulation, CARR Discussion

Paper 37, London, Centre for the Analysis of Risk and Regulation.

Ingram H. and S. Rathgeb-Smith (1993) (Eds.) Public Policy for Democracy,

Washington (DC): Brookings Institution

Ingram H. and A. Schneider (1993) ‘Constructing citizenship: the subtle

messages of policy design’, in Ingram H. and S. Rathgeb-Smith (Eds.) Public
Policy for Democracy, Washington (DC): Brookings Institution, 68-94.

Jacobsen, J.K. (1995) ‘Much ado about ideas. The cognitive factor in economic
policy’, World Politics, 47(2), 283-310.

James, O. (2000) ‘Regulation inside Government: Public Interest Justifications

and Regulatory Failures' Public Administration 78(2): 327-343.

Jobert, B. and P. Muller (1987) L´Etat en Action: Politiques Publiques et

Corporatismes. Paris: Presses Universitaires de France.

John, P. (2003) ‘Is There Life After Policy Streams, Advocacy Coalitions, and
Punctuations: Using Evolutionary Theory to Explain Policy Change?’ Policy
Studies Journal, 31(4), 481-498.

34
Katznelson, I. (1998) ´The Doleful Dance of Politics and Policy: Can Historical
Institutionalism Make a Difference?´, American Political Science Review, 92(1),
191-197.

Kingdon, J.W. (1995) Agendas, Alternatives and Public Policy, New York:
Longman, 2nd edition.

Knill. C. (2005) ‘Cross-national policy convergence: concepts, approaches and

explanatory factors’ Journal of European Public Policy, 12, Number 5 / October
2005 764 – 774.

Laffont, J.J. and D. Martimort (2002) The Theory of Incentives: The Principal-
Agent Model. Princeton, Princeton University Press.

Lindblom, C.E. (1965) The Intelligence of Democracy, New York: The Free
Press.

------ (1975) ‘The sociology of planning: thought and social interaction’, in M.

Bornstein (ed.) Economic Planning, East and West, Cambridge, MA: Ballinger,
23-67.

------ (1990) Inquiry and Change. The Troubled Attempt to Understand and
Shape Society, New Haven: Yale University Press.

Lindquist, E.A. (1988) 'What do decision models tell us about information use?',
Knowledge in Society, 1(2), 86-111.

Lowi, Th.J. (1964) ´American Business, Public Policy, Case Studies, and Political
Theory.´ World Politics, 16(4), 677-715.

------ (1969) The End of Liberalism: The Second Republic of the United States.
New York: W.W. Norton.

------ (1972) ‘Four Systems of Policy, Politics and Choice.’ Public Administration
Review 33(4), 298-310.

Mahoney, J. and G. Goertz (2004) ´The Possibility Principle: Choosing Negative

Cases in Comparative Research´ American Political Science Review, 98(4), 653-
669.

Mahoney, J. and D. Rueschemeyer (2003) (Eds.) Comparative Historical

Analysis in the Social Sciences. Cambridge: Cambridge University Press.

Majone, G. (1989) Evidence, Argument, and Persuasion in the Policy Process.

New Haven: Yale University Press.

March, J.G. and J.P. Olsen (1989) Rediscovering Institutions: The Organizational
Basis of Politics. New York: Free Press.

35
Marin, B. and R. Mayntz (1991) (eds.) Policy Networks. Empirical Evidence and
Theoretical Considerations, Frankfurt a.M./Boulder, CO: Campus/Westview.

May, P.J. (1992) ‘Policy learning and failure’, Journal of Public Policy, 12(4), 331-
354.

McCubbins, M., R. Noll, and B. Weingast (1987) ‘Administrative Procedures as

Instruments of Political Control.’ Journal of Law, Economics, and Organization
3(2), 243-277.

McCubbins, M., and T. Schwartz (1984) ‘Congressional Oversight Overlooked :

Police Patrols versus Fire Alarms.’ American Journal of Political Science 2(1),
165-179.

McGowan, L., and S. Wilks (1995) ‘The First Supranational Policy in the
European Union: Competition Policy’ European Journal of Political Research
28:141-169.

