Advanced Introduction to Public Policy

Second Edition
Elgar Advanced Introductions are stimulating and thoughtful intro-
ductions to major fields in the social sciences, business and law, expertly
written by the world’s leading scholars. Designed to be accessible yet rig-
orous, they offer concise and lucid surveys of the substantive and policy
issues associated with discrete subject areas.
The aims of the series are two-fold: to pinpoint essential principles of
a particular field, and to offer insights that stimulate critical thinking. By
distilling the vast and often technical corpus of information on the subject
into a concise and meaningful form, the books serve as accessible intro-
ductions for undergraduate and graduate students coming to the subject
for the first time. Importantly, they also develop well-informed, nuanced
critiques of the field that will challenge and extend the understanding of
advanced students, scholars and policy-makers.
For a full list of titles in the series please see the back of the book. Recent
titles in the series include:
Politics of International Human Creative Industries
Rights John Hartley
David P. Forsythe
Global Administrative Law
Community-based Conservation Sabino Cassese
Fikret Berkes
Housing Studies
Global Production Networks William A.V. Clark
Neil M. Coe
Global Sports Law
Mental Health Law Stephen F. Ross
Michael L. Perlin
Public Policy
Law and Literature B. Guy Peters
Peter Goodrich
Advanced Introduction to

Public Policy
SECOND EDITION

B. GUY PETERS
Maurice Falk Professor of Government, University of
Pittsburgh, USA

Elgar Advanced Introductions

Cheltenham, UK • Northampton, MA, USA

© B. Guy Peters 2021

All rights reserved. No part of this publication may be reproduced, stored

in a retrieval system or transmitted in any form or by any means, electronic,
mechanical or photocopying, recording, or otherwise without the prior
permission of the publisher.

Published by
Edward Elgar Publishing Limited
The Lypiatts
15 Lansdown Road
Cheltenham
Glos GL50 2JA
UK

Edward Elgar Publishing, Inc.

William Pratt House
9 Dewey Court
Northampton
Massachusetts 01060
USA

A catalogue record for this book

is available from the British Library

Library of Congress Control Number: 2020950949

ISBN 978 1 78990 826 8 (cased)

ISBN 978 1 78990 828 2 (paperback)
ISBN 978 1 78990 827 5 (eBook)
06
Contents

List of tablesix
Prefacex

1 Public policy: a design perspective 1

Introduction1
The design perspective 5
Design and evolution 8
Rationality, risk and design 9
The state and its allies 10
Contents of the book 11

2 Policy problems 14
Why have public policy at all? 15
Market failure 16
Social failures 18
Characteristics of policy problems  19
Boundary spanning problems 20
Public goods and divisibility 21
Scale22
Solubility24
Complexity26
Certainty and risk 28
Tragic choices 30
Monetization31
Summary32

v
vi ADVANCED INTRODUCTION TO PUBLIC POLICY

Unstructured and wicked problems 33

Some common issues in defining problems 35
Coordination as a policy problem 36
Summary 37

PART I MAKING DECISIONS ABOUT POLICY

3 Models of policymaking 40
Approaches to decision-making and their
relationship to policymaking 42
Rationality  43
Bounded rationality 47
Institutional analysis and development 53
Multiple streams and the garbage can  55
Political models of policymaking 58
Stages models of policymaking 58
Subsystems: policy causes politics 60
The advocacy-coalition framework  63
Powering, puzzling, and so on 65
Contingent models and choices 67
Summary: multiple models, but what is missing? 69

4 Agendas, agenda-setting and framing 71

Agendas and agenda-setting 72
The issue-attention cycle 78
Focusing events and agenda-setting 80
Framing82
Summary84

PART II POLICY INTERVENTIONS

5 Designing intervention and implementation  87

Implementation as a component of the policy process 89
Implementation outcomes as probabilities 89
Forward and backward mapping for design 91
Multiple actors: implementation structures and networks 92
 CONTENTS vii

Summary94
Gaining compliance 94
Drift and sabotage 95
Barriers to implementation 97
Street-level bureaucrats and implementation 100
What is implementation success? 102
Interventions and implementation 103
Summary105

6 Policy instruments 107

Classifying policy instruments 108
Political science approaches to instruments  110
Constructivist perspectives on instruments 113
The psychology and sociology of instruments:
nudge and its allies  114
Evaluating the classifications 115
Choosing instruments 116
Individual decision-makers 116
Institutions and instrument choice 118
Ideas120
Interests121
Summary 122
Evaluating instruments  123
Political features of instruments 124
Economic features of instruments 125
Administrative criteria of policy instruments 126
Ethical criteria 128
Summary 130
The Swiss Army Knife of government 131
Summary132

PART III EVALUATING POLICY

7 Evaluating public policy: an introduction 135

The process of evaluation 136
viii ADVANCED INTRODUCTION TO PUBLIC POLICY

Barriers to effective evaluation 140

Performance management as an alternative to evaluation 142
Experiments and evaluation 145
Evaluation as feedback 146
Evaluation and learning 148
Conclusions149

8 Evaluating public policy: the utilitarian dimension 151

The logic of cost-benefit analysis  152
Measuring costs and benefits 155
Problems in cost-benefit analysis 156
Extensions of cost-benefit analysis 160
Summary163

9 Normative and ethical analysis of policy 165

Fairness166
Autonomy and freedom of choice 168
Preservation of life 170
Stigma173
Truth-telling, many hands and dirty hands 175
Summary178

10 Conclusion: policy success and failure 180

Going beyond individual policies 183
Reactions to policy failure: governing in the shadows 185
Where do we go from here? 191

References192
Index211
Tables

2.1 Types of complexity 26

2.2 Choices with risk 32
3.1 Stages models of policymaking 59
3.2 Policy causes politics: Lowi 61
3.3 Policy causes politics: James Q. Wilson 62
6.1 Examples of types of policy instruments 110
6.2 Resources available to government 112

ix
Preface

The study of public policy is understanding what governments do and

their effects on citizens. Although policy analysis is central to that under-
standing, it is by no means a simple undertaking. From an academic
perspective the analysis is complex because it involves multiple disciplines
– political science, economics, law, sociology and philosophy, among
others. These disciplines all provide insights into policy, but they also
provide different and perhaps conflicting perspectives on policy issues.
Therefore, making decisions about what are good answers to policy prob-
lems is difficult.

At a more practical level some of the same concerns about public policy
emerge. Given all the different perspectives on policy and policymaking,
how can decision-makers make good choices when they make the deci-
sions that will shape policies? Can the analytics provided by academic
discussions provide effective guidance for those decision-makers or are
they better advised to rely on their own intuition and judgment? Although
this book is directed primarily to an academic audience, I would not
have written it if I did not believe that it was relevant for the real world.
The academic models developed here and elsewhere in the literature are
abstracted from reality, but they also illuminate that reality.

One way to consider policy analysis from both the academic and prac-
tical perspectives is to use a concept of policy design. The fundamental
argument behind this perspective is that making policy is a design science
very much like architecture and engineering. While the messiness of
the political process involved in making policy makes such a perspective
appear excessively optimistic, the analyst can still develop and attempt to
implement more coherent designs of policy. Even if that designing fails,
and few policies are adopted and implemented exactly as designed, begin-

x
 PREFACE xi

ning with a coherent conception of the policy is likely to produce a more

coherent result.

There are a number of colleagues and friends who should be thanked for
their part, some unknown to them, in the development of this volume.
Perhaps the deepest debt goes to Stephen H. Linder, my colleague at
Tulane University during the 1980s. We wrote several articles together
on ideas of policy design, as well as on other aspects of policy studies. He
was central to my thinking about policy as an exercise in designing, and
in attempting to understand the complex interaction among causation,
action and evaluation in studying public policy. Steve, and my other
colleagues at Tulane, were extremely important in helping me develop
ideas about public policy, and governing more generally. At the same time
I had the pleasure of working with several British policy scholars such
as Richard Rose, Brian Hogwood and Christopher Hood who helped
me create an even stronger foundation for developing a perspective on
public policy. Several former and current students at the University of
Pittsburgh – notably John Hoornbeek, Patrik Marier, Jose Luis Mendez
and Max Peterson – have also contributed directly and indirectly to
the development of this book, as has Maximillian Nagel from Zeppelin
University. Over the past several years I have learned a great deal from
my colleague Philippe Zittoun at the University of Lyon. And last, but not
least, I continue to profit from my personal and professional interactions
with Jon Pierre. We have been working together for some years on issues
of policy and governance, and that cooperation continues to be both fun
and productive.

I cannot, of course, blame any of the above-named colleagues for any

inadequacies of this book, although it would certainly have been less ade-
quate without them. And thanks also to Alex Pettifer and his colleagues
from Edward Elgar who have been both encouraging and patient.

B. Guy Peters
1 Public policy: a design
perspective

Introduction

Public policy is the set of activities that governments engage in for the
purpose of changing their economy and society (see Peters, 2013). Some
years ago Harold Lasswell (1936) defined politics as “Who Gets What”.
Lasswell was one of the first social scientists to study public policy in
a systematic manner, and his definition of politics emphasizes that the
political process is ultimately about producing benefits, and costs, for the
members of the society. In other words, public policy ultimately is about
the public. All the other things that we identify with politics – elections,
legislative voting, bureaucracy, and even the courts – are significant
primarily because they contribute to making and implementing public
policy.

Any number of traditional academic disciplines should be involved in the

study of public policy. This book is oriented toward the politics of policy
but economics, law, sociology and ethics also make important contribu-
tions to understanding how governments intervene into the economy and
society, and these disciplines inevitably will make numerous appearances
in the discussions here. In addition, various substantive areas of expertise
such as public health, education and climatology are relevant for indi-
vidual policy areas. Thus, to understand public policy involves bringing
together a range of information and expertise, all focused on changing
conditions in the economy and society.

Even within political science there are alternative approaches to under-

standing how policy is made and implemented. For example, the conven-

1
2 ADVANCED INTRODUCTION TO PUBLIC POLICY

tional political science approach has been to use the policymaking cycle
(Jones, 1984; Hupe, 2007) beginning with agenda-setting and ending with
evaluation and feedback as a frame for understanding policy. There also
have been a several approaches to policy grounded on bounded ration-
ality and the constraints imposed on decision-making by the absence of
full information. Still other approaches are based on the manner in which
the public agenda is determined (Baumgartner and Jones, 2010). And
even more approaches focus on characteristics of policy itself (Hood and
Margetts, 2009; Peters, 2021).

The various approaches to public policy, in political science and across

the variety of other disciplines, provide a number of insights into public
policy. However, to be able to link all these perspectives requires some
more encompassing framework of perspective. I will be attempting to
make the linkages in policy through a design perspective (see Linder and
Peters, 1984; Schneider and Ingram, 1988; Peters and Fontaine, 2021).
Like engineering or architecture, good policy analysis involves bringing
together a series of components and processes into a designed interven-
tion into the economy and society.

The political process and the complexity of the targets of policy may make
such well-designed interventions more difficult to achieve than for engi-
neering or architecture. Unlike metal or concrete, citizens have minds of
their own and therefore making policy depends upon the extent to which
citizens behave as expected. That said, however, for the analyst thinking
about what a good design would be enables him or her to at least begin
considering strategies for intervention. That design may have to be more
contingent than those in other fields, but a clear design is perhaps the best
way to begin the intervention. The clearer the design, the more possible
it is to determine what goes wrong if the policy fails to achieve its goals.

The fundamental argument of this book is that we can understand public

policy through a design perspective, and also that we should consider
the appropriateness of policy interventions through a design perspective.
The argument for a design perspective on public policy is that there
needs to be some clear connection between the assumed causes of the
problem being addressed, the instruments used to attempt to remedy that
situation, and an understanding of what a desirable outcome would be.
In addition, the design requires a strategy for implementation. A design
perspective on policy emphasizes those multiple linkages, as well as the
 PUBLIC POLICY: A DESIGN PERSPECTIVE 3

need to consider policy interventions in a more holistic manner than is

generally characteristic of policy analysis.

Saying that a policymaker or an analyst wishes to utilize a design per-

spective on policy, however, is only the beginning of addressing the issue.
The participants in the process must then elaborate exactly what is meant
by that design. On the one hand, designing could be considered from
a positivist, “scientific” foundation. In such a perspective the nature of the
problem to be solved is taken largely as a given, and then various forms
of evidence and analysis are brought to bear on the problem to produce
a “solution”. This can be seen as a form of “normal science” in which the
analytics are conducted within a largely agreed upon policy paradigm.

This scientific approach remains the modal form of addressing public

policy problems, simply because so much policymaking is done within
specific policy domains for which the nature of the problem and the range
of possible solutions have been well defined. Organizations within the
public sector, as well as their clienteles, generally have been able to create
some agreement within their own policy domains, so that decisions being
made are largely at the margin, rather than about fundamental policy
issues. While these “silos” may generate real problems for considering
resolving larger policy issues they do provide the participants with some
predictability and some control over their own policy domains (Genieys,
2017). Likewise, policymakers and their analysts are primarily in the busi-
ness of solving particular problems rather than thinking about the more
intellectual foundations of policy analysis.

Another strand of the political science literature on policymaking and

policy design has dealt with this issue of design primarily through con-
sidering the ideas informing a design, and to some extent through the
framing of possible policy interventions (see Stone, 1997). In this analytic
perspective policies, and the problems that they are meant to resolve, are
primarily social constructions rather than “real” in the positivist sense
of that term. Giandomenico Majone (2002b), for example, expressed his
discontent with the more scientific approaches to regulatory policy in
Europe and stressed the desirability of using rhetorical and legal methods
for understanding and shaping policy.

Therefore, in this constructivist view, policy designs are shaped through

argumentation rather than through a rational analytic process (Gottweiss,
4 ADVANCED INTRODUCTION TO PUBLIC POLICY

2007; Fischer and Gottweiss, 2012). Rather than having a single answer
to the policy problem different sets of argument ideas will be associated
with different designs, and may produce different types of interventions.
Further, the problem itself may not be taken as a given and will be subject
to interpretation – indeed defining the policy problem may be the most
important aspect of the entire process, given that once it is defined and
understood in a particular way then the remainder of the analysis pro-
ceeds from that perspective.

This constructivist conception of design argues that a particular set of

ideas defines the policy design (Béland and Cox, 2011). In addition,
those policy designs can be generated by picking ideas from other set-
tings, defined geographically or through policy domains (Schneider and
Ingram, 1988). This approach to design emphasizes the need for design
and the need for comparative analysis, but it does not go far enough in
identifying the elements of designs that need to be included, and linked
together, to construct an effective policy intervention. Constructivist
perspectives on policy design have become very popular among govern-
ments in the guise of “evidence-based policymaking” (see Pawson, 2006;
Botterill, 2013) but in this version the approach appears to assume that
policies may be transferred more easily from one setting to another than
may actually be possible.

In addition to those two seemingly opposite approaches described above,

Davis Bobrow and John Dryzek (1987) have discussed policy design from
the perspective of various theoretical perspectives in the social sciences
and their insights into policy. Whereas the ideas discussed by Deborah
Stone and others (Fischer and Gottweiss, 2012) pertain to the policy
issues themselves and their frames, the approach to design coming from
Bobow and Dryzek is more concerned with the social scientific theories
that can be brought to bear in understanding and solving those issues.
Bobrow and Dryzek argue that major theoretical perspectives on policy
each contains an intellectual design that also contains a set of possible
solutions. Many social scientists do not consider the policy implications
of their theories, but they are certainly present and need to be considered.

Yet another version of policy design stresses implementation and the

mix of policies and policy instruments that may be involved in particular
policy areas (Howlett and Rayner, 2007; van Hulst and Yanow, 2016).
This perspective makes the important point that policies are rarely acting
 PUBLIC POLICY: A DESIGN PERSPECTIVE 5

alone, and many policies and policy instruments may be acting upon
the same targets. On the other hand, however, this perspective on policy
design is perhaps too narrow because it focuses almost exclusively on
implementation and policy instruments rather than on the full range of
decision-making that must go into design. Still the emphasis on design-
ing in the context of crowded policy spaces forces the designer to adopt
a more comprehensive range of policies and instruments.

Although it is easy to say that policy design is a useful approach to public

policy, that is only the beginning of thinking about the process of design-
ing policies. There are alternative approaches to the process, and to policy
considered more generally, as well as alternative theoretical perspectives
on governing. And there is the political process itself that can place
numerous barriers in the way of any would-be designer. What follows is
an attempt to understand better what is needed for an adequate design of
a policy, and the process of producing such a design.

The design perspective

The design perspective presented in this book will contain three com-
ponents. The first is a model of causation. Policies are designed to solve
some problem or problems in society, and no policy can be expected to be
effective unless it has a clear conception of socio-economic dynamics that
are producing the problem to be solved. Further, policy problems them-
selves need to be understood and conceptualized adequately in order for
effective interventions to be designed (see Hoornbeek and Peters, 2017).
These policy problems may be thought of as somehow objective and
clearly identifiable, but they are subject to, and sometimes the product of,
politics and interpretation (Fischer and Gottweiss, 2012).

It is important to understand that there may be no single conception of

a policy problem, and therefore no single conception of how to address
that problem. For example, what type of problem is the drug problem?
Is it a law enforcement issue, or a health issue, or an issue of social dis-
location (see Payan, 2006)? Can gun control be framed as a public health
policy rather than a law enforcement issue (Hemenway and Miller, 2013)?
This framing of issues is a crucial political process, determining not only
6 ADVANCED INTRODUCTION TO PUBLIC POLICY

the types of intervention to be undertaken but also the actors within

the public sector who will gain the budgetary and personnel resources
associated with the policy. And, of course, the accuracy of the frame may
determine whether the policy problem is actually “solved”,1 and what
types of outcome individual citizens may expect.

A model of intervention is the second component of any design frame-

work for public policy. What options are available for the public sector,
and its allies in the private sector, to intervene and to alter conditions in
the society. Understanding interventions requires something of a strate-
gic or governance approach to policy. That is, to be effective policymakers
have to understand the policy process and the variety of points at which
they can intervene. To some extent understanding policy interventions
requires adequate understanding of the causal processes described above,
so that the policymaker knows which aspects of a policy are most amena-
ble to change.

These points for intervention also may be defined by the stage in the
policy process or it may be defined by institutions. For example, in federal
systems individuals or organizations must decide which of several possi-
ble levels of government they will select as the venue when they attempt to
change policies (Benz and Broschek, 2012). And it may be easier to alter
outcomes at the implementation stage than at the formal policymaking
stages of the policy process (Exworthy and Powell, 2004), given that
partisan politics will be less relevant, or at least apparent, during imple-
mentation. Therefore, the would-be policymaker needs to make a number
of careful strategic choices that are likely to influence his or her success.

A central element in designing interventions is the selection of policy

instruments. Governments have a substantial “tool chest” that they
can use to attempt to generate change during the implementation
process (Hood and Margetts, 2009; Lascoumbes and Le Gales, 2007) and
those tools have a number of characteristics that must be understood
for effective design. Further, as governments have moved away from
command and control instruments toward softer modes of intervention
(see Salamon, 2001a) the entire nature of policy instruments, as well as
intervention more generally, has to some extent been transformed.

The selection of instruments is not, however, a simple technical process

but rather is also political as the instruments themselves create winners
 PUBLIC POLICY: A DESIGN PERSPECTIVE 7

and losers just as do the substantive nature of the policies themselves.

The shift away from “command and control” instruments toward softer,
negotiated forms of intervention, for example, creates a very different
dynamic within policymaking that involves numerous stakeholders in the
selection of policy. Further, instruments such as the SNAP program2 in
the United States that builds coalitions across conventional policy areas
are likely to be more stable than are instruments based more on individual
organizations and their clients rather than coalitions.

Finally, understanding policy requires a model of evaluation. What is

good policy, and what is a good outcome from the intervention of govern-
ment? This evaluation of policy outcomes involves several sets of criteria.
The most commonly applied criteria are economic, and in particular those
associated with cost-benefit analysis (Mishan and Quah, 2007). But there
are also normative criteria, including political criteria, that also should be
used to assess how well the public sector has been performing as it makes
its interventions (Boston, Braddock and Eng, 2010; Poama, 2019). These
normative criteria rarely provide the neat, quantitative answers provided
by the economic assessments, but they are equally important in under-
standing success and failure in policymaking.

A final point about this design perspective on public policy is that I have
been discussing the designs policy by policy, but almost all policies are
embedded in complex patterns of cooperation and competition with
other policies and organizations. Although I will not emphasize issues
of coordination and collaboration in this book (see Peters, 2015), it is
important to remember that these connections do influence the success
and failure of any individual policy. And the services delivered to citizens
and businesses are sub-optimal because of the failures to integrate the
services being provided. Indeed, addressing some issues such as crime
may be better done, in the long run, by beginning with programs usually
labeled “social policy” and “economic inequality” than with those labeled
“policing” or “crime prevention”. The world of public policy is complex
and interconnected and taking the linkages into account may be crucial
for success.
8 ADVANCED INTRODUCTION TO PUBLIC POLICY

Design and evolution

The design perspective described above tends to assume that the policy
designers, whoever they may be, will create a template for policy that
then proceeds through other stages of the policy process. With a plan,
whether implicit or explicit, for intervention and then a model of evalua-
tion, a design perspective seeks to put the chosen template for policy into
effect. This top-down perspective on policymaking is common among
decision-makers within the center of government, but generally appears
excessively controlling to actors outside those inner circles.

The more purposive conception can be contrasted with a more evolution-

ary and adaptive perspective. In that adaptive view any initial design for
a policy is, and should be, subject to modification as it is elaborated and
implemented (see Browne and Wildavsky, 1984). In this view any design
made through a political hierarchy or by planning is likely to require
adaptation as it confronts the complexity of the environment. The chal-
lenge in such a perspective on policy is maintaining the intentions of the
formulators of the policy while attempting to make it perform adequately
in the real world.

For some political actors that “drift” of policy (see Béland, Rocco
and Waddan, 2016) during implementation is considered a significant
problem. On the one hand, the tendency of bureaucrats to act on their
own interpretation about the policy often concerns conservatives who
believe that those bureaucratic decisions tend to expand the role of the
agencies making the decisions (McCubbins, Noll and Weingast, 1989).
And even on democratic and legal grounds there are reasons to object
to decisions that go beyond the intentions of the legislators who had the
constitutional powers to make the decisions (Lane, 1983). Policy drift
may become even more of an issue as policymaking and implementation
become more collaborative, and involve social actors even more directly
in the process (Ansell and Gash, 2007).

The conflicting ideas about the desirability of evolution and adaptation

in policymaking reflect underlying legal, managerial and political con-
ceptions about policy. On the one hand, it is easy to argue that designing
policy for once and for all is difficult and perhaps misguided, and that
building in more adaptive capacity is important for success over an
 PUBLIC POLICY: A DESIGN PERSPECTIVE 9

extended period of time. On the other hand, legally there is some imper-
ative to implement the law as enacted. For managers in public adminis-
tration these different conceptions are reflected in the dual and perhaps
competing demands to make something happen and to follow the law.

Rationality, risk and design

This book will stress the creation of designs for public interventions in
the society. There are certainly some approaches to public policy that do
not appear to match this rationalist perspective, for example, the logic
of bounded rationality, incrementalism and the garbage can, and these
approaches tend to be dominant in political science. However, even
those seemingly unsystematic elements can be integrated into the design
framework. Or at a minimum these less structured elements of the policy
process can be judged against the realities of formal design of the policy.

Phrased the other way round, the concept of design provides a logic
against which the realities of policymaking can be evaluated. Even if the
dynamics of policy and the definition of the problems are not as clearly
delineated as might be desirable from a rationalist perspective, they must
still be understandable within the context of a policy process that will
produce some form of design at the end of the process. Risk and bounded
rationality are not so random that they cannot be investigated through
thinking about more systematic design and analysis of policy and policy
processes.

Approaches to policy such as bounded rationality and the garbage can

emphasize that uncertainty is always present when governments attempt
to intervene in their environments. Indeed, that uncertainty may be
increasing as an increasing number of “wicked” problems confront
governments (Rittel and Webber, 1973; Levin et al., 2012; Head, 2021).
A range of emerging issues such as climate change, resource depletion and
obesity all involve multiple values, conflicting goals and uncertain causal
processes, making addressing policy problems increasingly difficult, and
the politics of policy all the more contentious. The political contention
and the inherent instability of these problems require maintaining open-
ness in the design of interventions and in the implementation of those
10 ADVANCED INTRODUCTION TO PUBLIC POLICY

interventions. A hierarchical, top-down conception of designing may

exclude important actors and come to premature conclusions about the
right policy to address a problem.

Any meaningful conception of policy design, therefore, must build in

uncertainty and risk. This point has been emphasized by scholars such as
Yehezkel Dror (1986; Boin, 2004) but remains less integrated into most
studies of public policy than perhaps it should be. Further, the importance
of risk has been highlighted by the pandemic of 2020 and the resultant
economic difficulties (Watson and Mullen, 2020). It is one thing to argue
that the rationality involved in policymaking is inherently bounded, but
integrating that understanding of uncertainty into the design of public
sector interventions is more difficult. As we will point out in Chapter
4 there are analytic methods of utilizing risk when making decisions
about policies, and these methods should be incorporated into the design
process.

Finally, any consideration of policy design must include careful consider-

ation of the process that generates the design. Thinking about designing
appears to imply a top-down, hierarchical process that minimizes political
involvement of the public. In any democratic political system imposition
is generally impossible, given the complex political and administrative
process for formulating and implementing policy.3 That political process
has been increasingly opened to a wider range of political actors so the
capacity to control the final outcome has been further weakened (Torfing
et al., 2012). Although not hierarchical, and often apparently less than
fully rational in the usual sense of that term, this remains a design process.
Designing policy does not mean that the idea will come from one source
or contain one clear idea, but rather this is more an analytic model to
understand what is being generated through the policy process.

The state and its allies

This book will focus on the role of the public sector in making and imple-
menting policy. We are, however, cognizant of the role that non-state
actors play in this process. Both market actors and civil society organ-
izations are involved in making and implementing public policies, and
 PUBLIC POLICY: A DESIGN PERSPECTIVE 11

this role has been increasing in most industrialized democracies, and to

a lesser extent in other countries (Roberts, 2004). The ideologies of New
Public Management (Christensen and Laegreid, 2007) and neo-liberalism
more generally have tended to emphasize the use of market mechanisms
in policy and administration (see Gingrich, 2011). Alternatively, advo-
cates of collaboration and network governance have emphasized the role
of social actors in policy and administration (Donahue and Zeckhauser,
2011).

The majority of the ideas to be developed in this book will be as applicable

to policymaking with the involvement of non-governmental actors as
they will be for policymaking that is more state-centric. The same issues
of designing policy must be confronted with or without the direct involve-
ment of non-governmental actors. Some advocates of more collaborative
styles of policymaking would argue that the process will produce better
policies if all the stakeholders are involved, and/or if market-style instru-
ments are adopted for implementation. But other analysts would argue
that difficult policy choices can be made more readily if they are shielded
from outside influences and made by responsible government elites
(Fung, Graham and Weil, 2007; Gilens and Page, 2014).

Whether more effective decisions can be made or not, it is likely that

contemporary policymaking will involve a wider array of actors than in
the past. The political mobilization of more interest groups, the emphasis
on openness and transparency in governing, and additional legal require-
ments for involving social actors all lead to the greater involvement of
stakeholders in the policy process. Although justified primarily in demo-
cratic terms this inclusiveness may not, in fact, include all relevant actors
– most importantly the public in general – so that conventional repre-
sentative democracy continues to have a major role to play in legitimating
policy choices and in attempting to include a wide range of perspectives
on good policy.

Contents of the book

The following chapters of this book will address the three components of
policy design mentioned above. As already noted the book will emphasize
12 ADVANCED INTRODUCTION TO PUBLIC POLICY

the political dimensions of public policy, although it will certainly attempt

to include other elements (especially the role of economic analysis)
in public policy. When attempting to understand the nature of policy
problems, we will also have to think about those problems in the social
and economic context that define them as policy problems, as well as
considering their political aspects.

Although not discussed as a separate component of policy design,

decision-making is a pervasive element of policy. The other three com-
ponents of policy design all involve decision-making and there is an
extensive literature on how policies are made. The approaches to policy-
making range from very rationalist perspectives to perspectives based on
almost random confluence of streams. All these perspectives have some
validity but also have some intellectual and practical difficulties. Chapter
2 will assess these perspectives on decision-making and argue for a more
contingent perspective on their use.

The first part of the book will discuss models of causation of policy
problems. The first chapter in Part I will discuss the political processes
involved in framing policy issues and in agenda-setting (Chong and
Druckman, 2007). These are crucial aspects of initiating the policy
process, and include numerous political and social actors in that process.
The second chapter will discuss the nature of policy problems themselves,
pointing to the various dimensions of these problems that in turn shape
the policy process and the possible modes of intervention.

The second part of the book will discuss approaches to intervention. The
first of these two chapters will discuss intervention in strategic terms,
including political and governance strategies that establish priorities and
must also include assessments of risk. The second chapter within Part
II will be a detailed examination of policy instruments and the means
through which public actors, and their private sector allies, can attempt
to alter conditions in the society. Policy instruments are complex enti-
ties that require evaluation and careful comparison, and often are used
together to produce more effective interventions.

After governments intervene into their societies, they need to assess the
consequences of those actions. Therefore, the third part of the book will
discuss means of evaluating public policies. Just as there are alternative
models of causation so too are there alternative models of evaluation.
 PUBLIC POLICY: A DESIGN PERSPECTIVE 13

Economic evaluations of policy – generally cost-benefit and allied forms

of analysis – are the most common modes of evaluation. The alterna-
tive version of evaluation – using ethical criteria – may give alternative
answers about the success or failure of a policy. And in addition to these
criteria there is the ongoing political evaluation being undertaken by both
citizens and political leaders.

Finally, there will be a short concluding chapter that attempts to integrate

the three components of design and consider how they interact in the
practice of policymaking and implementation. The conclusion also will
discuss the problems raised by a design approach to public policy and
what future research will be required to make the approach even more
useful for understanding policy and actually designing interventions. The
principal emphasis of this book is on creating an academic understanding
of public policy, but that academic understanding can also be linked
directly to actual attempts to solve policy problems.

Notes
1. Solving a problem may be an excessively optimistic goal. In reality most pol-
icymaking is only amelioration and further, policymaking at one time may
address the problem but may also create new problems (Sieber, 1980).
2. Supplemental Nutrition Assistance Program, formerly Food Stamps, sup-
ports farmers by ensuring markets as well as the disadvantaged by subsidiz-
ing food purchases.
3. There are in emergency situations possibilities for governance by decree, but
even those may be constrained by procedural rules (Scheuerman, 2006).
2 Policy problems

In the best of all worlds there would be no social or economic problems

that would require the intervention of the public sector. But we do not live
in that world, and there are a myriad of problems in society that require
intervention. The public sector, and its allies in the private sector, may
not always want to intervene in these social conditions – whether for
ideological or for practical reasons – but often they are compelled to do
so. Public pressure and the very presence of some types of problems may
force intervention. To understand policy design we need to understand
the problems that are being addressed by public sector action, including
very broad questions about market failure and social failure that create
those problems.

In the most general sense a policy problem is a condition that some or

all citizens (and policymakers) find undesirable. These problems may
range from simple issues such as litter in the streets to pensions for
elderly citizens through to major foreign policy questions. Some of these
policy problems may be addressed through private action – citizens may
clean up litter in their own streets. Most of these problems will, however,
become a part of the political agenda. And with these problems also come
a range of alternative solutions that may or may not be suitable as means
of resolving the issue. This common sense conception of policy problems
is the beginning of all steps in making public policy.

This discussion of problems will identify some of the underlying char-

acteristics of policy problems. Most discussions of policy problems tend
to describe them according to the functional area within which they
occur – agriculture, education, defense, and so on. However, there may
be as much variance within each of these functional areas as there may be
across areas (but see Freeman, 1985). For example, the field labeled educa-
tion includes everything from pre-schools through research universities,
with markedly different issues, actors and politics. And most important

14
 POLICY PROBLEMS 15

for this discussion there are a number of different policy problems within
education that require very different forms of intervention. We therefore
need to think about not just the titles we see on government buildings
when defining policy problems but also their underlying features.

We also at times discuss policy problems in terms of the instruments that

may be used to solve them, or at least ameliorate them. We may say, for
example, that “this is a regulatory problem” because when confronted
with issues of this type most governments tend to choose regulation as
the appropriate instrument. But here there may be a number of options
for how the same problem is addressed by government. Labeling an issue
by a particular instrument may prevent more careful consideration of the
design options available to decision-makers.

This chapter will examine policy problems in three different ways. The
first is to examine the most fundamental issues in public policy – those
that define the need for the public sector to act at all. The second approach
to policy problems will be looking at the defining characteristics of the
problems that arise in more ordinary policymaking, for example, the
characteristics of the stakeholders in the policy domain. And finally,
and related closely to the second, I will discuss the concept of “wicked
problems” and its importance for policy studies. These problems (see
Rittel and Webber, 1973; Levin et al., 2012) are large-scale, complex and
extremely difficult to solve, but also are becoming more important for
governments with the emergence of issues such as climate change, food
scarcity and sustainable development.

Why have public policy at all?

The first issue about policy problems is why should the public sector
intervene at all in the functioning of the economy and society? This is
in part an ideological question, with individuals on the political right
arguing that the state is justified in intervening only in exceptional
circumstances, while those on the left believe that there are numerous
good reasons for action by the public sector (Madrick, 2009; Brook and
Watkins, 2010). But this question goes beyond ideologies, and there are
good economic and political reasons justifying the role of the public
16 ADVANCED INTRODUCTION TO PUBLIC POLICY

sector in contemporary societies. To some extent these reasons are also

viewed through ideological lenses, but there can be more objective analy-
sis that may help to resolve the differences, or at least provide support for
different ideological positions.

Market failure

From an economic perspective the fundamental argument for the inter-

vention of the public sector is market failure (Wallis and Dollery, 1999).
The neo-classical model of the economy is based on a set of assumptions
such as perfect information that often simply do not exist in the real
world. Given the absence of those preconditions, markets do not work
effectively as assumed, and the public sector may need to intervene to
rectify the problems being generated by market failure. Again there may
be an ideological discussion over how much market failure is sufficient
to justify the intervention of the public sector, but at some level the inca-
pacity of markets to function as expected will invoke the intervention of
government.

Public goods constitute a major category of market failures. A public good

is a good that, once created, is available to all consumers. Since individuals
cannot be excluded from consuming the good, it cannot be priced or
marketed. Examples of public goods include clean air, national defense
and flood control. Although having some characteristics of public goods,
public recreational areas and parks are subject to crowding effects – if
we all attempt to use the park none of us is likely to enjoy it very much.
If anyone attempts to create public goods and market them, those pro-
ducers will encounter free riders who will enjoy the good but not pay for
it. Therefore, governments and the use of tax money represent the only
possible efficient producers of that type of good.

The standard model of the market also assumes that the costs of pro-
duction are reflected in the selling price of the market, but in many cases
this is not true. The most obvious example is pollution. The social costs
of pollution – health problems, reduced property values, and so on – are
not reflected in the market price, and hence products causing that pol-
lution are underpriced (see Dasgupta and Ehrlich, 2013). To rectify that
 POLICY PROBLEMS 17

problem governments must either regulate the costs of these externalities,

or force the producers to pay for them through some mechanisms such
as pollution pricing (Sandmo, 2000). There can, however, also be positive
externalities but typically the producers of these cannot receive their
value. For example, if a hydroelectric dam creates recreational opportuni-
ties and increases property values of (now) waterfront property the firm
building the dam can rarely appropriate those increased values.

For markets to work effectively there needs to near perfect information.

Buyers in particular need to understand what they are buying to be able
to make efficient, and safe, choices among products. But the financial
scandals following 2008 revealed very clearly that consumers did not have
that information about the loans they were buying nor about many invest-
ment opportunities being offered to them by the financial sector. Given
that sellers have little incentive to provide full information, governments
must step in to try to force disclosure (in credit cards, for example) or to
regulate products for safety purposes.

Although monopolies are generally considered failures in the market.,

some products are natural monopolies and would be produced ineffi-
ciently through a competitive marketplace. For example, it would be inef-
ficient to have two suppliers of water operating in the same area, involving
multiple systems of pipes, pumps, and so on. It is therefore more efficient
to allow one monopoly supplier, whether it be government itself or a reg-
ulated private company.1 This preference for regulated markets is true
not only for goods like water where multiple systems of supply would
be inefficient, but also for goods that have very large returns to scale,
meaning that very large producers can be more efficient than multiple
smaller producers. In the case of natural monopolies or returns to scale
the public sector must intervene with regulations on rates and returns to
capital to prevent exploitation.

Collective action problems (Olson, 1965; Medina, 2007) can also be seen
as a form of market failure that may require the intervention of the public
sector. The tragedy of the commons (Hardin, 1968) is the classic example
of this type of problem. If there is common grazing land and all farmers
use it to the maximum possible then soon there will be no more grass and
everyone is worse off. The same problem arises with water for irrigation,
or the over-exploitation of fish stocks. While government intervention
18 ADVANCED INTRODUCTION TO PUBLIC POLICY

is the usual solution, private, cooperative actions can also address these
problems (Ostrom, 1990).

Finally, the economic distribution created by the market tends to be

skewed, with income and wealth being concentrated among a relatively
small number of individuals. High levels of economic inequality have
become largely unacceptable in mixed-economy welfare states, and
governments have been using taxing and spending measures to at least
build an income floor for citizens who have been less successful in the
market. This justification for public sector intervention is, however,
more ideologically contentious than the others already mentioned, and
redistribution has been off the political agenda for many countries for at
least the past decade . Further, the existing strategies appear to have been
ineffective in most counties, as levels of inequality tend to increase all over
the world (Xue, 2012).

Social failures

Failures in society that may necessitate the intervention of the public

sector are not so readily classifiable as are those in the economy, but they
are nonetheless real. Issues such as crime, poverty, family breakdowns,
school dropouts, and so on may have some economic element but they
also have a strong social and cultural component, And the absence of
clear categories, such as those in economics, make these issues all the
more difficult to address through public sector action. The dynamics of
these issues are subject to numerous interpretations and hence become
highly politicized.

The politicization of these real or assumed social failures has been most
apparent around the issue of poverty. The political right tends to attribute
the persistence of poverty to failures of the individuals and their family
structure, arguing that the disruption in social life and the absence of
a work ethic tends to be transmitted from generation to generation and
is exacerbated by most social programs (Nowrasteh and Cole, 2014). The
political left, on the other hand, argues that poverty and social exclusion
are more a function of the poor functioning of the market and the failure
of governments to intervene through social policies or economic regula-
 POLICY PROBLEMS 19

tions such as minimum wage laws that would produce a living wage for
anyone in work (Bernstein and Parrott, 2014).

The above discussion about market and social failures should not let us
forget that there are also governance failures (see Wolf, 1987). While
markets fail because of public goods, governance may fail because of
private goods when government power is used to advantage certain
segments of society with public money (see Lowi, 1967). Likewise, the
public sector may have “internalities”, meaning that public sector actors
sometimes make choices that move away from allocative efficiency and
therefore impose costs on society. These internalities reflect the unequal
distribution of political power and ability of some groups to extract more
from the public purse than might be justified on economic or moral
grounds.2 Further, at times governments create un-natural monopolies
for themselves that stifle competition in areas such as telecommunica-
tions and energy.

Characteristics of policy problems

The above discussion of market failure, social failure and governance

failure provides a broad interpretation of the problems motivating pol-
icymakers (see Peters, 2014b), but once we move from that very general
level to the consideration of individual policy initiatives a number of more
specific characteristics of policy problems become important for design.
As mentioned, simply thinking in terms of functional or instrumental
labels is inadequate for policy design, given the variance within individual
policy domains, and the multiple dimensions that may affect the capacity
for policymaking for each problem. Further, keeping proposed policy
solutions within the individual silos defined by public organizations and
interest groups may reduce the probabilities of finding more than mini-
mally effective solutions.

As well as classifying problems according to their functional categories,

policies may also be classified by the particular policy instruments used
to address them. This labeling is particularly true for regulation, with
a number of economic issues being described as “regulatory issues”. This
classification tends to assume that the only, or at least the most efficient,
20 ADVANCED INTRODUCTION TO PUBLIC POLICY

means of addressing an issue is through command and control regula-

tion. But that is not necessarily the case and some issues, for example,
pollution, that were once considered regulatory are now addressed
regularly through instruments such as taxes and charges (Morag-Levine,
2009). At even more of an extreme, the development of “nudge” and
other psychological approaches to policy can produce results without
direct interventions (Thaler and Sunstein, 2008). Again, this progression
of definitions of policies points to the necessity of considering the basic
issues of problems rather than using familiar categories.

Although I can identify a number of important characteristics of policy

issues there is no clear theoretical foundation that guides the selection of
these dimensions of analysis. The literature on public policy has developed
a number of ideas about problems but these have largely been developed
inductively (see Hoornbeek and Peters, 2017). However, although there
is no unifying theoretical frame these various characteristics of policy
problems remain useful for understanding the challenges for government
when they seem to intervene. The variations in problems can be related,
if only loosely, to the nature of the interventions that governments may
find effective.

Boundary spanning problems

Having said that the usual labeling of policy problems is inadequate,

one of the more important characteristics of problems is the extent to
which they are contained within the usual departmental and functional
boundaries. Those boundaries are usually discussed in the functional
terms discussed above, but may also include geographical boundaries.
With globalization and increasing relevance of multi-level governance,
policy problems clearly cut across geographical boundaries (McKibbin,
2007). This is especially true of wicked problems such as climate change
and food security. The boundaries between the public and private sectors
are increasingly permeable, and pose another variant of the need to cope
with policy problems across boundaries.

Although there are important variations within the functional policy

areas, policy problems that can be contained within a single functional
 POLICY PROBLEMS 21

area, geographical area or entirely within the public sector are easier to
manage than those that span boundaries (May, Jochim and Pump, 2010).
Within any functional area a limited number of public organizations and
policy ideas may be involved, while if the policy problem can be contained
to a single or a limited number of geographical areas the political conflict
may also be reduced.

While the simplicity of policy management may be enhanced by more

constrained policy problems, the opportunities may also be reduced.
Boundary spanning problems, whether real or framed as being such,
make apparent the possibilities for coordination and synergies among
programs (Peters, 2015). That coordination may be horizontal between
programs and organizations, or it may be vertical across levels of gov-
ernment, or it may be both, but in any case the overall performance of
policymaking may be improved.

Stated somewhat differently, the most important problems in govern-

ing cut across the conventional boundaries of policy and geography.
Therefore, to the extent that governments, and their counterparts in the
private sector, can find ways to cope with cross-cutting problems they are
more likely to be successful in addressing the major issues facing citizens.
For example, if economic policy is dealt with in the conventional manner
through standard monetary and fiscal policy mechanisms, some success
can be expected. If, however, this issue is conceptualized as “competitive-
ness policy” then a range of other possible contributions, for example,
education and technological innovation, can be used to address the
underlying issue and produce perhaps more dramatic results (see Sum
and Jessop, 2013).

Public goods and divisibility

We have discussed the need to create public goods as a general justifi-

cation for the intervention of the public sector into the economy and
society. That said, some particular policy problems require the creation of
public goods, while others involve creating private goods (those that allow
exclusion and can be provided for some individuals and not for others).
The difference between public and private goods helps demonstrate that
22 ADVANCED INTRODUCTION TO PUBLIC POLICY

the same nominal policy area can produce different types of problems.
For example, defense is usually discussed as a public good, given that the
military apparatus once created tends to defend all citizens. However,
defense procurement is more of a private good, with firms and areas of the
country competing for contracts that will benefit them. At the extreme,
in the United States Congress at times demands that the Department
of Defense purchase weapons systems it does not want, simply to keep
plants open in the districts of powerful members (Bennett, 2014).

In addition to understanding that some policy problems represent indi-

visible issues, for example, clean air, the difference between public and
private policy problems is important primarily because it may limit the
range of instruments that governments can utilize when attempting to
solve the problem. Many of the policy instruments available to govern-
ment (see Chapter 6) depend upon providing benefits or incentives to
individuals, but if the problem is indeed indivisible then more collec-
tive solutions (usually involving law and public organizations) will be
required.

Although in many instances the divisibility of a policy problem is objec-

tive, in other cases it may be politically constructed. For the advocate
of a particular policy one means of “selling” it politically is to convince
decision-makers that the problem is indeed an indivisible problem like
a public good. If that characterization is true then the problem can only
be addressed effectively through the intervention of the public sector.
And individual organizations within the public sector may also attempt
to define policy problems as being indivisible public goods so that their
particular remedies can be adopted and implemented. For example, social
programs may be justified on providing a more peaceful and harmonious
society as well as through assisting individuals who need assistance.

Scale

The concept of scale for policy problems is to some extent related to the
issue of public goods. The logic of scale is, however, that some problems
are inherently large scale and need to be addressed as a whole, or not at
all. Issues like building a dam or a bridge across a river are rather simple
 POLICY PROBLEMS 23

examples – half a dam or three-quarters of a bridge are useless. A more

interesting example may be the eradication of epidemic diseases such as
smallpox and polio. The World Health Organization has been attempting
to eradicate these diseases totally so that not only would no one become
ill with them, there would be no future need for immunizations.3 And
during the COVID-19 pandemic public health hoped to create herd
immunity within the population, with enough people either immunized
or recovered from the disease that it could not spread easily (Altmann,
Douek and Boynton, 2020).

Solving large-scale policy problems represents a challenge to political

systems that, like most, tend to function more incrementally. The norma-
tive argument for incrementalism, and for bounded rationality in general
(see Jones, 2001; also Chapter 3), is that humans tend to lack the capacity
to make comprehensive solutions to public problems because those
problems are complex and always changing. Therefore, making decisions
by “successive limited comparisons” can be argued to be a more rational
way to make policy (Lindblom, 1965) than more comprehensive inter-
ventions. Policies would be made by taking small steps, considering how
well the policy worked, and then adjusting the intervention. But that style
of making policy is simply not feasible for large-scale projects (Schulman,
1980), no matter how rational it may be in general.

As well as the normative issues in decision-making raised by large-scale

policy problems these problems also pose empirical problems within
governments. The multiple veto points (Tsebelis, 2000) that exist in most
governments make producing large-scale projects difficult. This problem
is more pronounced for presidential systems with numerous independent
actors but may be true even for parliamentary systems, and especially coa-
lition governments. The multiple actors involved in making decisions and
the multiple interests that must be served tend more toward governance
by the lowest common denominator and gradual adaptation rather than
making bold decisions about large projects (see Scharpf, 1988).4

The problem for large-scale projects may not be so much that decisions
cannot be made but rather that coherent decisions may be difficult. If
there are multiple actors involved, as there will be in almost any decision,
then movements away from the design intended by experts or a political
leader may be expected. We focus on design in this volume but design
can be easy in principle but is more difficult in political practice. Multiple
24 ADVANCED INTRODUCTION TO PUBLIC POLICY

interests attempt to add their favorite ideas to a project (“goldplating”

in defense contracts for example) or attempt to remove elements for
financial or policy reasons. The result may be more diffuse and gradual
adjustments to problems, even large-scale policy problems.

Although Schulman’s work on large-scale problems is important, we

should also consider the more general question of matching the scale of
the problem with the scale of the response, and the scale of the unit deliv-
ering the response. Some policy problems, for example, climate change,
have to be solved at an international level, while other problems such as
delivering water or sewerage services are best handled at a very local level.

Solubility

The concept of scale of problems is closely related to the question of

whether indeed a problem can be solved. If the problem is defined as
enabling people to drive across a river then building a bridge will solve
the problem. If, however, the problem is defined as providing effective
transportation for citizens then it may never be solved. The size of the
population may increase, requiring more facilities, and making automo-
biles less desirable as the focus for transportation policy. And technol-
ogies for transportation may also change, making some forms of mass
transit that could have been unaffordable at one time more feasible, or
tele-commuting reducing demand for transportation. And even lifestyles
may change, with citizens wanting more services and jobs near their
homes so they do not need to drive or take a bus.

Transportation is a policy area in which some issues may appear to be

solved, at least for a time, but other areas such as education, health and
social policy may have issues that can never really be solved. People will
always want to be healthier and happier, so that there will be continuing
demands for improving services in these areas. And lack of adequate
knowledge about causes of many social problems with weak technical
cores, or the risk involved in many economic and defense policies, means
that these policies are often best conceived as experiments, requiring con-
stant monitoring, and continuous attempts at improvement (see Nelson,
1977, 2011).
 POLICY PROBLEMS 25

The inability to solve most policy problems for once and for all, and the
continuing attempts to solve those problems, mean that most policy
spaces are very crowded. There are layers of attempts on the part of gov-
ernment to provide solutions to issues, sometimes building on previous
legislation and sometimes seeking to abolish all trace of the previous
legislation (see Mahoney and Thelen, 2010). But attempting to eliminate
the efforts of the past may be impossible, given that clients remember the
old programs and the organizations implementing the programs tend also
to remember the previous programs. While continuing efforts to solve
problems may represent, as Dr Johnson said, the triumph of hope over
experience they may also tend to produce cynicism and a lack of commit-
ment among clients and employees.

The good news, at least politically, is that policy replacement (see

Hogwood and Peters, 1983) is not attempting to make policy on a tabula
rasa, but rather is attempting to reform existing policy commitments of
the public sector. Once the problem has been addressed, and there are real
organizations and real clients, some of the political struggles over getting
on the active agenda of government have been resolved. Although the
existing organizations and clients may defend the status quo, they may
also favor change, seeing all too well the imperfections of the existing
programs.

If a problem is indeed solved, and even if it only ameliorates the con-

ditions it was designed to solve, it may then generate new problems
and new challenges for policymakers. In most areas of governing we do
not have adequate knowledge of the underlying dynamics in the policy
area to make as effective diagnoses of the problem as we would like. As
Richard Nelson (1977) pointed out with reference to attempts to resolve
the problems of the ghetto (racism, social deprivation, etc.), many if not
most policy interventions involve some degree of experimentation. And
likewise, we do not understand our policy instruments sufficiently well to
be able to intervene as effectively as we would like.

If I consider my transportation example above some of the impacts of

initial policy choices and the difficulties of making interventions become
apparent. The experience of road building in the United States and
elsewhere (Goodwin and Noland, 2003) has been that once roads are
built they attract new traffic and become outmoded almost before they
are completed. Likewise, the experience of building roads that facilitated
26 ADVANCED INTRODUCTION TO PUBLIC POLICY

travel into urban areas was that they also facilitated travel out of those
urban areas even more, thereby contributing to the decline of inner cities.
Therefore, both within the policy area and in other policy areas, making
policy may generate new and potentially more severe problems. As Aaron
Wildavsky argued (1980 [2018]), policy often is its own cause.

Complexity

Some policy issues are simply more complex than are others (Duit and
Galaz, 2008). And the idea of complexity itself needs to be considered
carefully5 in policy terms (Table 2.1), because it has at least two dimen-
sions: Technical and Political (see Bovens, ‘t Hart and Peters, 2000).
By technical complexity I mean that the underlying causal processes
in the problem are not understood fully, or they involve a number of
interactions of individual social and technical factors. Crime might be
an example of a complex problem, given that it is difficult to determine
exactly what causes people to become criminals. Some aspects of climate
policy would also be technically complex, even more so because the
interactions among the variables may be non-linear with small changes in
some factors triggering much larger changes.