Milward, B.H. (1980) ‘Policy entrepreneurship and bureaucratic demand

creation’, in H.M. Ingram and D.E. Man (Eds.) Why Policies Succeed or Fail,
London: Sage.

Moe, T.M. (1985) ´Control and Feedback in Economic Regulation: The Case of
the NLRB.´ American Political Science Review, 79(4), 1094-1116.

Moran, M. (2003) The British Regulatory State: High Modernism and Hyper
Innovation, Oxford: Oxford University Press.

Moran, M., M. Rein and R. E. Goodin (2006) (Eds.) The Oxford Handbook of
Public Policy, Oxford: Oxford University Press.

Muller, P. and Surel, Y. (1998) L'analyse des politiques publiques. Paris:

Montchrestien.

Natali, D. (2005) ‘Europeanization, policy arenas, and creative opportunism: the

politics of welfare state reforms in Italy’ Journal of European Public Policy, 11(6):
1077 - 1095

Owens, S., T. Rayner and O. Bina (2004) ‘New agendas for appraisal: reflections
on theory, practice, and research’, Environment and Planning A, 36, 1943-59.

Pierson, P. (2004) Politics in Time: History, Institutions, and Social Analysis.

Princeton: Princeton University Press.

Pollack, M.A. (2003) The Engines of European Integration: Delegation, Agency,

and Agenda Setting in the EU. Oxford: Oxford University Press.

36
Radaelli, C.M. (1995) ‘The role of knowledge in the policy process’ Journal of
European Public Policy, 2(2), June 1995, 159-183.

------ (2005) ‘Diffusion without convergence: How political context shapes the
adoption of regulatory impact assessment’ Journal of European Public Policy,
12(5) October: 924-943.

Regonini, G. (2001) Capire le politiche pubbliche, Bologna: Il Mulino.

Rhodes, R.A.W. (1997) Understanding Governance. Buckingham: Open

University Press.

Richardson, J. (2000) ‘Government, Interest Groups, and Policy Change.’

Political Studies, 48, 1006-1025.

---- (1982) (ed.) Policy Styles in Western Europe, London: Allen & Unwin.

------ and G. Jordan (1979) Governing under Pressure, Oxford: Robertson.

Rochefort D.A. and R.W. Cobb (1994) (Eds.) The Politics of Problem Definition,
Lawrence: University Press of Kansas

Rose, R. (1990) ‘Inheritance Before Choice in Public Policy.’ Journal of

Theoretical Politics 2(3), 263-291.

Ross, S. (1973) ´The Economic Theory of Agency: The Principal´s Problem.´

American Economic Review, 63(2), 134-139.

Sabatier P. (2007) (ed) Theories of the Policy Process, Boulder: Westview. 2nd
edition.

Scharpf, F.W. (1973) Planung als politischer Prozess. Frankfurt a.M.: Suhrkamp.

Shadish, W. R., Cook, T. D., and Leviton, L. D. (1991) Foundations of Program

Evaluation: Theories of Practice, Newbury Park, CA: Sage Publications.

Shepsle, K.A. (1979) ‘Institutional arrangements and equilibrium in

multidimensional voting models.’ American Journal of Political Science 23(1), 27-
59.

Steinmo, S. (2003) ‘The evolution of policy ideas: tax policy in the 20 th century’,
British Journal of Politics and International Relations, 5 (2), 206-236.

Swank, D. and S. Steinmo (2002) ‘The New Political Economy of Taxation in

Advanced Capitalist Democracies’, American Journal of Political Science 46 (3):
477-89.

37
Thelen, K. (1999) ´Historical Institutionalism in Comparative Politics.´ Annual
Review of Political Science, 2, 369-404.

Tsebelis, G. (2002) Veto Players: How Political Institutions Work. Princeton:

Princeton University Press.

Weiss, C. H. (1999) ‘The interface between evaluation and public policy’,

Evaluation, 5:4, 468-86.

Wilks, S. and M. Wright (1987) (Eds.) Comparative Government-Industry

Relations: Western Europe, the United States, and Japan, Oxford, Clarendon
Press.

Williamson, O.E. (1985) The Economic Institutions of Capitalism. New York: Free
Press.

------ (1996) The Mechanisms of Governance, Oxford: Oxford University Press.

38