Political complexity means that there are multiple and conflicting inter-
ests involved in the policy domain. These interests may also have funda-
mentally different ideas about causation or about what would be a good
outcome of a policy process. This political complexity has been visible in
economic policy, for example, when environmentalists clash with eco-
nomic developers over what the goals of the policy should be. Thus, even
if there is basic agreement over the nature of the policy and the causal pro-
cesses involved there can still be very strong political disagreement about

Table 2.1 Types of complexity

Technical Complexity

High Low

Political High Environmental Policy Education

Complexity
Low Science Policy Pensions
 POLICY PROBLEMS 27

what to do. Also, in economic policy responses to the post-2008 economic

crisis there have been fundamental differences between advocates of
austerity and those pursuing a more Keynesian approach to economic
stimulation (Krugman, 2014). And when that political disagreement is
coupled with technical complexity and disagreement the processes of
making policy become all the more difficult. The conflict between those
who wanted to open the economy during the COVID-19 pandemic and
those who wanted to continue isolation represented different views of
policy priorities, different understandings of the medical facts, and differ-
ent ideologies – a very complex political problem.

We can see the interaction of political and technical complexity by exam-

ining Table 2.1. In this table levels of complexity are classified as simply
high and low. While this may simplify the underlying dimensions it is
still useful for understanding the policymaking challenges posed by com-
plexity. The simplest possible pattern for policymakers would be to have
problems that are both relatively simple technically and politically. These
tend to occur in policy areas in which governments have been active for
some time and many of the political conflicts have been ameliorated, if
not solved. In contrast, making policies in domains in which there is both
technical and political complexity is extremely challenging, and unlikely
to produce highly effective policies.

The other two cells of the table also represent challenges of policymaking.
When the technical issues underlying a policy problem are relatively
simple and there is still political complexity reaching agreement on policy
may be easier than when the technical issues are in doubt. When there is
agreement on the logic of cause and effect in a policy area, then the poli-
tics in some ways may be more intense, given that there is basic agreement
on the nature of the policy. Relatively high levels of political agreement
on policies, associated with technical complexity, may be able to produce
a more experimental, or evidence-based, style of making policy. While
experimentation is an important means of addressing the unknown in
public policy, citizens may not appreciate being considered guinea pigs.

That said, some policy large experiments have been attempted, and have
had an impact on public policy. The government of Finland, for example,
wanted to know what impact having a guaranteed income would have on
the economic behavior of citizens. They gave a sample of the population
a monthly income and then observed how their behavior differed from
28 ADVANCED INTRODUCTION TO PUBLIC POLICY

those without the guarantee. They found that those citizens receiving the
benefit did not change their economic behavior significantly, but they
were happier (Lefebvre, 2019; New York Times, 2019).

Governing and making policy always involves coping with complexity,

but some problems and some policies involve more complexity. The table
here points to some aspects of that complexity, but in the extreme govern-
ments face so-called wicked problems. These are problems that are suffi-
ciently complex and unstructured that making policy choices is extremely
difficult. These problems are sufficiently important for emerging policy-
making and governance that I will discuss them separately (see below) as
a significant mechanism for understanding contemporary policymaking.

Certainty and risk

Some policy problems are very predictable and involve little inherent risk.
School officials can know with some degree of certainty that if a child is
born he or she will need to begin school in five or six years, so the school
buildings, teachers and chalk had best be ready. Migration in and out of
the district may affect the final total of pupils slightly, but there is enough
certainty to plan effectively.6 At the other end of the life cycle, pension
managers know with substantial certainty how many people will become
eligible for pensions in any given year and can plan accordingly for paying
those pensions.

Many, if not most, policy problems do not have that degree of certainty,
so policymakers must cope with risk and uncertainty (Dror, 1986).7
Uncertainty is perhaps clearest in international policy areas in which one
set of actors is developing policies knowing that in other countries other
policymakers are making contrary decisions. Uncertainty is also apparent
in policy areas that are heavily influenced by natural events. For example,
the Army Corps of Engineers in the United States builds flood control
projects based on estimates of the largest floods that would probably
occur every 50 years or 100 years (Army Corps of Engineers, 1996). But
sometimes their projects have to contend with the 500-year flood, and
may fail – the flooding in New Orleans after Hurricane Katrina is the
 POLICY PROBLEMS 29

obvious example. The Corps did its job as it is mandated to, but it had to
contend with uncertainty and events that were not parts of the planning.

The characteristic of complexity discussed above is also connected to the

presence of risk in a policy problem. Charles Perrow (1984) has famously
discussed “normal accidents” as a part of contemporary society, and con-
temporary policymaking. His argument was that as we depend upon more
and more complex systems such as nuclear power and air traffic control
we should expect accidents. Therefore, policymakers need to build these
risks into their calculations about creating and regulating those complex
systems. And citizens may have to be educated about the possibilities of
these accidents so they can make their own calculations, and so that they
do not expect miracles from technologies or their governments. But peri-
odic epidemics and pandemics can also be “normal accidents” that will
occur, although we do not know when.

The presence of risk in many policy situations introduces the need to

include risk in making choices. One means is to attempt to eliminate risk
entirely through mechanisms such as the precautionary principle used by
the European Union for issues such as genetically modified organisms
(GMOs) (Majone, 2002a). While that approach appears to eliminate risk,
it does not because it fails to take into account adequately the possibility
that there will be no (or limited) negative effects of GMOs and hence
there are potentially large opportunity costs being imposed on European
citizens and farmers.

We should think of the decisions being made on GMOs as involving risk,

that is, the probability of there being harm, and also expected costs and
benefits. In this example we will suppose that the risk of harm from intro-
ducing a specific GMO is, say, 10 percent, but the risk of harm produced
would be very substantial if indeed the crop is dangerous. On the other
hand, if the crop is not dangerous there is a potentially substantial benefit
from the higher productivity of the crop. We can put these outcomes
together in Table 2.2, and calculate the economic outcomes of the choice.
This monetary figure does not, of course, take into account environmen-
tal consequences that may be difficult to calculate, nor potential benefits
such as reducing poverty in poorer countries, that are also difficult to put
on the measuring rod of money (see Otsuki, Wilson and Sawedeh, 2001).
Still this rather utilitarian analysis provides a means of beginning to think
about the choice that must be made.
30 ADVANCED INTRODUCTION TO PUBLIC POLICY

Risk is an objective quality of policy settings, but the perception of risk

may be as important or more important than the objective conditions.
There is an extensive literature pointing to the misperceptions of risk
within society (Slovic, 2000; Weber and Stern, 2011). Given these misper-
ceptions, governments may invest heavily in safety in some areas and not
in others, or invest more heavily in some diseases than in others even
though objectively the relative risk of death or injury does warrant that
distribution of funds.8 We as analysts may be able to say rather facilely
that the perceptions of risk are irrelevant for making real policy, but in the
political world dismissing public opinions is not so easy.

Tragic choices

Governments must make a number of extremely difficult choices. The

most difficult of these policy problems have been classified as “tragic
choices” (Calabrese and Bobbitt, 1978; Brown, 2007), meaning that
making choices to benefit one group in the society will inevitably produce
deprivations, and often quite severe deprivations, for other members
of the society. In its original development, the concept of tragic choices
was used to describe decisions being made in the allocation of scarce
life-saving technologies, when giving one person access to that technol-
ogy inevitably meant that someone else would not receive the treatment
and would die. The allocation of ventilators in the COVID-19 pandemic
is a clear example of these choices (Truog, Mitchell and Daley, 2020).

To some extent all choices being made in the public sector are tragic,
given that any decision to benefit one group involves deprivations to
others, but the idea of tragic stresses that some policy decisions may
mean that some people die. So, deciding how to allocate livers or hearts
for transplants means that some people will survive and some will not.
But deciding to allocate money to highways or to other purposes may
also mean that some people will be saved and others may not. Are the
rules for making these allocations fair? And are there a range of cultural
definitions of fairness that differ across countries and that may produce
different outcomes for citizens? Thus, policy problems involving tragic
choices require more careful attention to normative standards than do
other sorts of policies.
 POLICY PROBLEMS 31

Monetization

Finally, some policy problems are fundamentally about money, and

can be expressed in terms of the need to redistribute money. Many of
the common policy programs depend upon monetizing the problems
involved. Subsidies to farmers, loans to university students, pensions of
the elderly and flood insurance for homeowners all assume that money
is sufficient to address, and perhaps even solve, the policy problem.
And generally this assumption of the underlying monetary nature of the
problem is correct and money is sufficient.

If a problem can be addressed with money then government should con-

sider themselves fortunate. Given that governments have large budgets
and have access to printing presses, they can generally marshal financial
resources to address a problem.9 If, however, the problem is based on
the deprivation of status, respect and on injustices then attempting to
redress those grievances through money would be ineffective and even
insulting. For example, some groups have advocated reparations for
African-Americans whose ancestors were brought to the United States as
slaves (Posner and Vermeule, 2003). While that money might be welcome,
it could not adequately address issues of race and deprivation that persist.

When we begin to discuss policy instruments, it becomes apparent that

relatively few instruments are clearly connected to altering problems
defined by status or respect. Nodality, or information, is the most obvious
mechanism for approaching non-monetary problems. But this category of
instrument tends to be less directly effective in producing change than do
other mechanisms such as those utilizing money (treasure) or authority
(law). These nodality-based instruments have become more popular
as government resources wane and citizens become more resistant to
authority (Thaler and Sunstein, 2008), but they generally lack the effec-
tiveness of more intrusive forms of intervention.
32 ADVANCED INTRODUCTION TO PUBLIC POLICY

Summary

This list of characteristics of policy problems is useful for understanding

the nature of the tasks confronting the public sector. Each of the cate-
gories is useful for understanding the problems, but there is as yet no
weighting of the relevance of each, nor how to link them together to create
more complex understandings of the challenges to policymaking. With
these characteristics, however, we can demonstrate the internal variations
of a functional policy area such as health (see Table 2.2).

These problems, and the understanding of the challenges they present,

will serve as the background for other aspects of the process of policy
design. If designers understand the nature of the challenges they face
they will be better able to understand how to create programs that have
a greater probability of success. Further, differences among the nature of
problems will affect the policy process itself. For example, more politically
complex problems (almost by definition) tend to involve a wider array
of social interests, while technically complex problems tend to exclude
interests unless they are members of specific epistemic communities.
The real political fun begins for problems that are both technically and
politically complex.

Table 2.2 Choices with risk

 POLICY PROBLEMS 33

Unstructured and wicked problems

The discussion above has been based on a fundamental assumption that

the policy problems we are contending with can be defined readily and
have an identifiable structure. Such problems are the foundation of most
policy analysis, given that they can be readily understood and offer some
hope of being resolved. Also, policy analysts and policymakers tend to
assume that problems are adequately structured and understood so that
they can proceed to attempt to resolve them (Hisschemöller and Hoppe,
1995).

The problem for policymakers, as well as for citizens, is that many

problems do not come neatly structured. Further, problem structuring is
a political process in which individuals with different conceptions of the
problem attempt to create a conceptual structure of the problem that can
be used to make policy, if not necessarily solve the problem (Dery, 2000).
Without such a structure, or in the terms of this chapter some under-
standing of the characteristics of the problem, policymakers will not be
able to address the issues involved effectively.

We will discuss the political process involved in framing and structuring

problems in Chapter 3, but here it is important to consider the underlying
nature of unstructured problems, and particularly an extreme version of
these – wicked problems. Understanding these difficult problems appears
to be becoming increasingly significant for governments, as the real world
is forcing more of these problems – climate change, obesity, substance
abuse – into the public sector, and citizens are expecting some form of
public action. These wicked problems are often the most dangerous issues
facing citizens and societies as a whole and hence demand some form of
public intervention. That intervention may at times be only symbolic as
governments have no real answers, but there must still be some response.

Rittel and Webber (1973) coined the concept “wicked problems” to

describe very difficult problems that were facing the public sector.
Although originally expressed in terms of planning theory, this concept
appears extremely relevant for more general studies of public policy.
Rittel and Webber discussed the concept of wicked problems in relatively
34 ADVANCED INTRODUCTION TO PUBLIC POLICY

abstract terms, but phrased in somewhat more operational terms there

are several criteria that can be used to characterize wicked problems:

1. Wicked problems are difficult to define. It is not easy to say just what
the problem is.
2. The problems are multi-causal and have many interconnections.
3. Therefore, wicked problems are often unstable, with small changes in
one possible cause producing large-scale effects.
4. These problems have no clear solution, and perhaps not even a set of
possible solutions.
5. Because the solutions are unclear, any intervention may have unfore-
seen consequences.
6. Wicked problems involve multiple actors and are socially complex.

This characterization of policy problems appears relatively similar to

some of the aspects of complexity already mentioned, but tends to
emphasize the difficulties in even defining the problem in operational
terms. Also, recent research on complexity in public policy (Duit and
Galaz, 2008; see also Termeer, Dewulf and Van Lieshout, 2010) has been
extending the consideration of complexity in ways somewhat similar
to the conception of wicked problems. That said, Rittel and Webber
appeared more interested in understanding the nature of the underlying
problems, while the complexity scholars appear more concerned about
the possibilities of solutions.

If wicked problems were not difficult enough, some scholars have argued
for the emergence of “super wicked” problems. These problems have all
the properties of wicked problems, but in addition have four additional
characteristics that confound policymaking even more. These four char-
acteristics are:

1. Time is running out.

2. There is no central authority, or only a weak central authority, to
manage the problem.
3. The same actors causing the problem seem to solve it.
4. The future is discounted radically so that contemporary solutions
become less valuable.

The concern with “super wicked” problems is obviously closely connected

to contemporary questions such as climate change and resource depletion
 POLICY PROBLEMS 35

(see Levin et al., 2010), but may also apply to other issues, especially those
that have a strong international component.

At the time that Rittel and Webber were writing, the major empirical
referents for wicked problems were urban and social problems, while
in the twenty-first century the major referents are global environmental
and economic issues. These problems produce severe disadvantages for
society (even when compared with “normal” wicked problems), and
because of their global nature also lack a sovereign which can make
decisions that may be able to solve, or at least ameliorate, the problem.
But even with the importance of the international environmental issues,
the social problems with wicked characteristics, such as inequality and
inclusion, remain important.

Some common issues in defining problems

For all types of policy problems there are some common questions that
must be considered in the design of responses (see Weimer, 1993). In
addition to attempting to understand causation and to frame the problem
in terms of causation (see Chapter 3), the problem must be understood in
terms of the manipulable variables that are contained within it. No matter
if the problem is large scale or small scale, or simple or complex, anyone
thinking of intervening must understand what variables can be used to
produce change and which cannot. And designers will always be looking
for those variables that are the easiest, and least costly, to manipulate.

Another important, and forward-looking, element of problem definition

is to consider what the policy area is meant to look like after government
has acted? These goals will be different for each policy problem but there
needs to be some sense of what the intervention will produce. If there is
a problem, then what will be required to eliminate, or at least ameliorate,
the problem? As noted at the beginning of the chapter, the common sense
notion of policy problems may initiate the process and some common
sense of remedy also motivates the process. But that common sense may
not be adequate for producing a more enduring design for addressing the
problem.
36 ADVANCED INTRODUCTION TO PUBLIC POLICY

Coordination as a policy problem

Finally, we have been tending to consider the nature of policy problems

one by one, but as mentioned in Chapter 1, coordination among the
numerous policies within the public sector represents a significant policy
problem in itself. Although the coordination problem tends to arise after
governments have acted, we can conceptualize this as a fundamental
policy problem. As already noted, most policies depend at least in part
on other policies to be effective. This interdependence is perhaps most
evident in social policy areas (Challis, 1988). Education cannot be effec-
tive if students are hungry, or come from troubled homes. And health
policies cannot be effective if the population is poorly housed and poorly
fed.

As a policy problem coordination has some of the same characteristics

used to describe other policy problems, For example, coordination has
some elements of a large-scale problem, given that reaching some agree-
ment among the actors involved is more effective if all relevant actors
are involved. Further, coordination tends to have a great deal of political
complexity. In this case, the political actors are primarily organizations
within government itself, rather than external political actors. Each
of these organizations is attempting to use the coordination situation
as a mechanism for enhancing its own success, as well as to advance
a particular policy agenda.10 Finally, there are interesting questions about
what is the purpose of the coordination and how many actors need to be
involved in order for the coordination to be effective.

This recognition of coordination as a policy problem in itself (see also

Peters, 2015) helps to emphasize the interconnections existing within the
public sector as it acts to make policy. We should attempt to understand
the nature of the individual public problems and provide some (hopeful)
solution for them, but that may only be the start. That action cannot
be undertaken effectively without some understanding of the already
crowded policy space of the intervention.

In addition to the horizontal coordination with other programs, there will

also be vertical coordination within the multi-level governance operating
in any government system. To some extent we must consider policies one
 POLICY PROBLEMS 37

by one, but we must not lose sight of the multiple connections with other
policies and actors.

Summary

Policymaking is directed at solving problems that confront the society

and economy. Ordinary citizens can identify most of the problems facing
them, although they may not recognize all the complexities involved.
These common sense definitions of problems are useful but they may
not be capable of moving the solution of the problems very far forward.
Likewise government organizations tend to think of policy problems in
terms of their own interests and their policy priorities. Resolving these
policy problems will require more thorough analysis and a good deal of
politics, as well as a clear analytic conception of the problems themselves.

Although both academic and popular discourse tend to link problems

with particular substantive policy areas, or with particular policy instru-
ments, the reality of policy problems is more complex. This chapter
has raised a number of points about policy problems and how their
underlying features can affect the types of policy choices made, as well
as the success of those policy choices. This discussion of policy problems
has been rather analytic, but the next chapter will examine more about
the politics of policy problems. This chapter will focus on the framing
of policy problems and on how those issues become parts of the public
agenda. Our rather academic perspective on the policy problems pre-
sented in this chapter will have to be tempered by the political process of
framing and defining public sector agendas.

Notes
1. Historically telephones were also considered natural monopolies but tech-
nological change has made the market for cell telephony competitive, if still
to some extent regulated to prevent collusion in a market in which the entry
costs are very high.
2. A good recent example would be the decision by the US House of
Representatives to pass a farm bill providing billions of dollars in subsidies to
38 ADVANCED INTRODUCTION TO PUBLIC POLICY

farmers, most of which goes to very affluent farmers and to “agribusiness”.

The same bill terminated the food stamp program, the benefits of which go
primarily to the working poor.
3. This apparently has been achieved with smallpox (Fenner et al., 1999), and
children are now no longer routinely vaccinated.
4. There are some significant exceptions to this rule. More authoritarian
regimes certainly can produce large-scale governance, as the success of
some of China’s many development projects demonstrates. Also, crisis, and
large-scale failure, may induce investment in large governance projects.
5. And complexity is different from complicated. A complicated problem may
have a number of moving parts but the interactions among those parts may
be readily understandable.
6. This certainty depends in part on the policymakers having basic information
about population movements. However, in many developing there may
be little accurate information on total births and deaths, much less on the
residence of the children being born, so that planning becomes an even more
difficult task (UNICEF, 2005).
7. See Chapter 9 on risk-benefit analysis.
8. Approximately 20 percent more people die of colorectal cancer in the United
States each year than they do from breast cancer, but funding for breast
cancer research is more than twice as high.
9. There are, of course, some real constraints on the capacity of governments to
tax and to print money. The experiences of countries such as Italy, Spain and
Greece after the fiscal crisis of 2008 demonstrate that those limits on public
finance are very real.
10. In this way coordination (and several other policy problems) resemble
Graham Allison’s concept (Allison and Zelikow, 1999) of bureaucratic poli-
tics, in which a decision situation (no matter how much of a genuine crisis it
may be) becomes a locus for pursuing organizational interests.
PART I

Making decisions about policy

3 Models of policymaking

The logic of designing public policies is to make decisions about policies

that then are translated into action. Therefore, I need to discuss the range
of models of decision-making for public policy and assess their utility
for understanding how the public sector, along with its allies, function to
make those decisions. The literature on public policy contains a number
of theories, models and frameworks for policymaking (see Petridou,
2014). Although making different claims about generalization and expla-
nation, each of these schemes argues that their particular perspective pro-
vides a useful, if not superior, intellectual foundation for understanding
and explaining how decisions about policy are made.

As well as having aspirations of being explanatory, most of these models

are also at least implicitly normative. That is, they are arguing not only
this is how decisions are made but also how decisions should be made.
The normative stance of several of these models of decision-making is
not explicit, but there is some sense even in those cases that if the logic
of the approach were followed then governments would be able to make
better decisions. In none of these models of decision-making can there be
a guarantee of good decisions, given that the information and assump-
tions going into the process may be faulty. The proponents of the models
argue, however, that at least the process itself, and the assumptions about
how to reach good policy decisions, will not be an impediment to good
policy.

Although each of the models of policymaking discussed here has its own
distinctive questions and answers, there appear to be two broad sets of
approaches. One is concerned with the ability of policymaking to be
rational in any meaningful sense of the term. The possibility of synoptic
rationality is one of the oldest questions in policy studies (Lerner and
Lasswell, 1951, pp. 4–6), and continues to motivate discussions in the
field, and most models of policymaking make less demanding assump-

40
 MODELS OF POLICYMAKING 41

tions about the process. That said, some form of utilitarian and rational
perspective persists, and even dominates, in evaluative models of policy
(see Chapter 8) if nowhere else.

Associated with the question of the possibilities of rationality in

decision-making is whether decisions are made best in a comprehensively
rational manner, or whether they should be “good enough” for the short
term, with decision-making being conceptualized as more continuous.
As well as the capacity to act rationally, this issue is concerned with
the availability of adequate information to make those comprehensive,
long-term decisions. Are decision-makers capable of marshaling enough
information to be able to predict the future state of the policy domain, and
to understand all the potential policy options?

The second set of questions about policymaking addresses the extent

and the manner in which politics is involved in making policy decisions.
As we are dealing with public policy then the public sector is inherently
one component of the decision-making, but the question is how are
government and politics involved. This question includes the extent to
which the formal institutions of government dominate policymaking,
and conversely the extent to which interest groups and other social actors
influence the process directly. Further, we need to consider the extent to
which expertise rather than politics is dominant in the decision-making.

Although it is at a different level of generality, the policymaking literature

also raises the question of the extent to which the process itself shapes
the policy outcomes produced. Much of the political science literature on
public policy emphasizes the process by which policy is made, and tends
to emphasize the position and role of particular political institutions in
that process (Hupe, 2007). Other models of policymaking are less con-
cerned with process and institutions but rather focus on decisions per se,
assuming that the same logic would apply regardless of the nature of the
institutions involved.

This difference in emphasis on institutions and processes does not mean

that the fundamental emphases cannot be combined. For example, we
may consider how rational actor approaches function within institutions
by considering institutions as collections of veto points and veto players
through which decisions must processed in order to make a final decision
(Tsebelis, 2000; but see Ganghof, 2003). We would expect that institutions
42 ADVANCED INTRODUCTION TO PUBLIC POLICY

with more decision points would make any comprehensive version of

rationality more difficult, with the boundedness of that rationality (see
below) becoming more evident.1 For example, presidential policymaking
systems such as the United States with multiple actors and multiple veto
players are characterized by difficulties in making clearly defined policy
design.

Although this chapter is entitled “Models of Policymaking” other impor-

tant models are located in other chapters. Most notably, the discussion
of the Punctuated Equilibrium Model, one of the more commonly used
models in political science, is located in the chapter on agendas. This
splitting of major approaches to policy reflects the difficulties of writing
about public policies, both in a practical and an academic sense. Although
to some degree issues about the individual components of policy and the
policy process can be treated separately, in other ways they are intimately
connected with the remainder of the policymaking.

Approaches to decision-making and their relationship

to policymaking

When we discuss policymaking and policy design we are ultimately

talking about decision-making. Having said that, we need to place
decision-making at the center of a study of public policy and we must now
proceed to discuss how we understand the processes through which deci-
sions are made, and the manner in which individual decisions are linked
across the governing process. The models we will discuss come primarily
from political science and public administration, and are inherently about
making policy decisions. Saying we need to focus on decisions, however,
brings forward a wide range of models and frameworks that do attempt to
help the scholar and the practitioner understand decisions.
 MODELS OF POLICYMAKING 43

Rationality

For contemporary political science the usual manner in which to begin

considering the process of making decisions about governing would be to
think of these as a rational process in which the actors involved attempted
to maximize their personal or organizational utility. That rational per-
spective can be developed either in a rather simple descriptive sense or
in a more formal manner. In either version of rationality there are a set
of assumptions that are fundamental to the approach and which at once
strengthen the analysis produced and place constraints on the range of
decisions that may be taken.

For our purposes the first and perhaps most difficult of these assumptions
is that the actors making decisions have a known order of preferences, and
these preferences are transitive. That is, if a decision-maker prefers A to
B and prefers B to C then he or she will prefer A to C. This appears to be
a rather simple condition for ordering preferences but in complex politi-
cal and strategic governance situations this condition may not necessarily
hold. The potential lack of transitivity is in part because governance
decisions involve multiple dimensions so that having to choose A or B
or C involves actually making choices about a number of factors – goals,
instruments, budgets – that may not all vary together in the preferences of
a decision-maker. And even if individual preferences are transitive once
more than one individual becomes involved in the decision, problems of
aggregating the preferences emerge (Dryzek and List, 2003).

A second assumption about rationality in public policymaking is that all

the actors involved will have the same array of preferences, or at least that
their preferences will be based on roughly similar foundations. Beyond
the simple idea that all actors are utility maximizers, that assumption
may be difficult to maintain. Policymaking involves a range of actors, all
of whom may want different things out of governing, and they may not
even be thinking about the same factors. The leader of a bureaucratic
agency may be thinking about maximizing the prestige or budget of the
organization while the political leader may be thinking about re-election,
controlling public expenditure, or even the best way of serving the public.
These conflicting rationalities then need to be integrated in some ways
through the processes of governing (see below). Further, the preferences
of the actors involved should be transitive. In policymaking that assump-
44 ADVANCED INTRODUCTION TO PUBLIC POLICY

tion is often violated because of ideological or strategic reasons. If that

assumption is violated then making decisions becomes more difficult,
especially when multiple individuals with seemingly unstable preferences
are involved.

A third assumption about rationality is that the actors involved have

perfect information so that they can make rational choices among the
available alternatives. And indeed there is an assumption that the actors
are aware of all the options so that they can make the appropriate choices
among them. In the real world of policymaking there is generally very
incomplete information and incomplete understanding of the options (or
even of the nature of the problems themselves, see Chapter 2). Further,
policymakers may attempt to limit the number of options considered in
order to reach a decision more quickly (Peters, 2018). Thus, given these
problems in information, among others, the possibilities for making fully
rational decisions is limited, even if the actors involved are attempting to
be as comprehensive as possible in their analysis.

Fourth,, this model of decision-making would assume that the outcomes

of a process can be assessed on a common utilitarian scale of benefits
and costs. In the world of politics and governing, however, process may
count for as much as substance, and image may trump reality. In the
terms of March and Olsen (1989) the “logic of appropriateness” may be
at least as important as the “logic of consequentiality” in understanding
how to evaluate decisions made in governance. That is, decisions should
be evaluated in terms of their normative content as well as their costs
and benefits. If this is the case then decision-making about public policy
may involve a set of incommensurable dimensions of evaluation which
renders making a reasonable decision difficult.

Finally, a rational model of decision-making for governance would

assume that the consequences of the choices being made are predictable
and reasonably stable. Unfortunately for policymaking and governance
making predictions about outcomes is difficult. There is, for example,
a large and interesting literature on the unintended consequences of deci-
sions made in governance, as well as on the massive failures of programs
and policies (Sieber, 1980; Bovens and ‘t Hart, 1996). And governments
may be forced into making choices that they know will not solve the
problem at hand, simply because they have to do something (McConnell,
2019). Thus, even if we were able to provide unambiguous rankings of
 MODELS OF POLICYMAKING 45

preferences for outcomes the ability to link outcomes with specific gov-
ernance choices may be limited.

While the assumptions about rationality discussed above give some cause
for concern about this approach to understanding decision-making for
governance, we can still gain some useful insights by acting as if the
assumptions could be sustained and then looking at the ways in which
decisions are being made. The simplest of these ways of understanding
choices are relatively simple utilitarian models of choice and evaluation.
For example, cost-benefit analysis and risk-benefit analysis assume that
the consequences of public policies can be assessed on a common meas-
uring rod of money and that more is better.

It is important to notice in these discussions of decision-making from

a rational perspective that the emphasis tends to be on single decisions
rather than on the complex chains of decisions that are inherent in
governance. Likewise, the emphasis is on a decision-maker, or a limited
number of decision-makers who make bargains. These bargains tend to
be conceptualized as one time deals based on a particular set of prefer-
ences. Governance, on the other hand, is a continuous process so that
reducing transaction costs among the participants may be as important
for the success or failure of the policy process as the rationality of any
particular choice.

One of the most familiar examples of rationality in the literature on

governance is Graham Allison’s study of decision-making in the Cuban
Missile Crisis (Allison and Zelikow, 1999). He examined the behavior
of American and Soviet decision-makers during the crisis using three
alternative frames, one of which was comprehensive rationality. The
version of rationality used by Allison was, however, rather informal and
the conditions that would underpin that approach to governing were
not fully specified. This crisis also contained multiple individual choices,
involving tacit bargaining between the two sides. Similarly, Yehezkel Dror
(1986) examined the rationality of decision-making under conditions of
adversity in governance, and attempted to understand how governments
could make rational choices when faced with a complex, and potentially
hostile, international environment.

A more detailed set of models about the application of rationality to gov-

ernance decision-making can be drawn from the literature on decision
46 ADVANCED INTRODUCTION TO PUBLIC POLICY

science, seeking to quantify the pay-offs of making choices under risk.

For example, what should a decision-maker do when faced with an event
(the landfall of a hurricane) that is to some extent uncertain? There are
costs and benefits associated with any action (evacuation or sit tight) and
therefore we can consider the likely costs of any action given different sets
of probabilities. The typical cost-benefit analysis associated with policy
analysis produces the same type of analysis albeit generally without con-
sideration of the probabilities of events.

More formal models of rational decision-making such as game theory

contain most of the formal assumptions mentioned above. The indi-
viduals are assumed to have a set of preferences to which they can rank
and can assign ordinal if not cardinal values. Game theoretic models are
largely based on the interaction of two or more actors who may have to
make choices in the absence of knowledge of the choices of the other
participant(s) in the game. The complex and extended discussions among
participants in the policy process do not appear to fit well into these rather
restrictive models (but see McCain, 2009).

For policymaking these game theoretic models are not particularly

valuable because most assume actors with conflicting preferences and
the absence of direct bargaining among the participants. The world of
public policymaking tends to involve multiple actors with some com-
peting and some compatible goals engaged in long-term interactions
with one another over programmatic choices. Where the game theoretic
approaches do help for understanding policy is when they are considered
over an extended period of time and also involve learning (Axelrod, 2007).
In these cases the actors involved learn about each other’s preferences and
strategies and may be able to produce more cooperative outcomes.

Finally, even if some of the formal conditions for rationality were to be

met, a fundamental political question remains about the model. Any
strictly rationalist model would tend to depoliticize policymaking, some-
thing that may also be virtually impossible in as political an environment
as designing and selecting public policies (Elster, 1991). Some functioning
policymaking systems such as the Planning, Programming and Budgeting
System (PPBS) have attempted to inject greater rationality into the
process, but that style of making budgets was criticized because it did not
fit well in a political environment.
 MODELS OF POLICYMAKING 47

However, even if rationality is unattainable in reality, does it provide

a standard against which to compare the processes and outcomes of
policymaking. As studies of implementation (see Chapter 5) have focused
on the extent to which outcomes differ from perfect implementation, so
too can we learn about policymaking by positing rationality and then
seeing where the deviations from that assumption occur. For example,
the analytic narratives approach (Bates et al., 1988), used primarily in
comparative politics, combines a rational actor perspective with case
study methodologies to understand how choices are made. The same sort
of approach could easily be applied to public policy to understand how
choices are being made (Shanahan, Jones and McBeth, 2018).

In summary, while we have an ingrained expectation, or at least hope,

for rationality in decision-making concerning public policies, producing
those rational decisions is difficult. Producing rational action is especially
difficult in governance when there are multiple sets of preferences, actors
may not even agree on the definition of the decision-situation, and where
there is limited and often biased information. Given these barriers to
comprehensive rationality, students of governance need to consider
alternative models if we are to have a more realistic understanding of the
manner in which decisions are made in the public sector, and through the
interactions of actors in and outside that public sector.

Bounded rationality

All the problems with a comprehensively rational perspective on

decision-making for governance lead rather naturally to consideration
of bounded rationality, and means of making decisions that reflect more
accurately the complex realities of governance. Indeed, some of the
assumptions behind bounded rationality represent marked contrasts to
those of comprehensive rationality in making decisions. These assump-
tions are, of course, also subject to criticism and whereas comprehen-
sively rational models may be too well specified and produce excessively
neat answers to governance problems, bounded rationality may provide
answers that are too indeterminate and not adequately precise.
48 ADVANCED INTRODUCTION TO PUBLIC POLICY

Bounded rationality is usually associated with Herbert Simon and his

studies of public administration (Simon, 1947). Simon argues that most
of the assumptions about decision-making by “economic man” – the
perfectly rational decision-maker – are not sustainable in reality. As
already noted, in the public sector (or indeed in much of the private
sector) preferences may not be transitive, and individuals may not really
know their own preference orderings until confronted with specific
decision-situations. Preferences in making public policy often emerge
based on the feasibility of outcomes rather than more fundamental ideas
about the desirability of policies.

Likewise, the information available to decision-makers is far from perfect

and indeed if individuals waited for perfect information, or a full array of
possible choices in the decision process, Simon argued that they might
never make a choice. Given that policymaking is inherently prospective,
and that in many policy areas there is not the clear linkage between one or
a limited number of causes and the effect being considered, then assum-
ing that one can make comprehensively rational decisions appears more
than a little questionable.

The alternative that Simon (1947) proposes to the seemingly severe

requirements for economic man is “administrative man”. This
decision-maker wants to act rationally but also understands that any
decisions would inevitably be made within a set of constraints on
information and analysis. Therefore the decision-maker is well advised
to make smaller, less comprehensive decisions rather than attempt to
synoptically rationalize. These decisions should be good enough for the
time being – they “satisfice” rather than maximize (see Brown, 2004). The
problem may not be solved but with good fortune it may be ameliorated.

The logic of satisficing is also that the decision is good enough for the time
being, and can be revised continually as conditions change, new informa-
tion becomes available, or new actors become involved in the governance
process. This version of rationality may therefore be more suitable
for governance than is the comprehensive rationality described above.
Bounded rationality tends to recognize clearly that decision-making is
being conducted in a much more complex world than that assumed by
many of the rational models of decision-making. Further, it also assumes
a more continuous process of making policies in which there can be some
learning and some adjustment.
 MODELS OF POLICYMAKING 49

In addition to satisficing as an acceptable, and again desirable, goal for the

outcomes of a governance decision-making process, bounded rationality
also tends to emphasize routines and standard operating procedures as
means of institutionalizing these formats for making decisions. The role
of routines within policymaking processes can be seen perhaps better
in March and Olsen (1989) on institutions than in the original work
by Simon. March and Olsen argue that the logic of appropriateness
within organizations involved in governance is defined by a series of
factors such as symbols and routines. These routines tend to simplify the
decision-making process by establishing bounds on what will be con-
sidered, when they will be considered, and other factors associated with
making the decision. Rather than writing on a tabula rasa, as might be
understood from the rational perspective, decision-makers in governance
are writing on a slate that is already crowded with policies and organiza-
tions, and within which only relatively minor adjustments are likely to be
successful.

Although not a produce of the same intellectual tradition of bounded

rationality, the logic of incrementalism (Hayes, 2017) has many of the
same features as does bounded rationality. Perhaps most importantly
there is the rejection of the possibility, or even desirability, of synoptic
decision-making in governance, and with that rejection the acceptance of
continuing processes of “partisan mutual adjustment” (Lindblom, 1965)
to make decisions. This model of decision-making therefore explicitly
accepts that there are multiple perspectives on policy and governance held
by multiple actors (the partisan element) and that decision-making in this
context is never finite but continuous.

Advocates of bounded rationality and of incrementalism will argue that

this form of decision-making may in fact be more rational than the syn-
optic attempts at rationality described above. They argue this position
on the basis of two elements affecting decision-making. The first is that
comprehensive decision-making may uncover a solution for a governance
problem, but it may also fail. When these decisions fail, they tend to fail in
a large, visible and expensive manner which is bad not only in substantive
policy terms but also undesirable from a political perspective. If politics
is indeed about blame avoidance (Hood, 2011) then making large-scale
comprehensive decisions may not be the best approach.2
50 ADVANCED INTRODUCTION TO PUBLIC POLICY

The other way in which incremental solutions may be more rational

is that they can minimize decision-making costs. Even attempting to
make decisions in a comprehensive manner requires marshaling large
quantities of evidence and analysis, and even then it will almost inev-
itably be inadequate for the complexity of most governance decisions.
Therefore, by making a series of smaller decisions, assessing the impacts
of those decisions, learning, and then adjusting to the evidence, better
decisions can be made with less investment in the decision-making.
The capacity of incremental decision-making to reduce decision costs
has been demonstrated most clearly in budgeting, when large numbers
of choices involving immense amounts of money are often reduced to
simple rules of thumb (Davis, Dempster and Wildavsky, 1966). This
makes decision-making possible even if those decisions may be less than
perfectly rational.

Whether in the lineage of Simon and other members of the school of

bounded rationality (see Jones, 2001) or within the incrementalist school,
decision-making for governance is being conducted within a set of con-
straints that tend to privilege certain types of decisions. These bounds are
in part cultural, so no American president would think of nationalization
as a mechanism for addressing economic policy problems.3 The bound-
aries may also be organizational. Governance is conducted primarily by
organizational actors and organizations tend to have their own internal
logics that they utilize to shape their internal dynamics as well as to
present themselves to the outside world (Goodsell, 2011).

The emphasis on organizational actors in governance points to the need

to consider the internal dynamics of organizations involved in govern-
ance. There is a tendency, especially in the rational models, to consider
organizations as unitary actors with a single perspective on policy and
governance. That internal integration is not necessarily true either for
public sector organizations or for social actors that interact with gov-
ernments to facilitate governance. Individual members of those organ-
izations may have divergent views (see for example the organizational
culture literature) that at times intervene in the decision-making.

Thus, there is a need to consider any set of positions taken in a bargain-

ing situation around policy issues to be a temporary equilibrium within
the organization and hence subject to change. Indeed, the process of
decision-making itself may produce some alteration of perspectives if not
 MODELS OF POLICYMAKING 51

of fundamental values. Even if the information being exchanged during

bargaining is imperfect it may still be new information that will alter the
perspectives of participants in the process, and also reveal new alterna-
tives for action.

Having mentioned Graham Allison’s study of the Cuban Missile Crisis

above with respect to rational models of the governance process, we
can also think of his “Organizational Process” model as to some extent
a version of bounded rationality. In this analysis of decision-making the
decisions were not structured so much by rational calculation as they were
by routines and institutionalized processes. A decision-maker, such as
a president, would have to find means of overcoming those institutional
routines if he wished to have the policymaking conducted differently.
These organizational routines were seen by the participants as means of
minimizing costs and of producing predictable responses to situations,
whatever that situation may be.

In summary, bounded rationality represents a seemingly more applicable

approach to public policymaking than does the comprehensive rational
model. The bounded rationality approach assumes that decision-making
involves actors that are constrained from making any comprehensive
decision by the complexity of the situations, the absence of complete
information, and the routinization of decisions within organizations and
established processes. Further, this model is not in any way remorseful
over not being able to make comprehensive decisions, but instead posits
that ultimately higher quality decisions can be made through this more
incremental and bounded process.

I have been singing some hymns of praise to bounded rationality as an

approach to decision-making for governance. There are, however, several
considerations about the approach that should cause some concerns,
both for academics and for practitioners. One is the assumption that
decisions can be made over again and again, so that any decision made is
reversible. While policies are constantly being remade, in the public sector
the constraint is that once organizations are created, personnel hired
and clients are being served, remaking the program becomes more dif-
ficult than assumed in these models. Indeed, the very logic of routinized
and path-dependent decision-making would seem to make continuing
replacement of programs very difficult for organizations and managers
(but see Mahoney and Thelen, 2010).
52 ADVANCED INTRODUCTION TO PUBLIC POLICY

Indeed, there is to some extent a paradox in the bounded rationality

perspective on decisions. On the one hand, there is the argument that
organizations, or other decision-makers, will learn from their own mis-
takes and adapt. But if their decisions are viewed through the perspective
of the bounds of their own culture and values – whether individual or
organizational – then they may be resistant to information that does not
correspond to their institutionalized conceptions. This difficulty may be
evident for initial adoption of programs but becomes more important
once initial decisions have been made and with them some understanding
of the policy area institutionalized.

Further, the bounded rationality approach may be conservative, and

prevent rapid adjustment to changing social and economic conditions.
The basic logic of bounded rationality is gradual adjustment of policies
to the environment, through trial and error, learning and multiple iter-
ations of policies. That gradual adjustment is suitable for problems that
are not severe or which change gradually, but not at all suitable for crises
or rapidly changing processes. Therefore, as we will develop below, this
model and indeed the others need to be understood and evaluated in
context.

Although the bounded rationality perspective on decision-making does

assume less structured decision-situations than does the rational perspec-
tive, it still is not especially well suited to the multiple players and inter-
active involvement that characterize much of contemporary governance.
Much of the discussion of bounded rationality occurs in the context of
a single, and generally governmental, actor whereas any realistic concep-
tion of modern governance is that it involves multiple players who may or
may not have the same set of intellectual bounds. Social and market actors
who may be directly involved in the decisions may well not conceptualize
the decision-situation in the same way as other actors involved. These
interactions among actors, each with their own bounds and own values,
do not fit easily into the seemingly more monolithic model of bounded
rationality.

And perhaps even more basically, bounded rationality assumes that

decision-makers do not necessarily have clear preference orderings. They
must, however, have some idea of their policy preferences if they are to
choose even an incremental step away from the status quo. They must
believe that the consequences of that decision will produce outcomes that
 MODELS OF POLICYMAKING 53

are better than the existing conditions, and although they may not have
considered all alternatives they are still making choices based on some
ordering of preferences.

Leaving aside for the moment the practical issues of application of ration-
ality and bounded rationality, there are several analytic questions that
arise. Perhaps the most significant of these is the simple question of how
big is an increment (Dempster and Wildavsky, 1979)? That is, if we are
attempting to understand policymaking, or attempting to advise a pol-
icymaker, how large a change can be expected to produce some effects
without being so large as to potentially upset existing social and economic
arrangements and waste large amounts of resources on a misguided
intervention? Unfortunately, the model does not provide any guidance
for either the analyst or the policymaker, so we are still left largely with
experimentation.

Finally, and to some extent related to the point about the conservatism
of the bounded rationality perspective, some problems simply require
large-scale interventions (Schulman, 1980). There is a tendency for
bounded rationality approaches to disaggregate policy problems and
attempt to solve them piecemeal. But large-scale and complex problems
need to be addressed more comprehensively, even if the approaches that
are adopted may not be comprehensively rational in any useful meaning
of that term.

Institutional analysis and development

Although she was a Nobel laureate in economics, Elinor Ostrom was

concerned with the range of influences on behavior in making and imple-
menting public policies. Her seminal work on resolving collective action
problems (1990) addresses means through which developing appropriate
institutions can overcome problems developing from rational individuals
pursuing interests that collectively create social problems. Also, along
with Larry Kiser (Kiser and Ostrom, 2000) she developed the institutional
analysis and development framework (IAD) that addresses public policy-
making in a more general manner, albeit to some extent still colored by
the concerns with collective action.
54 ADVANCED INTRODUCTION TO PUBLIC POLICY

The IAD framework recognizes the significant role that political institu-
tions play in shaping policy, but also recognizes that these structural influ-
ences are not the only factors shaping those policy choices.4 Unlike most
institutional approaches this framework is not content with structural
explanations but also attempts to integrate greater amounts of individual
agency into its explanations. Further, although it has a strong economic
foundation it also involves some aspects of the other social sciences as it
attempts to integrate individual and institutional levels of analysis. Trust
among the actors and institutions involved in making policy is central to
the dynamics of the framework (Henry, 2011a).

One of the more important foci for policy analysis utilizing the IAD
framework has been the same problems of collective action that much
of Ostrom’s other work has addressed. In addition to those issues, the
framework has been applied to complex governance arrangements such
as wilderness management and water systems that may involve multiple
actors with different priorities and differing levels of involvement in the
process. These socio-ecological systems present numerous challenges to
governing, as well as providing numerous opportunities for producing
positive outcomes (Ostrom and Basurto, 2011).

Although to some extent based on the bargaining among actors involved

in making and implementing policies, the operative element in the IAD
framework is a set of rules. These rules exist at three levels – consti-
tutional, collective choice and operational. Further, they may be both
formal and informal, moving this model away from the more structured
models associated with rationality and rational choice versions of institu-
tions. Thus, this model for making policy stresses the capacity for adap-
tation and bargaining among the actors, as well as the need to develop
understandings among the participants.

Advocates of the IAD framework argue that this approach is more capable
of describing interactions in complex, polycentric decision-making envi-
ronments than other approaches to policy (see House and Araral, 2013).
In addition to integrating perhaps the most important of the contri-
butions of the approach is the sensitivity to context and the attempt to
develop rules that match particular circumstances. This openness can, of
course, also be a problem given the absence of any definitive guidelines or
templates for designing policy interventions. Further, although the IAD
framework provides just that – a framework – it does not generate many
 MODELS OF POLICYMAKING 55

specific hypotheses. It continues to be used for a good deal of research on

collective action problems (Schlager and Cox, 2018) and in this research
continues to emphasize the importance of institutions in policy.

Multiple streams and the garbage can

If we move even further way from the model of comprehensive ration-

ality in governance the logic of multiple streams and the garbage models
can provide a perspective that introduces a good deal of randomness in
decision-making. Although written from somewhat different perspectives
and for different reasons the garbage can model (Cohen, March and Olsen,
1971) and the multiple streams model (Kingdon, 1985 [2003]) share many
of the same assumptions and appear to have relatively common implica-
tions for governing. Although discussed as models of decision-making
(and especially the Kingdon model) they both also contain significant
concerns with the formation of agendas for policymaking.

For both of these models there is an assumption that decision-makers may

not so much seek to make decisions as they have decisions thrust upon
them. While even in the bounded rationality model there is an assump-
tion (a principle of intended rationality in Simon’s terms) that individuals
and organizations responsible for governing are engaged in responding
to a perceived need to govern a particular domain or to solve a particular
policy problem, the actors in the multiple streams perspectives appear
somewhat more passive. They appear to realize that governing is difficult
and efforts to intervene in the complex chains of action involved in gov-
erning are fraught with dangers for the unwary (see Mucciaroni, 2013).
Thus, the most rational action may avoid making decisions rather than
seek to “solve” problems that may not be readily solvable.

Given those difficulties in making choices – the “organized chaos” with

which potential decision-makers are faced – the actors in the multiple
streams approach tend to wait for a “window of opportunity” to emerge.
In these windows of opportunity the several streams hypothesized to
be necessary for an effective decision come together. In the garbage can
model these various elements of decision are conceptualized as swimming
about in some primeval decisional soup, and when they come together
56 ADVANCED INTRODUCTION TO PUBLIC POLICY

there is the opportunity for decisions, and especially opportunities for

decisions that are likely to be successful.

There are several elements – streams – that must come together for
the decision to be made. The first of these elements is the problem.
Clearly both policymaking and governance involve addressing social
and economic problems. Whereas for the models above these problems
are almost inherently prior to the initiation of a decision process, in this
model they are initially merely there and become activated as the source
of action only when joined with the other elements (see Chapter 2). As we
will point out below concerning framing policy issues, the problems may
not exist for political reasons until they are recognized and conceptual-
ized as problems.

The second of the elements involved in a decision are the solutions, or

policies. These solutions may be conceptualized as a set of policy instru-
ments (Hood and Margetts, 2009) or as programs integrating resources
and law, but in either case these are possible means for solving policy
issues.5 The usual rational logic for policymaking would be that problems
pursue solutions, with decision-makers attempting to solve any problem
confronting them. In the multiple streams conception, however, solutions
may precede problems. Actors in the process may have a favorite solution
(a policy instrument) and then seek, or uncover by serendipity, new
uses for that solution. These solutions also can be seen as means for the
organization to enhance its own power and resources within the games of
bureaucratic politics (Peters, 2001b).

For governance the third element of this primeval soup may be the most
significant for explaining when decisions are actually made. This third
element is politics. The problems and the possible solutions can swim
around for very long periods of time until the political element aligns
with the other two elements. Although these three elements tend to be
discussed as relative equals in the decisions, politics create the opportu-
nities for attempting to solve the problem. This activation of the political
may be a function of a “focusing event” (Birkland and de Young, 2013),
or changes in the governing coalition, or even the result of the charisma
of particular leaders (Helms, 2012).

Above we noted that the confluence of the three elements of decision

explains, or at least describes, the timing of decisions. Unless and until the
 MODELS OF POLICYMAKING 57

elements come together there is unlikely to be a decision. It is important

to note here that this confluence of events is not particularly adept at
explaining the content of the decisions. The content of the decision tends
to be a function of which problem meet up with which solution in what
political context. This random element of the decision process may be
extremely unsatisfying for anyone attempting to impose a governance
style, or a particular policy design, on a policy area, but in this perspective
such “intended rationality” should not be expected to be successful. As
a result, any hypotheses about the manner in which the model functions
are difficult to develop (Herweg, Zahariadis and Zohlnhöfer, 2018), and
hence this model may remain more descriptive than predictive.

As mentioned above, in John Kingdon’s conception of the multiple

streams approach the confluence of events represented the opening of a
“window of opportunity”. When these windows opened decisions were
possible. This decision need not, however, be made on a conventional
ends-means conception of making a decision. Rather, it is equally likely
that solutions will chase problems as much as vice versa. For example, we
know that organizations and individuals have favorite policy remedies
(see Linder and Peters, 1989) and these actors may be looking for prob-
lems to which their favorite tools can be applied.

Although the above discussion has appeared to emphasize the passivity

of decision-makers in the multiple streams approach, they are also
characterized by Kingdon and others as “policy entrepreneurs”. The con-
ception that scholars working in this perspective have is that these actors
may have ideas about good policy, or good governance, and are merely
biding their time until they are likely to be successful in having their
ideas adopted and implemented. When the opportunity arises then these
entrepreneurs need to be ready to leap through the available windows and
push forward with a decision. This description of policymaking actually
conforms to the descriptions of many actors involved in government who
have ideas on the shelf for years, waiting for the right moment to put those
ideas into effect. Thus, these actors are not as passive as might be implied
by a quick reading of the model, but must be able to recognize an opening
for policymaking and be ready to seize the opportunity (Mintrom and
Norman, 2009).

The multiple streams approach represents an interesting confluence

of possibilities about decisions within itself. On the one hand, there is
58 ADVANCED INTRODUCTION TO PUBLIC POLICY

the underlying assumption that decisions arise when there is a virtually

random meeting of various streams required for the decision. On the
other hand, at least one version of this approach assigns a significant role
for a more activist entrepreneur to help shape the decisions. Indeed, one
could conceptualize the availability of an entrepreneur as yet another one
of the necessary elements that must come together in order for a decision
to be made. And if nothing else that entrepreneur may be crucial to the
actual content of a decision, given the policy ideas he or she may have on
their shelf, whether central or not in defining the opportunity for that
decision.

Political models of policymaking

Differentiating the following set of models as “political” appears to deny

the political content of those discussed above. That is not the intention,
and there is definitely political content in all the above models of deci-
sions, Rather, the intention of this label is to emphasize the political
content of the following set of models, or frameworks, for studying public
policy. These models place political and governmental institutions at the
center of the analysis and emphasize the ways in which institutions and
opportunities for participation affect the choices made. Further, they
tend to assume that political values will dominate over economic values
in policy choices.

Stages models of policymaking

The stages models of policy analysis (Hupe, 2007; Jann and Weigrich,
2007) has been the mainstay of political science approaches to public
policy. The stages model highlights the procedural logic of public
decision-making from goal setting to evaluation. This model is primarily
descriptive and heuristic, but it also assumes that the passage through the
various stages will shape the final policy selected. In particular this model
has a certain path dependency with decisions made early in the process
limiting the possible choices at later stages. This path dependence may
 MODELS OF POLICYMAKING 59

Table 3.1 Stages models of policymaking

Laswell Jones Dunn

Intelligence Agenda-setting Problem structuring

Recommendation Formulation Forecasting

Prescription Implementation Recommendation

Invocation Budgeting Monitoring

Appraisal Evaluation Evaluation

Termination

imply some elements of design, although designing policies is not the

central concern of the model.

The earliest of the stages models of policymaking was proposed by Harold

Lasswell (1956) as a general model of decision-making. He argued that
there were seven stages of decision-making, expressed in functional
terms, through which decisions must go (Table 3.1). The process began
with gathering intelligence and went through to appraisal and termi-
nation. In this framework termination was assumed to occur, perhaps
because the focus was on individual decisions rather than on programs
or policies. Most of the literature on policy termination argues, however,
that it is infrequent albeit not so infrequent as the critics of the public
sector would assume (see Geva-May, 2004).

The commonly used stages model, rather like that of Lasswell, posits a set
of stages necessary to move from the identification of an issue through
to evaluation and then feedback (Jones, 1984).6 This listing of stages does
identify a set of important steps in the process, but also begs a number of
questions. The most obvious of these questions is that this is meant to be
a process, but the emphasis is on a set of stages rather than the process
by which these various stages are linked. There is no clear internal logic
in the model that explains, or even describes, how issues move from one
stage to the next.

In addition, there is no internal logic in the model that explains what

happens at each stage, or why it happens. Each stage appears to have its
own internal logic and these are not linked by any common theoretical
perspective on decision-making. As demonstrated in the other chapters in
60 ADVANCED INTRODUCTION TO PUBLIC POLICY

this volume, stages such as agenda-setting and implementation have their

own theoretical perspectives that are largely unrelated to each other or to
other aspects of policymaking. Can we make a convincing argument that
some other dynamic, for example, bounded rationality, is working within
each of the otherwise discrete stages?

Further, the model implies that there is a natural order through which
issues will progress. In reality, however, the process may actually begin
at almost any stage, and then fill in the missing parts later. For example,
difficulties during the implementation of a program may lead to recasting
that program and initiating reconsideration. This framework also appears
to assume that policymaking is always writing on a tabula rasa, while in
reality most policymaking is actually modifying policies, sometimes fol-
lowing on from numerous previous modifications. This continuing mod-
ifications of policy make some of the stages, agenda-setting for example,
less difficult, but also may make others more difficult.

This simple version of the stages model has been supplemented by

other stages models that have somewhat greater theoretical content.
For example, Hill and Hupe (2014) developed a Multiple Governance
Framework that is somewhat analogous to Elinor Ostrom’s institutional
analysis and development framework (see above). This model assumes
that there are multiple levels of action within government, ranging from
higher level constitutional levels to operational levels. The process of
making policy therefore involves moving across these levels from grand
designs to the actual delivery of the services to the public. This model does
the important task of linking those levels of action but it also appears to
lack agency, meaning that there is no clear identification of the role of
individuals in policymaking.

Subsystems: policy causes politics

In addition to the stages model Theodore Lowi’s (1964) argument that

policy causes politics represents another of the political science mainstays
in this area of research, especially in the United States.7 The more conven-
tional argument would be that political action produces policies. Lowi,
however, argued that the nature of the policies would produce different
 MODELS OF POLICYMAKING 61

Table 3.2 Policy causes politics: Lowi

Applicability of Coercion

Individual Environment
Conduct of Conduct

Likelihood Remote Distributive Constituent

of Coercion
Immediate Regulative Redistributive

types of political action – regulation, distribution, redistribution and

self-regulation8 – as they are processed through the political system. Even
if the direction of causation is not always clear, it is clear in this argument
that certain political configurations are associated with certain types of
policies (Table 3.2).

The two variables defining these patterns of policy are the likelihood
of coercion and how that coercion is applied – directly or through the
environment of conduct. For example, regulatory policies tend to be
associated with direct controls over individual behavior that are expected
to operate quickly. In contrast, distributive policies, such as subsidies,
are expected also to influence individuals directly, but to do so more
remotely. In all these cases the logic is that attempts to reach certain types
of policy goals will necessarily invoke different constellations of political
forces and different styles of political action.

The argument that policies cause politics is analogous to some of the

discussion of policy problems mentioned below. The need to redistribute
income or other resources among groups in society is a particular type
of policy problem that in turn requires developing political coalitions of
a specific variety. In Lowi’s analysis the issue of designing interventions
raised by different types of policy is not so significant as the political
consequences, but in both perspectives on policy the nature of the policy
problems drive policy and political decision-making.

As well as being linked to the analysis of policy problems, the “policy

creates politics” logic also has been the foundation of a large literature on
policy subsystems (see Freeman and Stevens, 1987; McCool, 1998). The
argument for the existence of subsystems has been in part tied to the four
policy types that Lowi identified in his initial analysis. And the argument
is also more general, arguing that to understand policy we need to under-
62 ADVANCED INTRODUCTION TO PUBLIC POLICY

stand the numerous subsystems – also called iron triangles, whirlpools,

and a host of other terms. While we might want to talk about government
as a unified entity, this perspective on subsystems emphasizes the divi-
sions government and policy.

James Q. Wilson (1980) has developed another understanding of policy

subsystems based on the distribution of costs and benefits associated with
the policy (Table 3.3). Wilson’s argument is that both the benefits and the
costs of policy can be either concentrated or diffuse, and those character-
istics will define the type of policy intervention For example, if both the
costs and the benefits of a policy are concentrated there tends to a direct
political confrontation between the interest groups that would be either
the beneficiaries or the payers if the program were adopted.

Lowi’s and Wilson’s insights into policy were important in activating

policy studies in political science, and they continue to generate signif-
icant debate (McCool, 1998). While the insights have been influential,
they also raise a range of difficult questions. The most important is the
direction of causation. For example, if one adopts the multiple streams
perspective above, the confluence of the various political actors and
conditions may produce a particular policy choice. In addition, the model
does not really provide an explanation for the decision process or the
decision outcomes. The model does identify an important dimension of
policy within the public sector, but does not necessarily provide explana-
tions for policy choices.

The subsystems approach has been extended into an approach referred

to as “policy regimes” (May and Jochim, 2013; Workman, Shafran and
Bark, 2017).9 This perspective also attempts to link policies and politics,
and is centered on the policy problem that is the target of policy actions.

Table 3.3 Policy causes politics: James Q. Wilson

Costs of Policy

High Low

Benefits Concentrated Interest Group Client Politics

of Policy Politics

Dispersed Entrepreneurial Majoritarian Politics

Politics

Source: Wilson (1980).

 MODELS OF POLICYMAKING 63

By beginning with the problem the range of institutional, political and

ideational forces that affect the resolution of the issues can be integrated
and understood as a collectivity. Despite its attempts to integrate actions
addressed at particular problems the regimes approach does concentrate
on the role of government actors, tending to ignore the role played by
non-governmental actors in most contemporary policymaking.

The advocacy-coalition framework

A fourth political model for decision-making for public policy govern-

ance has been described as the “advocacy coalition framework (ACF).
As developed by Jenkins-Smith and Sabatier (1999; see also Weible,
2014) this framework addresses policy change more than initial inter-
ventions by governments into a policy domain. Further, the ACF model
of decision-making is based more on ideas than are the others discussed
above. Although the model is based more on policy change than are the
others as noted in several places most policymaking actually involves
adjustments to existing policies rather than making policy entirely from
scratch. Thus, change may be more of an issue for governance than it
might be thought to be.10

The fundamental assumption of this ACF model is that there are a set of
core ideas or beliefs that define a policy. Although there is a core of ideas,
these ideas vary in terms of their centrality for the policy (or governance
area) in question. An organization or other actor will attempt to protect
their core ideas against change, but will be willing to change more deriva-
tive ideas at the perimeter of their policy space. This change in policy will
come about when the status quo in the policy domain is confronted with
a challenge to the programs from some other actor that has ideas that to
some extent overlap and to some extent contradict the dominant set of
ideas operating within the governance domain.

As implied above, this approach to policy change and governance

assumes that subsystems are the dominant arenas for governing. Much
of governance theory tends to look at whole systems of governing and
to make assumptions about them, but the ACF focuses on more tightly
defined subsystems, with actors involved having core beliefs that tend
64 ADVANCED INTRODUCTION TO PUBLIC POLICY

to be stable. Thus, there are some similarities with the multiple streams
approach in that shifting away from the status quo tends to be difficult.
However, unlike the seemingly random nature of the multiple streams
perspective governance here is more purposive, and more driven by con-
flicting ideas than by the random confluence of events.

This conception of decision-making may be especially useful for the more

interactive versions of governance we are including along with more
state-centric conceptions of governance (Torfing et al., 2012). In these
cases we would conceive of the social actor – be it market or non-market –
as presenting some ideas that challenge the status quo. Although in most
cases this status quo will be defined by an actor within the public sector it
may also be that the dominant approach to governance is defined by other
social actors, or a coalition of public sector and social actors. Whatever
the source of the dominant governance arrangements, there will be chal-
lenges and those interactions must be understood to understand govern-
ance. These challenges play out over a more extended time span than in
many models of policy change, and also involve a number of mechanisms
such as policy learning (Weible et al., 2011).

The interaction of the dominant governance arrangements and the chal-

lenges coming from external actors, whether they are state or non-state
actors, in the process of producing a new equilibrium should be concep-
tualized as a decision-making process. In this perspective on policy, the
actors tend to bargain over changes in the status quo, but more impor-
tantly they learn about the ideas of the other actors and the potential for
constructive change. This conception of decision-making in governance
is therefore more open to external influences than are many, and assume
that interactions among actors produce the capacity for improvement in
governance.

The advocacy coalition model of decision-making adds to the other

approaches discussed here by bringing in the need for bargaining more
explicitly. The other approaches appear to have a more hierarchical con-
ception of the manner in which decisions are made, differing primarily
in the extent to which rationality, in any comprehensive sense, can be
expected to emerge from the process. The ACF assumes that there are
at least two possible policy coalitions available, each with a set of beliefs
about policy and governance. Through a process of negotiation and
learning, a dominant coalition will be able to impose some or all of its
 MODELS OF POLICYMAKING 65

wishes on the area. That said, the process through which that bargaining
is conducted is not specified as clearly as it might be. Further, the confron-
tation of different ideas may, however, exacerbate polarization among
approaches rather than producing compromise (Henry, 2011b).

Powering, puzzling, and so on

The stages model of policymaking addresses some elements of the

political process, but does not address questions about what actors
are involved and influential. Understanding the course through which
policies are adopted and enacted is important, but identifying the key
actors is at least as important. While perhaps not as clearly articulated
as a theory of policymaking as some of the approaches discussed above,
Hugh Heclo’s observations about decision-making in pension reform in
Britain and Sweden (1974) do point to important issues in making policy.
Heclo observed that there were two dominant governmental styles of
making policy. One depended upon formal hierarchy, and the associated
political power, to make policy decisions and to ensure that they were
enforced. The authoritative policymakers involved in policymaking may
take advice, but in the end they would be able to make their decisions
based primarily on their formal positions in the political system.

The alternative perspective identified by Heclo was characterized as

“puzzling”, as opposed to powering. In this second approach to making
policy decisions information and analysis are more important than the
relative power of the actors involved. This puzzling approach assumes
that the best ideas will prevail in policy, and that the policy process can be
designed to maximize the role of knowledge. While the puzzling approach
cannot function without being backed by adequate political power, it will
use that power to validate policy choices made by more analytic means.

The initial puzzling and powering dichotomy provided an interesting

political perspective on decision-making, but other scholars have added
other dimensions to that analysis. Most importantly, Robert Hoppe
(2010) added a third component of “participation” to the possible styles
of decision-making. This addition reflects the increasing importance of
mechanisms for direct public participation in shaping public policies,
66 ADVANCED INTRODUCTION TO PUBLIC POLICY

as well as the established patterns of accommodation in many Northern

European countries. That participation is being manifested through net-
works, deliberative mechanisms, or a range of other possibilities.

As well as maintaining the alliteration, the addition of participation to

the initial dichotomy reflects different influences on policy than might
be manifested through the two other styles of policymaking. Both
the puzzling and the powering metaphors imply more elitist styles of
decision-making, while adding participation adds more of a populist
dimension. Making decisions about policy, especially complex policy
issues, may always have a strong elite component but at the same time
there are means of involving the public and adding different perspectives
to the process.

Pepper Culpepper (2002) added yet another alliterative component to this

analysis – “pacting”. By this he meant the use of social pacts as a means
of institutionalizing the interactions between social actors and the policy-
making process. Social pacts (Rhodes, 2001) are analogous to corporatism
(Molina and Rhodes, 2002) as a means of formally linking interest groups
and other organized interests to the state and to the policy process. Thus,
“pacting” is a specialized version of participating, but that participation
is more structured than the perspective on participation discussed by
Hoppe. Further, pacting adds another source of political power to the
process by adding the active involvement of interests potentially affected
by the policies.

Finally, Mark Blyth (2007) added “persuading” to this litany of political

approaches to policymaking. This addition points out that discussion
and persuasion play a major role in the policy process. The “persuading”
perspective does not go as far as the argumentative approach discussed
earlier (see Chapter 1; see also Stone, 1997) but it does emphasize the
use of ideas in debates over policy – and not just the ideas of experts
as implied by the “puzzling” perspective but also broader participation
through means approaching deliberative democracy. This perspective
further can be seen to constitute a general version of the ACF in which the
advocacy of different ideas about policy is also central to understanding
policy dynamics.

All these words beginning with “P” make several important points about
how policy gets made in the political arena. First, as we are talking about
 MODELS OF POLICYMAKING 67

public policy, there is a role for political power, but that power often is
incapable of being exercised as raw power, especially in a democratic
system. Experts, social actors and even the general public may have some
things to say about the policy choices being made. Further, these versions
of policymaking do not emphasize the policy process to the extent of some
other models, but they do assume that process, especially the interactions
among numerous actors, is significant in shaping the policy decisions.
And finally, these models are all about making decisions as crucial for any
understanding of public policy.

Contingent models and choices

We have now identified several important models for policymaking,

We have made some allusions to the settings in which one or the other
frameworks may be particularly useful, but is this section we want to
expand that discussion and attempt to make even closer linkages between
policymaking models and environmental considerations. The remainder
of our discussion of policymaking models will be focused on the manner
in which decisions are made in different policy areas and in different
institutional settings.

The examples drawn from the Cuban Missile Crisis point to one impor-
tant dimension of variation – the extent of real or perceived crisis
(see Boin, 2004) – but there are a host of other variables that might be
included. Perhaps the most fundamental point here is that policymak-
ing is not the same across policy areas and dependent on the degree of
urgency in making a policy choice. As Chapter 4 following here tends to
focus on differences in decision-making across different political systems,
this section will place greater attention on the role of particular types of
issues, and the role of different types of situations, in shaping governance.
This list will not be exhaustive, given the wide variety of influences there
may be over governance, but it is intended to provide some insights into
the variability of decision-making for governance.

The most fundamental difference among policy issues is the extent to

which they are programmed. That is, to what extent are the actors and
the premises for decisions well established through law or through insti-
68 ADVANCED INTRODUCTION TO PUBLIC POLICY

tutionalized patterns of interaction. The policy regimes literature with

its emphasis on more or less formalized interactions appears particularly
well suited to address highly legalistic, programmed decisions. The
advocacy coalition approach also appears well suited here, given that it
assumes that there is an established pattern of policy which is being sub-
jected to pressures for change.

For example, neither the causes nor the remedies of so-called “wicked
problems” (see Chapter 2) such as poverty or many aspects of environ-
mental policy are well identified. Therefore, rational, and perhaps even
bounded rationality, approaches would not be particularly useful for
these cases. More politicized and experimental interventions may be
the only available options for difficult policymaking situations such as
those. Unfortunately, however, governments cannot avoid responsibil-
ity for intervening in these difficult issues, so the less formal means of
decision-making become crucial.

Conversely, highly programmed decisions may approach the criteria

required for rational decision-making. This is perhaps particularly true
if the decision is remaking an existing policy for a second, or third, or
nth time (see Hogwood and Peters, 1983, p. 198). In such situations the
preferences of the actors should be relatively clear, the consequences of
decisions may be more knowable, and information sources may be better
developed. These conditions may also be seen as reflecting bounded
rationality, with the experience of the actors in this particular policy area
representing the bounds that can make decisions more feasible.

As mentioned above, crisis also may be the antithesis of programmed

decisions. Crises tend to reduce the capacity for rational decision-making
given that preferences may not be knowable in a fluid situation and the
range of decision premises will not be constrained (Boin, ‘t Hart and
Sundelius, 2005). To some extent crises are examples of decision-making
through multiple streams, with the crisis being analogous to a window
of opportunity in which the demand for decisions may force bringing
together solutions with the problem presented by that crisis.

In addition to the nature of the problems being confronted, the political

environment of policymaking also affects the relative utility of these
frameworks. Frameworks such as the stages model are likely to be more
effective in institutionalized political systems. Also, political systems that
 MODELS OF POLICYMAKING 69

have greater policy analytic capacity will be capable of governing more by

“puzzling” than by “powering”. These systems with greater policy capac-
ity, including analytic capacities, also may be able to approach rational
policymaking more closely than can those without such capacity.

Summary: multiple models, but what is missing?

There is no shortage of models and frameworks directed at making

and understanding public policies. Indeed the multiplicity of models
makes doing them all justice in this one chapter extremely difficult. The
models all provide some insights into these processes, and all offer some
explanations for the outcomes of those processes. And they all have weak-
nesses and blind spots that prevent their being complete and dominant
explanations for policy. This combination of strengths and weaknesses
in the frameworks leads to at least two significant questions about how to
advance the understanding of policymaking.

One means of attempting to address this broad array of approaches, each

of which does some things well, but others not so well, is to consider com-
binations of approaches. For example, the multiple streams perspective
on policy describes the almost random elements of some policymaking
well, but it tends to be largely institution-free. Can we combine that
approach with the stages approach that emphasizes institutions and more
formal aspects of government? This marriage of approaches might illu-
minate not only the processes at each stage of policymaking, but also the
role that institutions can play in developing and channeling the streams
of influence.

These models might also be enhanced by adding yet additional elements

to the models. For example, most of these approaches to policy tend to
downplay the role of agency. Therefore attempting to include individual
level behaviors into these models should not only help to explain more
about the manner in which decisions are made, but also remove some
of the more mechanistic features of the models. The role of the policy
entrepreneur is one example of adding agency to a model that might oth-
erwise depend excessively on randomness, and the role of entrepreneur or
70 ADVANCED INTRODUCTION TO PUBLIC POLICY

“animateur” might be added to several of the other approaches to provide

more of a locus of decision-making.

Further, these models are all discussed in rather generic terms, even if
they are largely derived from the experiences of industrialized democ-
racies, and therefore greater understanding of context would also add
to the utility of these approaches to public policy. Some of the models
may be less applicable in other social, economic and political settings.
For example, the ACF appears to depend upon multiple sources of policy
ideas that can contend in a relatively open policymaking arena. Both the
absence of contending sources of policy ideas and the closure of possible
answers because of ideology or partisan domination may make this model
less relevant in many international settings.

Notes
1. This argument is similar to that of Pressman and Wildavsky (1974) concern-
ing implementation, with larger numbers of veto points (clearance points in
their terminology) likely to produce failure, or certainly deviations from the
intentions of the policymakers. See Chapter 5.
2. But sometimes this is the only viable approach to the issue being confronted
(Schulman, 1980). Large-scale issues, as described in Chapter 2, require
large-scale and comprehensive solutions.
3. Except, of course, when they do. Both George W. Bush and Barack Obama
partially nationalized automobile manufacturers and banks and insurance
companies in response to the financial crisis beginning in 2007.
4. For a discussion of the roots of the IAD see Aligica and Boetkke (2009).
5. Solving may be an excessively optimistic word. Most policy areas, and
domains of governance, are never really solved but represent overlays of
multiple attempts at solution (Hogwood and Peters, 1983; Carter, 2012).
6. For another version see Dunn (2012).
7. Jeremy Richardson’s ideas about policy styles (1982) are not dissimilar from
Lowi, and have been more influential in European political science.
8. In another version of the argument, Lowi (1972) replaces “self-regulation”
with “constitutive” policies. That latter term means policies that define
structural relationships within government and/or between government and
policy actors.
9. See also the literature on urban regimes (Pierre, 2014).
10. Indeed, the tendency of the literature on governance and policy to write as if
decisions were innovations may seriously misrepresent governance.
4 Agendas, agenda-setting
and framing

Before the public sector can act on a policy problem it must find a way to
move it from a matter of general social concern into the public arena for
consideration. There is no shortage of important issues that governments
could, and should, act upon, but for governments to make those policies
the issue must work its way from some concern in society, or perhaps
a negative segment of society, into the institutions of government. That
movement onto an active agenda of government requires political action,
often by political parties or interest groups. That said, governments are by
no means totally passive, waiting for issues to come over their proverbial
transoms. Rather, many actors in government, perhaps especially the
public bureaucracy, are active in creating their own agendas for collective
action.

Issues that do make it to some collective agenda for action are not nec-
essarily successful – they only have an opportunity for being considered.
The absence of guarantees for success is in part a function of how quickly
issues move on and off active agendas. This “issue-attention cycle” reflects
not only the fickle nature of public opinion, but also the pressures of
numerous issues on governments that have limited time and resources
(Downs, 1972). There is always something else occurring in the economy
and society that can excite the public, the media and government, and
hence any one issue may have a relatively short appearance on the
political stage. The issue-attention cycle is perhaps accelerated by the
“mediatization” of governing, with the media setting the timeframe for
government rather than vice versa (Schillemans and Pierre, 2016).

This chapter will examine the process of agenda-setting in government.

I will first discuss the nature of agendas themselves, and some means of
shaping issues so that they have a greater probability of making it to an
agenda. This includes the basic issue of how to frame issues for action in

71
72 ADVANCED INTRODUCTION TO PUBLIC POLICY

the public sector. I will then conclude the chapter by discussing several
theories about agenda-setting and the politics of forming agendas. This
chapter will deal very much with the politics of issues and agendas, while
the following chapter will discuss issues and policy problems more from
the perspective of policy analysis.

Agendas and agenda-setting

I have already made the point that agendas are crucial for making policy,
but just what is an agenda. In its simplest form an agenda is the set of all
issues that governments (again, along with their allies) will act upon (see
Cobb and Elder, 1972). To examine this aspect of the policy process in
somewhat greater detail there are at least three forms of policy agendas.
The most general version of agendas is the systemic, or informal, agenda,
meaning all those issues that have been accepted onto the agenda for
consideration, whether they are actually being actively considered at any
one time or not. The systemic agenda includes a large array of issues that
have been accepted as legitimate objects of action by the public sector.
Some may be considered settled for the time being, but can be reacti-
vated when change is desired. For example, European governments have
a well-developed interest in the welfare state but may not be legislating or
adjudicating about it during any one period.

At a second level of generality there are institutional agendas that contain

the issues that an institution is working on actively at any time. Different
institutions may or may not be processing an issue at any one time, and
issues may move back and forth among institutions. One important polit-
ical challenge for agenda-setters, therefore, is to move issues along. The
bureaucracy, of course, is almost always processing issues as they imple-
ment programs, and the individual administrators tend to develop ideas
about how to improve the policies for which they are responsible. Their
task therefore is to find ways to move those concerns into legislatures or
the political executive in order to have the laws changed.

Politically, it is crucial to understand how an issue moves from the infor-

mal agenda to some formal institutional agenda. After the initial recog-
nition of the underlying problem in the society, the second stage is that
 AGENDAS, AGENDA-SETTING AND FRAMING 73

those perhaps vague worries about a social situation have to be specified

in a manner suitable for being addressed through the policy process. After
that specification of the issues to be considered, concern with the issue
must be expanded to include a wider range of political and social actors
so that some coalition can be built to enact the desired reforms. After the
issue has been made sufficiently broad it can then be moved onto an active
agenda in some institution or another.

Different types of political systems provide more or fewer institutional

agendas for the would-be agenda-setters. For example, federal and other
decentralized systems provide another whole set of institutional opportu-
nities for agenda-setters (Chappell, 2002) other than central government.
Likewise, systems such as the United States and Germany that have pow-
erful and active court systems provide another option, and one that has
proven significant in dealing with issues such as civil rights that were dif-
ficult to process through the more political institutions (Rose-Ackerman,
1996). That said, multiple points of access for agenda-setters may mean
that although access is easier, moving issues around among the insti-
tutions and making definitive decisions may be more difficult than in
simpler systems.

The availability of multiple points of entry into the policy process pro-
vides the participants opportunities for venue-shopping (Baumgartner
and Jones, 2010). This search for the institution offering the best oppor-
tunities for any particular policy may be highly strategic, but often there
is inadequate information to make such choices. Further, different venues
may provide opportunities for advocacy groups that go beyond winning
or losing on a single issue and open new opportunities for action, and
for gaining new partners in their political struggles. In particular, the
increased internationalization of policies offers new venues with great
potential for groups seeing policy change (Pralle, 2003; Ley and Weber,
2015).

Finally, some issues are on recurrent agendas. The most obvious example
is the public budget that in most political systems comes onto the agenda
every year (see Walker, 1977). And given that all public programs require
budgetary funds, this means that to some extent every program comes
onto the agenda regularly, even if the substance of the program may not
be debated to any great extent. In addition, many programs are author-
ized for only a limited period of time, requiring them to come on to the
74 ADVANCED INTRODUCTION TO PUBLIC POLICY

agenda regularly. This recurrence may be beneficial to popular programs

but is a threat to less popular programs that may be better served by
having issues remain settled and off the active agenda.

Shaping issues for the agenda

Getting an issue on any one of these agendas is important for the advo-
cates of a program, but these advocates are often at the mercy of other
forces. That said, however, those advocates can attempt to shape issues
so that they are more likely to place them on the agenda. Some aspects of
issues may not be controllable by their supporters, but others can be. In
either case, understanding characteristics of issues that enhance the ease
with which they can be placed on the agenda can inform the strategic
choices of advocates, as well as their understanding of the likelihood of
success in the process.

Issues are easier to place on the agenda if their effects are, or are perceived
to be, severe. It is easier to place AIDS, Ebola or COVID-19 on the polit-
ical agenda than it is the common cold, even though many more people
are likely to catch colds. And influenza outbreaks may be even easier than
AIDS because there is no stigma potentially associated with the disease.
Similarly, governments – even the neo-liberal American government –
were able to make much more extensive interventions in the economic
crisis following 2008 and during the COVID-19 crisis than they might in
minor downturns of the economy.

As well as having severe effects, it may be beneficial to the would-be

agenda-setter if the effects cover a large number of people. Further, the
possibilities of placing an issue on the agenda are increased if the effects
are concentrated, either geographically or socially. For example, higher
levels of unemployment are always important politically, but may be
more of an issue if that unemployment is concentrated in one area of the
country, or in one social group. And that effect may be magnified if the
group or region has been the subject of other deprivations, for example,
immigrants in most European countries.

The previous two criteria can justify the use of terms such as “emergency”
or “crisis” as a means of moving issues onto the agenda. For example,
labeling long-term unemployment insurance in the United States as
emergency was used as a means of bringing it back onto the agenda for
 AGENDAS, AGENDA-SETTING AND FRAMING 75

additional consideration and extension in 2014. As we will discuss below,

the literature on “focusing events” as a means of explaining agenda-setting
also utilizes crisis as a means of explaining how issues can be moved onto
an agenda. Even though a policy problem may exist for some time, it will
not be addressed until an event brings the problem into sharper focus.

Placing an issue on the agenda is also facilitated by attaching important

political symbols to the issue and to the programs used to address the issue.
For example, children are commonly a powerful political symbol, so even
programs that may go against the normal political grain of a country can
be successful if they can be “sold’ as creating benefits for younger citizens.
Also, at least in the United States, defense and national security have been
important symbols so it is no surprise that the interstate highway system
is formally labeled the National Defense Highway System.

One particular version of symbols involved in setting agendas is the

availability of analogies of policy situations and policy interventions.
The use of analogies is more common in making foreign policy than
in making domestic policy (Houghton, 1998), but being able to make
a contemporary issue appear like an older issue can be useful politically.
If policymakers can say “We have done this before”, or “When something
like this happened previously it was a major problem” those policymakers
have a better chance to persuade others that they should intervene. The
analogies are often false, but they can still be persuasive (Boscarino, 2019).
And analogies can also be used to prevent action, if it can be argued that
when this type of problem was addressed before it proved to be insupera-
ble, or that something that appeared to be an impending crisis was really
quite minor.

The active political agenda-setter is capable of assisting in getting issues

on the agenda. While some aspects of issues are givens, other can be
molded in ways that can make them more likely to be placed on an
active institutional agenda. This activity by interested individuals can
be crucial, given the large number of issues that compete for attention
by government. This advocacy and shaping of issues is a rather basic
way of understanding agenda-setting, but the importance of this stage of
policymaking has produced a range of theoretical models of this process.
And the possibilities of manipulating issues in ways that can make them
more amenable to being placed on the agenda leads on to a discussion of
framing issues.
76 ADVANCED INTRODUCTION TO PUBLIC POLICY

The politics of agendas

While all the above discussion is to some extent political, we should con-
sider some specifically political models of how issues are developed and
then placed onto agendas. The most fundamental issue in understanding
the politics of this part of the policy process is the extent to which it is
open to popular ideas, as opposed to being controlled by political elites
or powerful interest groups. As is so often the case, there is some evi-
dence supporting all these positions and the apparent strength of one or
another of the perspectives on agendas may depend upon the policy area
being considered or the political system within which the agenda-setting
process is being conducted.

The elitist position on agenda-setting is perhaps the easiest to maintain.

This argument is simply that political and economic elites dominate the
process of setting agendas, as well as politics more generally. Even in
political systems that are nominally democratic, or indeed truly demo-
cratic, the capacity of money for campaigns, party support, or even direct
bribery may give the more affluent greater influence. Even if money is not
involved, the political and economic elites tend to move in the same social
circles, and often come from common backgrounds, so that the elite will
have greater opportunities for influence and control. And the elite also
may control the media so that the information available for policymakers
may also be skewed in certain directions (Carnes, 2013). The strengths of
elites may be especially powerful within specific policy domains where
expertise reinforces power (Genieys, 2010).

The pluralist or more egalitarian position is that democratic systems

tend to be open to a variety of influences so that there is roughly equal
opportunity for all to influence agendas (McFarland, 2004). In this view
the public sector is an open arena within which various interests contend
for influence, with no certain winners and no certain losers. Adherents
to this position can point to a variety of successes for the interests of less
affluent citizens and minority interests. This position appears particularly
viable in countries such as Germany, Canada and the United States where
multi-level governance and a reasonably activist judicial system provide
more opportunities for access than in more confined political systems.1
Even in more centralized regimes there may be structures, as well as
a political culture, that facilitate the influence of non-elite actors.2
 AGENDAS, AGENDA-SETTING AND FRAMING 77

When we consider agenda-setting and the relative power of different

actors to influence the agenda we usually consider the evidence con-
cerning who is successful in getting their issues moved from some broad
agenda onto an active institutional agenda, and then in winning once the
issue is acted upon in government. We should, however, also recognize
the capacity of powerful actors to prevent issues from being considered
at all. This “second face of power” (Bachrach and Baratz, 1962; see also
Lukes, 2004) demonstrates that the most certain way of preventing any
threat to the positions of entrenched interests would be to have the con-
cerns of outsiders totally ignored by decision-makers in the public sector.

Another means of classifying the agenda-setting process is to consider

whether the impetus for new policies, or for policy change, comes from
outside government or from inside. We tend to think of the politics of
agenda-setting as involving individuals and groups on the outside of the
public sector attempting to press their views on the public sector. While
that is certainly true for most policymaking, there are also numerous
instances in which government organizations and actors are the prime
movers. There is some tendency to consider the bureaucracy as faceless
automata without policy ideas of their own, but that is rarely true, and
public servants who have been working in a policy area for years will
clearly have ideas about how to improve policy.

The agenda items coming from within the public sector may not be major
transformations of public policy but there are numerous instances of
change, especially reforms of existing programs. The role of actors within
the public sector as agenda-setters – usually the public bureaucracy – may
be especially relevant within the “multiple streams” approaches to poli-
cymaking such as that of Kingdon or the “garbage can” (see Chapter 3).
That is, members of the bureaucracy, or organizations with clear policy
ideas, may wait for an opening presented by a change in government, or
a focusing event, to bring those ideas forward for action.

Summary
From the above it should be clear that the agenda for policymaking in
most instances is made – it does not just happen. Political actors, both
inside and outside government, are actively involved in moving issues
onto the active agenda of political institutions. Other actors may be
working as vigorously to prevent those issues from being considered. That
78 ADVANCED INTRODUCTION TO PUBLIC POLICY

said, however, these actors are to some extent at the mercy of external
forces in society that heighten the attention of the public and political
elites in certain issues and dampen concern about others. The role of the
media in both reflecting and activating attention to policy issues has pro-
duced a significant strand of research in public policy, a subject to which
we now turn.

The issue-attention cycle

One of the more important models of agenda-setting within political

science began with a rather simple observation that has been referred
to as the “issue-attention cycle”. Anthony Downs (1972) observed that
issues came onto the political agenda with “alarmed discovery” but once
the real difficulties of actually doing anything significant and successful
about the issue become apparent then the interest tends to wane. Further,
concern about most issues is quickly replaced by interest in the next issue,
and then concern about that first issue becomes quiescent. That first issue
remains part of the systemic agenda, albeit not an active agenda, until it
becomes activated again.

The second root of this strand of literature is E.E. Schattschneider’s (1962)

discussion of the expansion of political conflict. His argument was that the
political system in the United States tended (and still tends) to segment
policymaking and to prevent actors not usually involved in a particular
policy domain from breaking through and influencing choices. This
segmentation tends to maintain existing patterns of policy and permit
only incremental forms of change. Once settled, the institutions and
the powerful actors who work within them prevent an issue from being
considered by a wider range of actors, or perhaps at all. This is analogous
to the arguments of the historical institutionalists (Robinson and Meier,
2006) that policies are path dependent, with change coming through
major events – punctuated equilibrium. Although this rather pessimistic
model of government was developed in reference to the United States,
similar patterns can be observed in other political systems.

Following from this initial observation a more elaborated version of

this approach has been developed as “punctuated equilibrium theory”
 AGENDAS, AGENDA-SETTING AND FRAMING 79

(Baumgartner and Jones, 2010; Jones, 2001). This is one of several models
in the social sciences that assumes the presence of an equilibrium that is
disturbed by some external force. In terms of setting agendas the presence
of an equilibrium means that the policy process can be extremely con-
servative. In this perspective issues and particular ways of conceptualizing
issues will persist without some sort of exogenous shock, or perhaps
without the presence of a policy entrepreneur.

Information processing plays a significant role in the punctuated equi-

librium model of agenda-setting (Jones and Baumgartner, 2012). In this
perspective information concerning the exogenous shock (see focusing
events below) is imported into the policymaking system and depending
how the information is understood and processed, it will produce some
action, and perhaps very high levels of action, from policymakers. Even
then the system will tend to revert toward something approaching the
status quo ante without reinforcement.

The conservatism of most decision-making in punctuated equilibrium

models can be seen as having much the same root as incremental theories
of policy (Dahl and Lindblom, 1953; Lindblom, 1965). Given the huge
number of decisions with which policymakers must contend on a regular
basis, and the complexity of each of those individual decisions, policy-
makers adopt simplifying rules of thumb that enable them to contend
with policy. This means that most decisions being made are made incre-
mentally, with small movements from the status quo, a style that preserves
much of the existing policy frame. This tendency is heightened by the
domination of policy areas by specific interests and the negative feedback
that may result from attempts at change (McCool, 1998).

Given the tendency of negative feedback and incremental decisions to

maintain particular patterns of policy, once attention is focused on the
existing policy then the change that is triggered is likely to be extensive.
That trigger may be from a focusing event (see below) or perhaps from
greater general awareness of problems in the policy area. Once the
interests and leaders who have been protecting the policy subsystem
from external influences are incapable of maintaining that isolation, then
the problems that may be pent up within the area become more widely
apparent.
80 ADVANCED INTRODUCTION TO PUBLIC POLICY

Somewhat analogously to Down’s idea of “alarmed discovery”, once the

underlying policy problems do become salient they tend to result in what
scholars working in this approach have called “disproportionate informa-
tion processing” (Workman, Jones and Jochim, 2009). That is, whether
because of the accretion of problems or simply the opportunity to address
issues that had been protected from broader consideration by members
of the subsystem, there tends to be a large-scale response to the available
information in the field, and perhaps non-incremental change Again,
however, that spate of change will tend to be short-lived, with a return to
quiescence when other issues take center stage in the policy world. Moshe
Maor describes this as “policy overreaction” and provides more extensive
explanations for the phenomenon (2012).

The mass media and other sources of information are major players in
the issue-attention cycle. In some instances the media can be the source
of the attention focused on the policy area. In other cases the media may
merely reflect changes in attitudes and concerns among the public. At
times the shift in attention may be the product of political entrepreneurs
who press for opening the policy subsystem, and the pressure may come
from social movements. However the change is initiated, the shift away
from incremental decision-making to more significant change represents
an opening of agendas for transformation, albeit one likely to be followed
by another period of incremental changes.

Focusing events and agenda-setting

The use of so-called focusing events is a special case of the agenda-setting

process. As already noted above, proclaiming a problem or an issue
a crisis can be a useful technique for placing an item on the agenda, but
there are also genuine crises and other extraordinary events that require
governments to move out of their well-established routines and add new
items to the agenda for action. The analysis of these events began with
their definition in John Kingdon’s seminal work on agendas (Kingdon,
1985 [2003]), and were seen as providing a “little push” from events that
can permit individuals who have ideas about policy the opportunity to
move those items from some limbo populated by good ideas into active
consideration.
 AGENDAS, AGENDA-SETTING AND FRAMING 81

While Kingdon proposed this concept, it remained rather vague (see

Birkland and de Young, 2013). To some extent it conflates sudden, major
events with more gradual learning about policy opportunities and options
(May, 1992). It also tends to conflate natural events and crises with oppor-
tunities for policy action created by intentional mobilization of interests,
for example, the civil rights or anti-war movements. Finally, the idea has
relatively little predictive capacity, although one can argue that such an
event did occur after the fact. Why do some events produce major policy
shifts and others do not? This concept, and Kingdon’s analysis of it, do
not provide a ready answer to that question.

Birkland (1998) discusses “potential focusing events” and considers why

some events are effective in moving an issue onto an active agenda and
others are not. The answer to this question is complex but somewhat
like the punctuated equilibrium model discussed above, the cumulation
of problems and their revelation, along with potential solutions tends to
produce policy change (Birkland, 2006). Again, negative outcomes and
failure tend to drive issues back on to the policy agenda to a greater extent
that do the opportunities to improve policies that are working reasonably
well (Best, 2010). The familiar adage of “If it ain’t broke don’t fix it”
appears to work in policy as well as other areas of human endeavor.3

It is also important to note that focusing on policy problems may not just
be a function of cataclysmic events. Sometimes it may be a simple act of
civil disobedience such as Rosa Parks on the city bus in Montgomery,
Alabama. Or it may be the publication of a book like Rachel Carson’s
Silent Spring, or the decision of a vegetable seller in Tunisia that he could
no longer accept repression by the government. It is almost impossible to
predict when an event or any other action will produce a disproportionate
policy response. What does seem clear, however, is that there are powerful
forces maintaining the status quo that may require some spark – natural
or human – to overcome.

Further, institutional structures may facilitate or reduce the impact of

focusing events on policy change. The more organizations and institu-
tions that are involved in the process of recognizing and then acting upon
focusing events the less likely there is to be quick, and perhaps dispropor-
tionate, response to the event (Maor, 2019). Similarly, focusing events that
fall between the conventional “silos” of government may be less likely to
provoke a strong response from the policymaking system. That said, the
82 ADVANCED INTRODUCTION TO PUBLIC POLICY

economic and social fallout from COVID-19 has made it more apparent
that policy problems cannot be understood adequately without under-
standing the connections to other policy domains.

Framing

Much of the above discussion of agenda-setting has taken as a given the

nature of the issues being considered. While for many well-established
policy problems that may not be a particular concern, for newer issues
coming on to the agenda their definition cannot be assumed. And even
for more established issues changing politics, and changes in the economy
and society, may alter the manner in which the issues are regarded. Thus,
in order to understand the process through which issues are placed onto
an agenda, and the form in which these issues actually appear on those
agendas, understanding how they are “framed” is important.

At the extreme, issues must be framed to appear on the policy agenda, or

even to be recognized as problem. For example, while spousal abuse is
now widely recognized as a policy problem in earlier times this behavior
was largely assumed to be a normal part of family life. Therefore, for
governments to be able to criminalize the actions, and to provide more
supportive forms of intervention for the victims, this behavior had to be
framed as a problem, and then as a public problem. The same pattern has
been identified for child abuse (Nelson, 1984) and for abuse of the elderly,
with these problems remaining widely accepted until political agitation
identified them as major social problems.

Framing is important in the initial movement of an issue onto an

active agenda of institutions in government, but it also is significant as
a mechanism for understanding policy change and creating coordination
among programs. One possible cause of an absence of coordination and
cooperation among programs and organizations is that they represent
different ideas and different frames of reference (see Bardach, 1998). For
example, although social policies and labor market policies may have
some common goals, and some common clients, they tend to have very
different underlying ideas about the causes of unemployment, inequality
and poverty.
 AGENDAS, AGENDA-SETTING AND FRAMING 83

Policy problems can also be framed in a collaborative manner, rather

than having conflict over the frame. Botterill and Fenna (2019, pp. 53–4)
present the contrast of how drought was framed in Australia and the
United States. In Australia it was framed as an agricultural issue and other
possible frames were excluded, while in the United States it was framed
to include a wide range of interests. But that collaborative framing also
made formulating policies more difficult than when there is a singular
understanding of the issue.

Even when policy issues have been accepted onto the agenda, and have
been framed as an issue suitable for public sector intervention, the ques-
tion of framing has not be answered entirely. It is not only a question that
there is an issue, the real problem arises when there are multiple interpre-
tations of the issue. The classic example of this framing problem is drug
policy (see Payan, 2006). This issue is usually addressed as a problem of
law enforcement. There are, however, several other possible frames for
this issue. For example, drugs constitute a major health issue, and also
may be a consequence of social problems such as family breakdown and
abuse.

These several possible frames for drug issues, and indeed multiple frames
for any policy problem, can engender political conflicts. Many of these
conflicts are organizational, given that the frame selected as the definition
of, and remedy for, the perceived problem will determine which organi-
zations within government will receive the funds and the personnel allo-
cations associated with the program. While that conflict may be driven by
utilitarian goals, there are also genuine policy debates over the best way
to address the problem. Policy organizations are not only committed to
their self-preservation and growth, they are also committed to ideas and
to means of addressing policy problems.

The differences in possible frames for policy pose problems for policy-
making. Donald Schön and Martin Rein (1994) have argued that refram-
ing is a viable means of addressing difficult policy problems, especially
those such as drug policy involving fundamental conflict of values and
conceptions of what the policy problem may be. While in principle devel-
oping a frame for policy that could be acceptable to all participants has
the potential of producing an enduring solution to policy conflicts, it also
can be difficult to produce. If the organizations and individuals working
in the policy area had such basic disagreements about the policy then
84 ADVANCED INTRODUCTION TO PUBLIC POLICY

finding a frame that is agreeable to all is a major challenge. The Schon and
Rein discussion of reframing points out that framing is a highly political
process, and also involves a process with elements such as story-telling
and narratives, as well as the exercise of political power (Van Hulst and
Yanow, 2016).

Summary

Agenda-setting is a crucial activity for policymaking. Unless an issue

actually makes it to an active agenda it is not capable of being addressed
through the policy process. Placing the issues on an agenda may appear
rather simple, but actually may involve substantial political mobilization.
There are numerous barriers to having any issue placed on the agenda,
and those barriers may be even higher if the issue conflicts with the inter-
ests and values of economic and political elites.

But simply getting the issue onto an active agenda is only one aspect of
the political process. Another question revolves around the nature of the
issue when it actually arrives on that agenda. The process of framing the
issue shapes the politics of the issue, both at the mass level and the organ-
izational level. Political interests that want to have their views reflected in
the final policy therefore must invest time and resources in the process
of framing. Further, to produce policy change for existing programs
reframing may be required, involving some of the same components of
the initial process.

In terms of the design approach being used in this book, framing and
agenda-setting more generally are the ways in which policy problems are
interpreted so that they can be acted upon in the remaining parts of the
policy process. Whereas the previous chapter has attempted to define
these problems in more or less analytic and objective ways, what may
matter more is the way in which the issues are understood as they are pro-
cessed through the political system. Understanding the technical nature
of the issues can be important for policy analysts but in the end politics
may be trumps and the political framing of an issue may matter most for
the final policy emerging from the process.
 AGENDAS, AGENDA-SETTING AND FRAMING 85

Notes
1. Despite that, critics such as E.E. Schattschneider (1962, p. 58) argue that
“the flaw in pluralist heaven is that the heavenly chorus sings with a strong
upper-class accent”.
2. For example, corporatism and corporate pluralism in Northern Europe allow
for a large range of actors to be involved in policymaking. In this case, in con-
trast to the usual critique of pluralism, this second pillar of democracy has
tended to empower business and agricultural interests against the dominant
role of labor and the political left in the electoral institutions.
3. This also corresponds to the finding in prospect theory (Bellé, Cantarelli and
Belardinelli, 2018) that when faced with losses people tend to favor riskier,
more innovative solutions, but when presented with opportunities they tend
to accept more incremental solutions.
PART II

Policy interventions
5 Designing intervention and
implementation

Once policies are formulated, legitimated and have resources attached

to them the task of actually delivering public services has to some extent
only begun. The basic design of intervention may be contained within
the formulation, but the implementation process itself requires careful
design and consideration. Although implementation is sometimes dis-
missed as “mere administration”, in reality administration is far from
“mere” and the final impact of policies may be shaped as much by the
manner of implementation as by the formal design of the programs. If we
return to that initial definition of policy and politics as “Who Gets What”
then the final decisions about those allocations may be made through
implementation.

In addition to the direct impact that implementation has on the distri-

bution of benefits and costs among members of the public, this process
also affects the overall legitimacy and policy capacity of political systems.
The majority of contacts between the state and the society come through
the public bureaucracy during the process of implementation. These
contacts provide the public with a sense of the probity, representative-
ness (see Peters, Schröter and Von Maravić, 2013) and responsiveness
of the bureaucracy. A number of studies of the public bureaucracy have
indicated that their treatment of citizens is reflected in evaluations of the
public sector, and therefore in the legitimacy of that political system.

In design terms, implementation should be considered as a central com-

ponent of a more general intervention strategy. It is relatively simple to
say that policymakers want to alter the behavior of individuals in some
manner, or they want to shape the economy in certain ways. It is more
difficult to determine how those changes can be actually brought to
fruition. At what points in the underlying social and economic processes
is it most effective and efficient to intervene? And through what mecha-

87
88 ADVANCED INTRODUCTION TO PUBLIC POLICY

nisms? And what are the barriers to making those interventions effective?
This chapter and the next will discuss interventions designed to produce
change. This chapter will discuss implementation in general, and the
following chapter will discuss policy instruments as an important analytic
approach to the process of implementation.

Much of the discussion of implementation could be characterized as

“the horrors of war” (Linder and Peters, 1987). The study of the concept
was initiated by examining failure (Pressman and Wildavsky, 1974), and
much of the subsequent literature has also focused on failures to imple-
ment successfully. This emphasis on failure appears in other areas of
policy studies (Bovens and ‘t Hart, 1996) but has been a dominant theme
in the study of implementation. That emphasis on failure may reflect in
part the difficulties inherent in implementation, as well as an intellec-
tual perspective that demands perfect implementation. That intellectual
perspective was informed in part by a legalistic perspective that assumes
perfect compliance with law, even if that law is vague and perhaps unen-
forceable (Lane, 1983).

Although interventions and implementation remain a central compo-

nent of the policy process the academic interest in implementation has
to some extent waned. This declining interest may merely represent fad
and fashion in policy studies, but also represents some of the research
difficulties involved in identifying the effects of administration on policy
outcomes. To be able to understand why programs are not as effective
as hoped by their “formators” requires tracking through the process of
implementation, much as did the original Pressman and Wildavsky study.
Even then, any failures in the policy may not be attributable entirely to
implementation failures. The need to rely heavily on case studies and the
difficulties in assigning causation make implementation incompatible
with the canons of much of contemporary social science.1

The above being said,, there has been some revival of interest in imple-
mentation (Winter, 2011) that has addressed some of the fundamental
administrative and policy issues involved in this stage of the policy
process. For example, there has been greater investment in understand-
ing compliance with law and the associated effects on implementation
(Saetren, 2014). In addition, there has been a revival of theorizing about
street-level bureaucracy (Hupe, 2019) and with that more general con-
sideration of the discretion, and the control of discretion, among the
 DESIGNING INTERVENTION AND IMPLEMENTATION 89

lower echelons of civil servants (Maynard-Mooney and Musheno, 2003;

Brodkin, 2011).2 There has also been a concern with the effects of imple-
mentation on racial and class inequalities (Baumgartner, Epp and Shoub,
2018).

Implementation as a component of the policy process

Governments have been implementing policies for as long as they have

been in the business of governing. Few if any programs can make them-
selves perform automatically, so some means of having them achieve
their objectives must be devised. The principal actor in implementing
programs in government is the public bureaucracy. That said, however,
increasingly the public sector utilizes non-governmental actors to assist
in implementation, or even to be the primary source of implementation.
And even if the public sector does retain control over implementation it
is unlikely that any single organization is fully responsible and adminis-
tration of public programs, and therefore implementation, involves the
interaction of a number of organizations and even individual actors.

Implementation outcomes as probabilities

The multiplicity of actors involved in implementation motivated the

initial discussion of this concept by Pressman and Wildavsky (1974).
Prior to their popularizing the term “implementation” the activity of
translating programs into action would have been discussed as just plain
public administration. Pressman and Wildavsky, however, emphasized
the difficulties involved in making programs work, especially in the
multi-organizational environment that characterizes most public pro-
grams. The extensive subtitle to their book emphasized that good policy
ideas often do not go into effect as planned, and that scholars and practi-
tioners need to understand those implementation failures.

To understand the difficulties involved in implementation Pressman

and Wildavsky developed the concept of “clearance points”.3 These were
90 ADVANCED INTRODUCTION TO PUBLIC POLICY

points along the process of implementation at which a positive decision

was required for the process to continue as expected by the “formators” of
the policy. This analysis indicates just how complex implementation can
be, especially in a federal country such as the United States (see also Hanf
and Toonen, 1985; Doonan, 2013). In their study of the implementation
of an economic development program developed in the federal govern-
ment in the San Francisco bay area, Pressman and Wildavsky identified
over one hundred clearance points at which the implementation process
could have been stalled and killed, whether intentionally or simply by
lethargy or poor administration.

The problem presented to policymakers by the large number of clearance

points in programs such as that program discussed by Pressman and
Wildavsky is that one of these may be sufficient to derail a program. Even
if actors at each of these clearance points are in fact committed to making
the program work and are skilled administrators there is still some possi-
bility that implementation will not be successful. Even if the probabilities
of successful implementation at each clearance point is 0.99 then the
probability of successfully implementing through 100 points is less than
0.00001. Implementation therefore requires active political involvement
if the program adopted by some organization in government is actually to
go into effect in the field. Thus, policy entrepreneurs may be as important
in implementation as they are in policy formulation (Ridde, 2009).

The need for political involvement in pushing through the implementa-

tion process raises the question of just how to alter the behavior of actors
involved in making these decisions. Elinor Bowen (1982) began with the
fundamental conception about implementation used by Pressman and
Wildavsky and then developed ideas about how the active implementer
could overcome some of the difficulties identified in the original research
on implementation. Rather than accepting passively that there is a chance
of failure at each stage, the would-be implementer can affect the process
positively and increase the possibilities for success.

For example, that active implementer could attempt to create a band-

wagon effect, using success at some clearance points to create momentum
that could be used to spur other actors to accept the program. That imple-
menter can also be persistent, and not take “no” as the final answer when
implementation fails at one point. That implementer can also attempt
to skip over potential bottlenecks in the process to place pressure on the
 DESIGNING INTERVENTION AND IMPLEMENTATION 91

recalcitrant actors. There are other tactics that can be used but the basic
point is that any simple, passive conception of implementation does not
take into account the possibilities for moving the process forward even in
face of apparent opposition.4 This argument also emphasizes the extent
to which implementation, and other aspects of public administration, are
fundamentally political activities.

Forward and backward mapping for design

The initial discussion of implementation was conducted as if this compo-

nent of the political process were simply an exercise in probability theory.
This perspective involved relatively little sense of design but instead
tended to take both the political structure and the process of implementa-
tion largely as a given. While that analysis provides a background against
which to compare the realities of administering policies, it does little to
help policymakers who may be attempting to develop a strategy for inter-
vention and implementation.

A second stage in the development of thinking about implementation

represented more of an attempt to assist would-be implementers in build-
ing more effective systems for putting their programs into effect. Most
thinking about making public policies assumes that the formulators of the
policy decide what they want to do, and then attempt to put those poli-
cies into effect. That image of policymaking to some extent conforms to
a democratic vision of governance in which elected officials make policies
and then work to implement those policies.5

One proposal for designing programs for enhanced implementation was

discussed as “backward mapping” (Elmore, 1985). To some extent this
proposal involves taking the idea of clearance points from Pressman and
Wildavsky and thinking what can be done in the design of programs to
minimize the negative impact of those constraints on implementation.
The basic idea is to consider what the actors implementing a prospective
program prefer and designing the program around those preferences. The
fundamental goals of the program would be driven by perceived needs to
intervene, but the structure of that intervention might be shaped more
92 ADVANCED INTRODUCTION TO PUBLIC POLICY

by the preferences of the implementers, and the preferences of the target

population.

From a design perspective on public policy backward mapping makes

a great deal of sense. Rather than thinking about the formulation of
a program in an administrative vacuum, backward mapping considers
the likelihood of success of an intervention. This strategy for shaping pro-
grams at least in part on the basis of the probabilities of implementation
success may increase the overall success of the programs. In addition, this
perspective on implementation corresponds well with the more evolu-
tionary discussion of policy, assuming that firm designs are difficult to
achieve within the complexities of the political world, and the even more
complex social and economic systems. Therefore, this potential linking of
formulation with the targets through considering feasibility may open the
way for more adaptive policy.

Although seemingly desirable from the design perspective, backward

mapping also tends to imply that policies would be driven by feasibility
rather than by policy priorities. It is not difficult to argue that implemen-
tation is difficult and therefore finding means of reducing friction with
both clients and implementers can be desirable. But as feasibility tends
to dominate real policy preferences then the likelihood of producing
effective programs and producing real changes in the target popula-
tion is reduced. And feasibility, or infeasibility, also can be politically
constructed to affect the probabilities of a program being adopted (see
Knoepfel, 2018).

Multiple actors: implementation structures and

networks

As already discussed, almost any implementation effort will involve

a number of actors. This observation is to some extent a defining feature
of implementation studies, but it is also the foundation for several addi-
tional theoretical approaches to implementation. For the two approaches
to implementation discussed here the fundamental question is under-
standing how the interactions among actors involved in implementation
occur and how they shape the outcomes of the administrative process.
 DESIGNING INTERVENTION AND IMPLEMENTATION 93

The more basic of these approaches is concerned with “implementation

structures” as the defining feature of the process of putting programs
into effect. Bennie Hjern and David Porter (1981) argued very early in
the study of implementation that “the single lonely organization” is dead
and that it had been replaced by structures involving a number of actors.
The death of that single organization may be exaggerated, and there are
certainly cases in which implementation is confined to one organization,6
but implementation tends to involve increasing numbers of actors.

While the large number of actors involved in many attempts at imple-

mentation may create problems, as indicated by Pressman and Wildavsky
and all implementation scholars since, it is also possible to exaggerate
those difficulties. Bowen pointed to several means that implementers
could employ to avoid passive acceptance of the impacts of clearance
points, and other students of implementation have also pointed to means
of pushing the process forward even against opposition (Hill and Hupe,
2014).

The potential problems created by multiple actors in implementation

are also mitigated in part because these actors are not drawn at random
at the outset of each new program, but rather represent institutionalized
patterns of interaction (Peters, 2014b). The organizations and other
actors involved in implementing social programs, or any other type of
program, in a community tend to work together for an extended period
of time and often work together on a number of different programs. They
therefore develop routinized patterns of interaction and generally develop
trust among themselves. Both routines and trust reduce transaction costs
among the participants (Calista, 1994) and make implementation proceed
more smoothly. This institutionalization does not mean that stability and
comity will be preserved across all programs, but it does serve as a good
starting point for implementing programs.

Conceptualizing multi-actor implementation as networks is yet another

variation in this general theme (O’Toole, 2011). Network theory has
become increasingly important in the study of public policy and gov-
ernance (see Sørenson and Torfing, 2007), and this approach has been
applied to implementation in a number of settings. Further, network
thinking about implementation tends to integrate some aspects of the
backward mapping approach as network actors tend to be involved in the
formulation of policies as well as in their implementation. Therefore, the
94 ADVANCED INTRODUCTION TO PUBLIC POLICY

members of the networks to some extent are designing policies in ways

that will help themselves as actors involved in the implementation of
those programs. Further, network actors may be able to work across con-
ventional boundaries of policy domains to help overcome coordination
difficulties (McGee and Jones, 2019).

Network implementation implies a number of actors working together to

implement a program. The assumption in these models is that networks
are composed of relatively equal actors who bargain among themselves
to deliver the program in question. In the other models, even that of
implementation structures, there is an assumption that the process is
driven by government actors. In the extreme, non-governmental actors
are conceptualized as much as potential barriers to the process as they are
partners in producing the program. In less extreme views, the social and
market actors are partners involved in implementing programs designed
primarily within the public sector.

Summary

Beginning with the Pressman and Wildavsky study and moving through
the other approaches to implementation, some common strands of
analysis appear. Perhaps the most important of these is that implemen-
tation is a “game” with a large number of players (see Bardach, 1977),
and understanding the interactions among the participants is crucial for
understanding the outcomes. Further, this is a political process as well as
an administrative process. The politics involved may be organizational
rather than partisan, but they are still politics. And finally, because of the
first two characteristics, simple legalistic conceptions of implementation
are inadequate for understanding what actually happens.

Gaining compliance

Much of the discussion of implementation is done in terms of laws and

the use of political and legal instruments to achieve policy goals. There is,
 DESIGNING INTERVENTION AND IMPLEMENTATION 95

however, also an important psychological element involved in implemen-

tation – the need for compliance by citizens and organizations. While the
targets for policies may be threatened, coerced or cajoled into compliance
with a law, it is much easier for governments attempting to implement the
law if they can gain willing acceptance and compliance.

As much as being a characteristic of the policies in question, gaining

compliance may be a function of the legitimacy of the programs and of
government more generally. Some policy instruments – those relying on
information and persuasion – are heavily dependent on the legitimacy of
government, or perhaps on the legitimacy of the private sector organiza-
tions actually delivering the programs (see Chapter 6). For those instru-
ments to gain compliance the organizations attempting the persuasion
must be legitimate and accepted by the targets of the policy.

It is also important for program designers to realize that compliance

through using rigid enforcement and hierarchy may produce the desired
outcome, but it can also be extremely expensive. That expense is not only
in terms of the actual expenditure of resources in the public sector but also
in terms of the potential loss of trust by citizens. Citizens may resent being
forced into accepting programs and using more coercive instruments may
make implementing other programs in the future even more difficult to
implement.

Drift and sabotage

Even if the process of implementation does go forward, it is likely that the

policy being implemented will to some extent be altered in the process.
As already noted, the real policy is the policy that is implemented, and
the complexity of the structures involved in the process tend to produce
deviations from the intentions of the framers. At each of those clearance
points decisions are being made and even if the actors involved are acting
in good faith they may not interpret the intentions of the lawmakers cor-
rectly, or they are often faced with political pressures

The problem of multiple actors in the system for implementation can

be seen through the lens of delegation and discretion. When laws are
96 ADVANCED INTRODUCTION TO PUBLIC POLICY

made and then are implemented through administrative organizations

or non-governmental actors those implementing organizations are given
discretion to make decisions. At the extreme, administrative organiza-
tions make many more rules than do legislatures (Kerwin and Furlong,
2017). And the implementation of programs is delegated to lower levels
of public organizations as well as to quasi-autonomous agencies within
government (Pollitt and Talbot, 2004).

The delegation to these organizations can leverage resources but may also
produce difficulties in implementing programs in the manner desired by
the legislature or other sources of legislation. The implementing organi-
zations – whether within the public sector or in the private sector – will
have their own values and their own priorities, and therefore will interpret
legislation in ways that may not have been intended. One clear example of
this was the Environmental Protection Agency (EPA) in the United States
interpreting its mandate on clean air to include carbon dioxide, hence
providing a backdoor method for the Obama administration to address
the issue of climate change (Wald, 2012).

The issue of bureaucratic “drift” in implementation, especially the imple-

mentation of regulatory policy, has been addressed by scholars advocat-
ing the development of “tamper-proof legal instruments” that can insure
that the program is implemented in the way intended (see McCubbins,
Noll and Weingast, 1989). These attempts to limit discretion represent
the continuing conflict within government over implementing programs
as intended by lawmakers and building in sufficient flexibility and
discretion so that programs can adapt to changing circumstances. This
is, of course, in part a political as well as an administrative battle – con-
servatives hated the use of discretion by the EPA mentioned above while
liberals applauded it.

But even “tamper-proof” instruments may be little value if the public

bureaucracy is opposed to the programs they are implementing and
choose to shirk, or to sabotage, a program (Brehm and Gates, 2002). The
issue of sabotage became more relevant as populist regimes were elected
and confronted public bureaucracies that had their own ideas about
policy, and tended to regard the populist leaders as outsiders with little or
no understanding of the complexities of public policy (see Guedes-Neto
and Peters, 2021).
 DESIGNING INTERVENTION AND IMPLEMENTATION 97

Barriers to implementation

The complexity and the attenuation of the implementation process are

themselves fundamental barriers to effective implementation, but there
are other factors in the process that can reduce even more the probability
that legislation will be implemented as intended. These factors may to
some extent be overcome through careful planning of the implemen-
tation strategy, but some may simply constitute problems that must be
understood. Implementation is rarely perfect (see Hood, 1986) and those
failures of implementation are often major cause for failures in the policy
process considered more broadly (McConnell, 2010).

Legislation
The legislation itself often constitutes a major barrier to effective imple-
mentation. Legislation almost inevitably represents compromises of polit-
ical forces, and those compromises often produce vague legislation that
cannot be implemented readily. Perhaps more accurately, the legislation
may be so vague that it is difficult to determine if indeed it has been
implemented effectively or not. Legislation often contains lofty goals
necessary to have it passed, and useful for appealing to the voting public,
but that rhetoric may be of little use to a civil servant attempting to put
the program into effect.

In addition to the rhetorical flourishes that may be contained in legis-

lation, legislation may also create extremely complex implementation
structures. Perhaps the most egregious example of this complexity is
the Affordable Care Act (“Obamacare”) in the United States. This act
was over 1,000 pages long and involved a variety of measures designed
to alter the manner in which health care is financed and delivered. Even
the centerpiece of the act, the insurance exchanges, involved complex
interactions between private insurers, potential purchasers and state (and
the federal) governments. It is perhaps not surprising that the first few
months of the program were, to say the least, trying (Clemmitt, 2014).

Finally, contrary to the logic of backward mapping, legislation may be

written in ways that do not recognize the strengths and preferences
of the organizations that will implement it. I have been making the
point throughout this book that organizations within government are
98 ADVANCED INTRODUCTION TO PUBLIC POLICY

crucial for the success of policies, and failure to take the preferences of
those organizations into account can readily produce implementation
failures. These difficulties in implementation may be manifested when
there are sharp ideological differences between members of the imple-
menting organization and the formulator,7 but may occur simply when
the designers of programs do not recognize the organizational routines
and preferences for particular instruments that exist within most public
organizations.

Politics
In addition to the coalition politics involved in the creation of legisla-
tion, other aspects of politics also make implementation difficult and
sometimes move the program as implemented away from the intentions
of the formulators of the program. These politically inspired devia-
tions are most likely to appear in presidential and federal systems. In
parliamentary systems the legislature and the executive responsible for
implementation will almost certainly be aligned politically, whereas in
presidential systems those institutions may have conflicting ideas about
policies and their implementation. Further, most parliamentary systems
are consensus systems (Lijphart, 1984) so that the policy change across
governments tends to be less than in presidential systems that function
more as majoritarian systems.

Inter-governmental relations may present a particular challenge for effec-

tive implementation in federal regimes (Stoker, 1991). The constitutional
autonomy of sub-national units in federal systems, and the concurrent
jurisdiction of levels of government over more policy areas, means that
the central government is generally negotiating rather than mandating
policies when working with sub-national governments.8 However, imple-
mentation within systems of “multi-level governance” can be a challenge
in any regime (Hooghe and Marks, 2003). Even if lower levels of gov-
ernment are controlled by the same political party they may still have
different priorities from the central government and therefore may
implement programs differently than intended. And if there are different
political parties involved across the levels of government the threats to
implementation are exacerbated. Conflicts between Democratic gover-
nors and the Trump administration over a range of policies such as the
environment, immigration and health care illustrate the importance of
politics in implementation.
 DESIGNING INTERVENTION AND IMPLEMENTATION 99

Multiple actors
As governments begin to utilize more and more third-party actors to
assist in implementation, or to take full responsibility for the implemen-
tation of public programs, potential barriers to effective implementation
arise. Just as sub-national governments may have their own priorities
and therefore attempt to implement public programs in their own ways,
so too do third-party implementers – market or non-market – also have
individual preferences and priorities. Although there has been a great deal
of enthusiasm concerning the involvement of networks and other social
actors in implementation (O’Toole, 2000), this form of implementation
can also present difficulties.

Although much of the discussion of multiple actors is based on poten-

tial problems with organizations in the private sector, reforms in the
public sector have tended to create more problems of this sort within
the public sector itself. Part of the logic of the New Public Management,
beginning in the 1980s, is that efficiency and accountability in the public
sector can be enhanced by dividing large public organizations into
multiple single-purpose organizations usually called agencies (Laegreid
and Verhoest, 2010). While these organizations may at times increase
the efficiency of administration, their autonomy may also provide them
more opportunities to make their own decisions about the interpretation
of legislation. Indeed, their pursuit of efficiency in their own terms may
undermine their willingness to pursue.

Although we can think of the difficulties created by this multiplicity of

organizations involved in implementation as purely an administrative
issue, it can also have significant policy consequences. First, although
usually discussed as leveraging private actors to improve policy, the
involvement of multiple actors may create conflicts that in the end may
lessen the success of a program. Further, implementation involves making
as well as strictly putting a policy into effect, so that the intentions of the
legislation may be altered when being implemented by individuals and
groups with perhaps different policy values. While some of this same
policy drift may occur when implementation is entirely within the public
sector the drift may be less controllable when being produced by private
actors.
100 ADVANCED INTRODUCTION TO PUBLIC POLICY

Performance standards
Somewhat paradoxically the increased use of performance management
as a part of the reforms associated with the New Public Management may
actually reduce the capacity to implement programs as intended by the
formulators of the policies. The basic logic of performance management
is that by establishing standards and targets for the organizations and
individuals responsible for implementation that implementation will be
improved (Bouckaert and Halligan, 2007). While that goal is well inten-
tioned, it may set up situations in which the quality of programs provided
actually diminish.

The difficulty in performance management is that when people have to

reach targets they will find ways to reach them, regardless of the actual
effects on the intentions of the program. For example, if the police have
a performance standard for clearing a certain percentage of reported
crimes they may simply refuse to record some crimes that may be difficult
to solve (Barrett, 2014). Or they may reclassify crimes as less serious so
they can demonstrate that rates of serious crimes such as homicide are
falling (Eterno and Silverman, 2012). And it is certainly not only the
police who find ways to reach targets for implementation without actually
implementing the programs as intended.

Street-level bureaucrats and implementation

Above we discussed some of the potential problems of drift in the

implementation of public policies in general terms. Although there are
a number of potential influences on implementation, the impact of the
lower levels of the bureaucracy represents an especially important influ-
ence on the decisions that are finally made by governments. This influence
highlights the point made above that the real policy of government (and
its collaborators in the private sector) is the policy as it is implemented.

Despite attempts by legislatures and others who make law, discre-

tion is always necessary in implementation (Thomann, van Engen and
Tummers, 2018). Indeed, legislatures may not want to write laws that
are too precise and too detailed. They both lack the expertise in many (if
 DESIGNING INTERVENTION AND IMPLEMENTATION 101

not most) areas of policy but also are faced with social and technological
circumstances that often change rapidly. Drug law enforcement makes
this point rather clearly. Most countries outlaw the use of certain drugs,
and place them on schedules of differing degrees of severity for punish-
ment. This worked reasonably well until drug dealers found ways to alter
a chemical compound slightly, making it outside the law but having the
same effects as the illegal substance (McElroy, 2013). Lacking discretion
to confiscate any drug that appears harmful, enforcers must wait for new
law.

The impact of the lower echelons of bureaucracy – the “street-level

bureaucrats” – on policy is apparent in social policy and in regulatory
policy. The term was coined by Michael Lipsky (2010) primarily in rela-
tion to social policy. He was concerned with the extent to which workers
in social service organizations could bend the rules either positively or
negatively and therefore advantage or disadvantage clients.9 His research,
and a good deal of other research on the powers even of receptionists in
social service organizations, indicates the extent to which clients may be
vulnerable, and programs vulnerable to sabotage.

The lower echelon of workers in regulatory organizations also have

substantial discretion in making decisions, and therefore a good deal of
power. Research in a number of settings (Lundqvist, 1980; Bardach and
Kagan, 1982, 2009) has pointed out that inspectors have a good deal of
discretion in how they administer environmental and health and safety
laws. Further, exercising that discretion is often key to the success of
programs, as excessive rigidity in the rules may lead to inspectors ignor-
ing violations that could have disastrous consequences for a regulated
industry. The capacity to negotiate enforcement can lead to more effective
policy implementation,10 although it may also result in deviations from
the intentions of the formulators of the policy.

The role of police as “street-level bureaucrats” has become very obvious

in the early 2020s. Especially in the United States there is evidence of the
police differentially enforcing laws against minority groups, and in using
violence against those groups (Baumgartner, Epp and Shoub, 2018). But
it is not just in the United States that there has been evidence of the use of
discretion by police to the detriment of minority groups (Breeden, 2020).
102 ADVANCED INTRODUCTION TO PUBLIC POLICY

Just as the New Public Management mentioned above has tended

to enhance the autonomy of agencies delivering public services, the
“empowerment” strand of reform in the public sector (see Peters, 2001a;
Fernandez and Moldogaziev, 2010) has tended to give lower-level officials
within organizations of all sorts greater autonomy to make their own
decisions. Thus, a great deal of the change within the public sector over
the past several decades has tended to delegate more and therefore to
make implementation more difficult for legislatures and political execu-
tives to control. This in turn means that policy outcomes for citizens are
likely to be more variable, raising significant questions of accountability
and even of democracy.

What is implementation success?

All the above discussions of implementation have discussed the process of

implementation, and many, if not most, of these have described the out-
comes of that process as failures. Indeed, most studies of implementation
have emphasized failures and discussed how and why implementation
fails. While those failures are often obvious, in many other cases they may
be in the eye of the beholder. The question of implementation failure is
important for understanding policy success and failure more generally.

Implementation failure is not just a simple empirical question, but rather

the failure is related to the numerous theoretical approaches to imple-
mentation. If one adopts the top-down perspective then failure can be
seen as a policy outcome that does not conform to the intentions of the
initial policy formulation. If, however, one has a more flexible, bottom-up
perspective then failure is more difficult to define, and also less likely
to occur. These two fundamental theoretical approaches then are also
potentially related to different approaches to measurement, in degree if
not in type.

These alternative conceptions of success in implementation can be

related to the fundamental issue of policy design. The more constrained
top-down conception of implementation success is derived from the
more rigid conception of policy design, assuming that initial design
perspectives will be carried through the process. In such a view imple-
 DESIGNING INTERVENTION AND IMPLEMENTATION 103

mentation failure implies a failure of the initial design. On the other hand,
if one adopts a more argumentative perspective then the program may
evolve more as it is implemented (but see Browne and Wildavsky, 1984).
Rather than being set rigidly by decisions at the outset the policy inevita-
bly adapts, with bottom-up implementation representing one dimension
of that adaptation.

From either perspective, there is still not a clear measure of what would
constitute failure. Even in the top-down perspective it is not clear how
much drift constitutes failure. Any demand for perfect implementation of
a policy would appear unreasonable, given the complexity of the process
and the difficulties in affecting large-scale social and economic processes.
On the other hand, simply saying that something happened, and perhaps
something good happened, seems too weak a criterion to argue that
implementation has been successful. Further, does one attribute success
or failure to the implementation process or to the initial policy design?
These difficulties then constitute a useful link to one of the later sections
of this book – the role of evaluation in the policy process.

Interventions and implementation

The title for this chapter includes the word “intervention”, although
almost all the discussion until this point concerns implementation.
Thinking about how to intervene effectively into ongoing social and
economic processes involves design elements such as policy instruments,
but may also require thinking about broader implementation strategies as
well. These intervention strategies to a great extent mirror the implemen-
tation literature, with three broad conceptions about how to put policies
into action in their socio-economic environment, and indeed within the
political system itself.

The first and dominant strategy remains one depending on law and
formal hierarchy to achieve policy goals. The foundation of the imple-
mentation literature is an assumption (unstated at times) that laws
should, and could, be implemented perfectly with the intentions of the
framers of the legislation going into effect. Implementation in this context
is assumed to be done primarily or totally by the public sector itself. The
104 ADVANCED INTRODUCTION TO PUBLIC POLICY

policy might still fail, but it would not fail because of implementation
deficiencies. Especially in administrative and political systems that are
highly legalistic, this conception of implementation persists, and the
probability of perceived implementation failure is consequently greater
(Hammerschmid and Meyer, 2004).

The second significant strategy for implementation emphasizes an under-

standing of the context into which a policy is being introduced. The fun-
damental idea of backward mapping is to understand the environment of
policy and then design programs that will be most acceptable within that
environment. While this strategy may lead to greater implementation
success, there is a clear question of whether it can produce policy success.
That is, if the program is designed for easy implementation and for
acceptability, it may not be capable of addressing the underlying prob-
lems it was meant to solve.

The backward-mapping approach to intervention also points to the

potential danger of the term “feasibility” in policy analysis (Meltsner,
1972; Vanderbroght and Yamasaki, 2004). One almost sure way to defeat
a policy proposal is to say that it is not feasible. That judgment of a lack
of feasibility may be passed without any great analytic support, or it may
be expressed in political rather than analytic terms, but in either case it is
a rather facile means of defeating a program. While backward mapping
may still lead to the adoption of a program, that program may be so
watered-down that its success may be relatively insignificant.

The third approach to implementation also is concerned with context

and the socio-economic environment of policy, but tends to use the social
actors rather than to have policy necessarily shaped by them. The use of
horizontal patterns of implementation involving networks and market
actors may facilitate policy delivery, if not have policies shaped by ex ante
by those socio-economic forces. Of course during implementation the
social and market actors will almost certainly influence the nature of the
policy as it actually goes into effect “on the ground”. We should in fact
expect more drift from the intentions of the framers of the legislation,
an outcome that can be compensated for by the ability to leverage the
resources of the non-governmental actors involved.

These implementation strategies may also be seen as broad strategies for

intervention, and for policy design more generally. One version of poli-
 DESIGNING INTERVENTION AND IMPLEMENTATION 105

cymaking would be an étatiste form of governing almost entirely through

public sector institutions, using law and hierarchy for control over the
society. An alternative could be seen as permitting self-regulation of the
society with very loose guidance based on guidelines, frameworks, and so
on. And finally policy may be pursued through collaboration with social
and market actors, attempting to direct those actors in the direction of
legislative and administrative goals, but also recognizing the importance
of the interactions between state and society.

However governments and their allies attempt to govern and to make and
implement policy, it is not an easy task. The implementation literature
has emphasized those difficulties and the significant probability of failure.
That sense of failure, however, may be a product of high expectations and
of a legalistic frame of reference that assumes that there should be little
deviation in the implemented programs from the formal law. But we are
still left with little guidance in understanding what is adequate implemen-
tation, and when failure has indeed occurred. And further, despite the
numerous failures in perfect implementation, there are still important
successes and many policies do work.

Summary

If public policies are to mean anything they must be implemented.

That is, however, easier said than done, and the history of the study of
implementation is built primarily on the study of failure. That emphasis
on failure is, however, perhaps excessively negative (see Compton and ‘t
Hart, 2019). Governments have been able to make programs work, and
sometimes work very well, although few if any programs are implemented
in exactly the manner in which they were intended. Thus, while failure
has helped to bring scholarly attention to the issue of implementation in
reality the outcomes are much more mixed.

The study of implementation also emphasizes the complexity of public

policy. Whether the actors involved are public, private or a combination,
almost all implementation involves multiple actors interacting in a variety
of ways for the program to function. Recognizing the existence of these
multiple actors in the process emphasizes the importance of discretion
106 ADVANCED INTRODUCTION TO PUBLIC POLICY

in implementation, and the extent to which policies may change during

implementation as a result of that delegation. That change does not mean
failure, but does invoke the need to evaluate what has happened and to
use that evaluation to continue improving policy.

Notes
1. Process-tracing methodology facilities the analysis of implementation to
understand better when and how processes fail. See Beach and Pedersen
(2013).
2. This concern about discretion was in part sparked by administrative reforms
during the latter part of the twentieth century that encouraged delegation
and empowerment for lower echelon workers in government.
3. This concept is similar to the concept of “veto points” in institutional theory
developed by Tsebelis (2000).
4. Ernst Alexander (1989) argues that Bowen’s optimism is founded on a set of
assumptions about the malleability of the implementation process that may
be applicable in only a limited number of circumstances. His argument, and
hers, are indicative of marked contrasts in views about the formality of public
administration.
5. It in some ways conforms even more to patterns of policymaking in
less democratic regimes in which the hegemonic party, or the individual
hegemon, will expect his or her policies to be implemented without question.
6. For example, most tax legislation is implemented by a single powerful
organization rather than by multiple, dispersed organizations. That said,
the private industry of accountants, lawyers and advisors that attempts to
lessen the tax liabilities of individuals and corporations must be considered
participants in the process.
7. For example, the Environmental Protection Agency in the United States
pushed back rather hard against the intentions of several Republican presi-
dents to weaken environmental programs.
8. The level of control of the center does vary significantly within federal
regimes (see Fenna and Hueglin, 2010).
9. For a very negative conception of this role of social service workers see Piven
and Cloward (1993).
10. This acceptance of negotiation also becomes crucial for writing regulations,
as “soft law” becomes more widely accepted. See Mörth (2004) and Dahl and
Hansen (2006).
6 Policy instruments

The preceding chapter discussed implementation and intervention in

a relatively general manner, but we now need to consider more carefully
the means through which that implementation is conducted. Public
programs are in essence designed using policy instruments, or tools, for
implementation. These tools (see Hood, 1986; Salamon, 2001a; Hood and
Margetts, 2009) are necessary to take the intentions of legislators or other
“formators” of policy and translate them into effective action.

This point about the central role of policy instruments appears rather
simple, but behind that apparent simplicity lies a good deal of complexity.
Policy instruments are not simple, mechanical means of intervention, but
have political impacts of their own. Further, they can be evaluated along
a number of different dimensions and there may be little agreement on
the desirability of one over another in specific policy situations. Finally,
instruments themselves involve a number of different underlying mecha-
nisms for affecting the society, and these need to be understood in order
to make informed choices about instruments.

Perhaps even more basically, to be successful policy instruments require

compliance from the members of the society. In some societies that is
relatively easy, and governments are able to use relatively “soft” policy
instruments because they can be sure the public will obey laws. In other
societies, however, direct command and control, and direct enforcement,
may be required to get the public to obey laws (Salamon, 2001b; Héritier
and Rhodes, 2012). And of course the targets of the laws being imposed
may affect these choices, with most people readily complying with laws
against murder but relatively fewer citizens being willing to obey speed
laws.

For policy designers there is some tendency to think about choosing

“the” right instrument to address a problem. That strategy may not be as

107
108 ADVANCED INTRODUCTION TO PUBLIC POLICY

productive as the designer might like, and there is a need to think about
using multiple instruments to achieve the goals for a program (see Weber,
Driessen and Runhaar, 2014). In particular, it may be valuable to mix con-
ventional “command and control” instruments with softer instruments in
an attempt to gain compliance through normative or collaborative means.

This chapter will discuss the role that policy instruments play in imple-
menting policy, and how they relate to more general designs for policy
interventions. This discussion will require first categorizing the instru-
ments, then discussing how to select them for utilization in a policy
design, and then evaluating them. The selection of instruments is a func-
tion of a number of factors, which emphasize the political dimensions
of instrument selection. That evaluation is multi-faceted, so that instru-
ments that are evaluated well on one dimension may be rated poorly
on others. Further, that evaluation of instruments is contingent, so an
instrument that may be evaluated positively in some circumstances may
be considered negatively in others.

I am discussing policy instruments in the context of implementation,

but we could also address them as a major subject of policy formulation.
Policy designers have to select not only the more general nature of the
program but also the individual instruments that are used to achieve the
goals of the program (James and Jorgensen, 2009). As described earlier
when discussing agenda-setting, excluding options from consideration
can be a significant means of controlling the outcomes of the selection
process. In this case, however, the principal actors involved may be policy
advisors and public administrators who present policy options, and
instrument options, to the political leaders responsible for the choices.
Thus, the exclusion of instruments may bias policy outcomes just as much
as the exclusion of issues at the agenda-setting stage.

Classifying policy instruments

Governments have a very large tool chest at their disposal as they attempt
to influence the society and economy. These tools are not, however, each
sui generis but rather fall into a number of categories. These categories
are useful not just for academic discussions of public policy but also for
 POLICY INSTRUMENTS 109

making real policy choices. In addition, the categories of instruments

help in understanding the underlying nature of the instruments and their
linkages with other instruments. Further, it helps both to understand the
effectiveness of instruments, as well as to understand the politics involved
in their selection. Again, these instruments are not neutral tools like
a hammer but do have their own political economies.

In this discussion of policy instruments we will need to remember that the

tools are to some extent empty vessels into which a good deal of additional
policy content must be poured. For example, we can say that government
is using a tax instrument to achieve some policy purpose. That is helpful,
but is it an income tax or an expenditure tax? And if it is an expenditure
tax is it a general imposition or an excise tax affecting only a limited
number of products? And is that excise tax levied at the point of sale or
earlier in the distribution process (like the value-added tax) so that the
consumer is not necessarily aware of the scale of the tax being charged?
The fine print on an instrument is important, so as we discuss a range of
instruments (Table 6.1) it is important to remain cognizant of the internal
variations.

The first significant attempt at classification, or at least enumeration,

of the policy instruments available to government was supplied by E.S.
Kirschen and his colleagues (1964). This effort listed 64 alternative policy
instruments, all considered options for economic policy. Thus, even
within one policy domain, albeit an important one, there is a very large
number of options for the public sector to intervene. This enumeration is
perhaps too extensive, but it does demonstrate the wide array of options
for governments, and these were indeed for governments and did not
include many alternative forms of intervention.

While the simple enumeration of policy instruments may not appear to

advance the cause of policy analysis, it is still important to understand just
what options are available to decision-makers who want to design a policy
program. Table 6.1 provides a list of instruments arranged within some
broad categories. This listing contains the principal policy instruments
but each of these might be further differentiated. In short, there is no
shortage of ways for governments to intervene, so we must now consider
how policy designers should choose their instruments when they formu-
late policy.
110 ADVANCED INTRODUCTION TO PUBLIC POLICY

Political science approaches to instruments

For political scientists and public administration scholars Christopher

Hood’s book The Tools of Government (1986) represented the first signif-
icant discussion of instruments and their role in policy implementation.
Rather than begin with a simple enumeration of policy instruments,
Hood began by considering the more basic resources available to govern-
ments. He classified these using the acronym NATO, meaning Nodality
(Information), Authority, Treasure and Organization. In other words,
governments can use information, legal authority, money and people to
influence their surrounding society.

As well as these four categories of the resources available to governments,

Hood argued that instruments could be considered as “detectors” or
“effectors”. We tend to think of policy instruments primarily as produc-
ing change in the environment (effectors), but they also can be used to
detect changes in that environment. For example, government personnel
are in touch with their clients or are patrolling neighborhoods (police)
and hence know what is happening within the society that may require
intervention. This role of detection is perhaps as important for instru-
ments as is producing change, because if government is blind to environ-
mental change it is unlikely to make good policy decisions.

The Hood taxonomy of policy instruments, or perhaps more precisely the

resources of government, also makes it obvious that the majority of the
programs actually used by government are hybrids, involving more than
one of those basic resources. For example, tax expenditures are obviously
dependent on Treasure, but also require monitoring (Nodality). They,
and any tax, depend on the legal authority of government, and finally

Table 6.1 Examples of types of policy instruments

Economic Legal Persuasive Other

Grants Regulation Information Monitoring

Subsidies Contracts “Nudges”

Taxes

Tax Expenditures
 POLICY INSTRUMENTS 111

tax authorities (Organizations) monitor and perhaps directly implement

those laws. Thus, although the resources of government can be separated
analytically, and that separation is useful for the analyst, in practice indi-
vidual tools involve some or all of those resources.

Hood’s four categories of tools, or resources, was the beginning for other
categorizations of policy instruments in political science. One classifi-
cation (Bemelmans-Videc, Rist and Vedung, 1998) described instru-
ments as “Carrots, Sticks, and Sermons”. An even simpler categorization
(Gormley, 1989) was that instruments can be understood as “Muscles
and Prayers”. The latter dichotomy makes the point that government can
either choose to exercise its power, or it can merely rely on moral suasion
to attempt to gain compliance. The former classification differentiates the
use of government power into incentives and disincentives, but at times
the only option is to preach at the public.

Peter Knoepfel (2018) has a broader classification of the resources avail-

able to government to produce change in the society (Table 6.2). The ten
resources Knoepfel mentions have some similarities to those in Hood’s
classification, but also contain some that are not included. For example,
this classification contains some political aspects of policy such as consen-
sus building and political support, and the use of time to force decisions,
But even with the difference the basic idea that government has a number
of basic resources to influence society remains.

The above classifications of instruments emphasize the mechanisms of

intervention, but instruments can also be classified according to their
political and social characteristics. In particular, some scholars have
argued that the most appropriate way to understand instruments is
through the level of legitimate coercion they impose on actors in the
society. Their argument (Macdonald, 2001) is that less coercive instru-
ments are, everything else being equal, more acceptable in liberal societies
than are the more coercive. In this context instruments can be classified
from self-regulation being at one end of the continuum and direct gov-
ernment provision of a service at the other.

Yet another classification system (McDonnell and Elmore, 1987) con-

tains four major categories – mandates, inducements, capacity-building
and system-changing. These two scholars were working primarily in
112 ADVANCED INTRODUCTION TO PUBLIC POLICY

education policy, but their scheme has relevance beyond that one area.
Perhaps the most significant element in this analysis is the difference
made between individuals as the targets of the instruments and institu-
tions or systems as the focus of change. This analysis makes the point that
although we tend to think about changing individual behavior or the out-
comes for individuals with policy, the most efficient way of doing so may
be to address deficiencies at the systemic level. This emphasis on process
and the environment of decisions is echoed in part in Michael Howlett’s
(2001) concept of procedural policy instruments. His argument is that
each of the four categories in the NATO scheme has procedural as well as
substantive elements, and indeed each of those can be further divided into
positive and negative.

Finally, Lester Salamon (2001b) has labeled tools as simply “old” and
“new”. This is not as simple a classification as it appears, given that
what is meant by old in this case are instruments that depend heavily on
command and control. The new policy instruments, on the other hand,
depend more on negotiation and collaboration. These new instruments
are, in other terms, “soft law” rather than hard law (Mörth, 2004). As with
the classification based on coerciveness described above, the assumption
is that the degree of intrusiveness and control is central to understanding
policy instruments, and further contemporary politics and society are

Table 6.2 Resources available to government

Hood Knoepfel

Nodality Information

Authority Law

Treasure Money

Organization Organization

Personnel

Force
Property

Consensus

Time

Political Support
 POLICY INSTRUMENTS 113

more resistant to the direct use of government power than was true in the
past.1

Constructivist perspectives on instruments

In addition to the classification schemes based on more or less objective

characteristics of policy instruments, other scholars have argued that
instruments are best understood in constructivist terms (see Bressers and
Klok, 1988). That is, most categorizations of instruments assume that
the instruments have objective characteristics that will produce relatively
similar consequences regardless of the setting within which they are being
utilized. The constructivist position, analogous to some of the framing
literature mentioned above, is that policy instruments are generally dis-
cussed in objective terms, but they can also be seen as social and political
constructs.

For example, Lascoumbes and Le Gales (2007) point to the extent to which
most studies of instruments have tended to look at them in a rather tech-
nical manner. They argue that instruments, like other aspects of policy,
need to be understood in a political and social context as well as simple
“tools”. The tools metaphor has been useful, but may undervalue the
social aspects of these means of creating public action. Further, the focus
on individual tools tends to undervalue the importance of constructing
policy mixes that involve multiple instruments. Even a simple tools logic
might alert the designer that not all instruments will work together easily,
but the more political and constructivist position may make the need to
consider interactions even more visible. The social perspective on instru-
ments may also facilitate their adaption to changing circumstances in the
environment of the policy (Howlett, 2019).
114 ADVANCED INTRODUCTION TO PUBLIC POLICY

The psychology and sociology of instruments: nudge

and its allies

There has been increasing interest in a set of tools relying on psychologi-

cal rather than more material forces to produce the desired outcomes on
behalf of the public sector (Thaler and Sunstein, 2008). The basic idea of
nudge is to provide subtle incentives and disincentives to the public to get
them to behave in the ways desired by policymakers without having major
intrusions, or at least obvious intrusions. In this way nudge is perhaps at
an even farther end of the intrusiveness scale than the examples given
by by Salamon. For example, rather than having real police personnel
to deter speeders it may be possible to achieve the same thing simply by
parking empty police cars in strategic places.

Although nudge is an important addition to the armamentarium available

to policymakers, it is a rather general conception of a style of intervention,
Much of this literature provides interesting examples of how less intrusive
mechanisms provide often subtle cues to citizens to behave in certain
ways. Can these seemingly subtle interventions be related to the four cat-
egories in Hood’s NATO scheme? For example, the example above of the
empty police car is actually invoking authority, or the threat of authority.
Further, many, if not most, of these interventions depend on nodality and
the diffusion of information, assuming a particular reaction of the target
individuals based on that information (John, 2013).

As well as having a psychological dimension, policy instruments also can

be seen as social and political constructs. The capacity of instruments to
achieve their ends may be a function of the manner in which they are per-
ceived by the actors involved, especially by target populations. The nudge
mechanisms, and indeed many other informational instruments, tend to
rely heavily on accepted social norms for their success. They assume that
citizens want to do the right thing, so that if they are reminded what is the
right thing to do – pay one’s taxes for example – then they will be more
likely to comply. This approach may therefore be effective in societies
with common, and publicly regarding, value systems but less effective in
more fragmented and individualistic societies. Further, the use of policy
instruments can alter underlying conceptions about government inter-
vention and spillover from one policy area to another (d’Adda, Caprarp
and Tavoni, 2017).
 POLICY INSTRUMENTS 115

Evaluating the classifications

The attempts to classify policy instruments have been significant for

the development of policy studies in political science, but the multiple
versions of classification also raise some analytic questions. What is the
relative utility of the various classifications? Are they mutually really that
different or are they all looking at some fundamental features of instru-
ments and merely discussing them from slightly different perspectives?
Further, does the emphasis on differences in tools mask the important
reality that most instruments are really hybrids. Even in those classifica-
tion schemes focusing on a single dimension, for example, coercion, any
individual instrument may have some features which are highly coercive
and others which allow a good deal of choice on the part of individual
citizens.

The simplest versions of classification of instruments are simple enumer-

ations, for example, Kirschen and to some extent the Salomon Handbook.
Simply understanding the range of instruments available to government,
as well as some of the characteristics of those instruments, is not without
its uses, especially for a policymaker thinking about options for a program.
That simple enumeration does not, however, provide the policymaker, or
the student of policy, much guidance about what to expect from any of
the possible choices.

The other classifications schemes provide at least some of the needed

answers, but none really links any particular instrument to either a type
of problem or a likely set of consequences. These classifications do give us
some insight into the ways in which governments can and do intervene,
and emphasize some important analytic points about the modes of inter-
vention. The classification scheme based on levels of intrusiveness, for
example, forces the policymaker to consider very carefully the political
impacts of instruments, as indeed does the seemingly simple classifica-
tion used by Salamon. The NATO scheme, with the amendments from
Howlett, does provide some inklings about the relative utility of different
instruments, but is not concerned directly with their relative utilities.

The discussion of instruments is usually done individually – this is a good

instrument for this policy situation or it is not. But most policy programs
use several different instruments and each of those may use more than
116 ADVANCED INTRODUCTION TO PUBLIC POLICY

one of the resources for instruments mentioned by Hood or Knoepfel.

Therefore, it is important to think about the hybrid forms of tools and
about the ways in which their use can adapt over time (Snyder, 2015).

Choosing instruments

Once we have an understanding of the characteristics of policy instru-

ments, we can begin to consider the ways in which political actors make
choices about instruments. This choice may seem like simply matching an
instrument to the demands of a policymaking situation, but unfortunately
there is no simple algorithm to make the linkage between the situation
and the instrument. If there were such an algorithm policy design would
be very easy. Therefore, like most things in governing, judgment is
involved in making those decisions.

Although judgment is certainly involved in the selection of policy instru-

ments, that statement alone does not go far toward understanding how
the choices are made. Further, we should not expect those choices neces-
sarily to be rational in the usual sense of that term, but rather may reflect
a number of factors that are more political, personal and organizational.
Indeed, many of the choices will be made from habit and routine rather
than through rational calculation. Although we are stressing the design
of public policies, that design process will involve a wide range of factors
rather than simple answers for complex questions. Further, those choices
are made by individual decision-makers and by policymaking organiza-
tions, with each of these sets of actors bringing their own perspectives to
the choices.

Individual decision-makers

Ultimately, individuals make decisions about which instruments to use

in any situation. We (Linder and Peters, 1989) therefore asked a group of
decision-makers in government how they made these choices. Although
these respondents provided a variety of answers, their answers could
 POLICY INSTRUMENTS 117

be divided into four groups. The first, and largest, group we labeled
“Instrumentalists”, and these respondents said they would select the
same instruments almost regardless of the circumstances. Many of these
decision-makers had strong professional backgrounds, so that economists
tended to opt for mechanisms depending upon economic incentives,
while lawyers depended on law-based instruments such as regulations.

A second group of respondents were labeled “Managerialists” (approxi-

mately 20 percent of the sample). These actors argued that the selection
of instrument was to some extent irrelevant to the success of a program.
They believed that they would be able to make any instrument work
effectively. As might be expected, the majority of these respondents were
trained in public administration, and believed that the management of the
instrument was more important than was the nature of the instrument
itself.

The third group of respondents were those whom we hoped we would

find in this research project. These we labeled “Contingentists”, meaning
their answers were that the choice of instrument depended upon the
problem and the situation in which it was to be applied. They possessed,
whether from experience or academic training, some sense that tools did
indeed have appropriate as well as inappropriate uses and some matching
was required. Unfortunately, we found relatively few of these individuals
(only 10 percent) in our sample of policymakers. Further, although they
understood the potential utility of linking problems and instruments they
had only intuitive ideas about how to do that in practice. That said, they
hoped to find some means of linking contingencies and policy problems
to the choice of instruments.

Finally, there was a small group (5 percent) of respondents, whom we

labeled “Constitutivists” who had something of a post-modern, or at least
constructivist, conception of instruments and the public policy process
more generally. These respondents argued that instruments could not
be understood out of the context of the policy problem, and the problem
and the instruments to address it had to be constructed simultaneously.
This perspective to some extent reflects the complex reality of making
policy decisions, but it also tends to make analysis of policy in other than
a fully contextualized manner difficult. Further, these respondents may
have had, perhaps unwittingly, a perspective on policy similar to that of
118 ADVANCED INTRODUCTION TO PUBLIC POLICY

the “garbage can” in which instruments combine with problems almost

by accident.

In summary, individuals within public organizations have their individ-

ual perspectives on how to design policies, and on which instruments are
the best to achieve their policy goals. As well as waiting for opportunities
to utilize their favorite policy instruments, individuals – especially the
Instrumentalists and the Contingentists – are likely to be policy entre-
preneurs and advocates (Kingdon, 1985 [2003]). As has been argued in
the garbage can model of choice (Cohen, March and Olsen, 1971; see
also Béland and Howlett, 2016), solutions chase problems just as much
as problems chase solutions. The individuals who have clear ideas about
how to solve policy problems are therefore very likely to try to find situa-
tions in which they can employ their favorite tools.

Unfortunately, this research did not take into account the manner in
which individuals with different perspectives on instruments interact,
or the extent to which the organizational framework within which they
make those decisions influences their choices. No matter how influential
they as individuals may be, they must still function within an organiza-
tional setting and the choice of instruments may develop through negoti-
ations among many actors. Therefore, in the next section of this chapter
we will discuss a range of other factors that may influence the selection of
instruments.

Institutions and instrument choice

In the public sector one of the most common explanations for any deci-
sion is that institutions are relevant. This is as true for instrument choice
as for other decisions, and institutions can have a significant impact
on these decisions. In the case of instrument choices, institutions have
influences in several ways. The first is that institutions have routines and
habits, and they have certain instruments to which they are committed.
For example, when social service organizations in most countries are
faced with a social policy problem they are likely to think first of social
insurance as the most desirable option. These instruments appear to have
been effective in the past, so why would they not be in the future?
 POLICY INSTRUMENTS 119

This selection of instruments through familiarity is not necessarily irra-

tional on the part of organizations. If the organization is familiar with an
instrument and knows how to make it function, then it is quite sensible
to continue to employ it when possible. This logic for selection makes the
additional point that instruments do not work automatically but rather
have to be administered. And even minimizing decision-making costs,
and not investing in extensive analysis to find the “perfect” instrument,
is far from irrational. The proceduralists (see above) may be correct and
good administration, or administration that is familiar with the dynamics
of an instrument, may be capable of overcoming many difficulties result-
ing from a less than perfect match between instruments and problems.
And attempting to train employees to manage a new instrument may be
a waste of resources when the familiar instrument may be as effective.

As well as selecting instruments out of habit, institutions may select

certain instruments for political reasons, meaning primarily bureaucratic
politics. Just as instruments must be administered to be effective, they
also have political characteristics that should be considered when being
adopted. Emphasizing one type of instrument or another can create
a political advantage for an organization. For example, regulatory organ-
izations will benefit if governments attempt to address economic issues
through those legal types of instruments rather than through more direct
forms of intervention. Or is the best way to address issues of malnutrition
through food subsidies (favoring a ministry of agriculture) or through
cash transfers to the less affluent families? Either could work but the
politics, and the political economy of the programs will be very different.

The institutional role in choosing instruments may not be simply a matter

of their choice, however, and may be a function of the range of legal
opportunities available to institutions. For example, an organization may
want to use loans or guaranteed loans as the mechanism for achieving
its goals. But to do that it requires the authority to make those loans, or
more generally to have access to funds that can be loaned. Of course new
legislation could provide the opportunity to make loans, but that is new
legislation and therefore involves building a political coalition necessary
to enact the law.
120 ADVANCED INTRODUCTION TO PUBLIC POLICY

Ideas

Ideas can also be the source of choices of policy instruments, as argued

by the constructivists mentioned above. We generally think about ideas
influencing the substance of policies rather than the instruments for the
delivery of those policies, but there are definitely ideas about the relative
virtues and vices of different instruments. For example, the neo-liberalism
as a seemingly dominant ideology during the 1980s and beyond tended to
denigrate the direct delivery of public services by governments, produc-
ing a spate of schemes for contracting out programs to both market and
non-market actors.

Ideas and academic theories provide policymakers with a set of ideas

about causation, and these ideas often lead on to a set of mechanisms for
intervention. For example, Keynesian economics provides a means of
understanding the business cycles that have beset capitalist economies for
centuries. If indeed changing levels of effective demand are at the heart
of the problem then this understanding leads on to policy instruments
that regulate that demand. Likewise, if one accepts the ideas of monetary
economics then another set of economic policy instruments becomes
appropriate for addressing the business cycle.

As noted above concerning the “Instrumentalists” in our sample of poli-

cymakers, the professions also provide their members with a set of ideas
about which instruments to utilize. Professional training tends to provide
the recipients of that training with a set of ideas about how the policy
world is organized and what the most effective means of intervention may
be. Professionals may observe the same social or economic problem and
not only assume different causes but also have different remedies. While
these different ideas at times may be useful, the potential incapacity of
participants in the process to understand, much less accept, alternative
conceptions of the issues may make effective policymaking difficult.2

And it is not only the professions that provide specialized and potentially
narrow conceptions of policy problems and solutions. Any group of
experts will tend to constitute an “epistemic community” that under-
stands policy problems and solutions through the lens of a body of spe-
cialized knowledge, and who tend to exclude others who do not possess
this knowledge, or even if they come from a different “school” within
 POLICY INSTRUMENTS 121

the community (see Zito, 2001; Dunlop, 2013). Having this specialized
knowledge is of course a virtue but its exclusivity and the restricted vision
may not be so virtuous.

Ideas can influence the selection of policy instruments through processes

of learning and diffusion (Dunlop and Radaelli, 2018). Although policy
diffusion has been in operation for decades, if not centuries, the more
recent emphasis on “evidence-based policymaking” has made the possi-
bilities of learning about how instruments have worked in other settings
more popular with policymakers (Pawson, 2006). The use of the world as
a source of ideas for policy does provide opportunities for expanding the
range of instruments being used, but also can lead to misunderstandings,
and excessive optimism, about the ease with which instruments (or other
parts of policy) can be transferred from one setting to another.

Finally, ideas can be used to attempt to create change across the full range
of policy being implemented by government rather than just a specific
policy. “Mainstreaming” policy ideas involves requiring policymakers
in all domains to consider the impacts of their decisions on some social
or economic variables. For example, many countries now have gender
mainstreaming, requiring policymakers to consider the effects of policies
on gender equality. And an increasing number of countries now require
considering “health in all policies” (Leppo et al., 2013).

Interests

Institutions may affect the choice of policy instruments, but so too can
the interplay of social and economic interests.3 When we think of the
role of these interests in policymaking we tend to think of the goals of
policy and the structure of benefits being created for members of the
society. Interests, usually organized in the form of interest groups, are
also concerned with the instruments through which those services are
delivered. Instruments can benefit and disadvantage interests just as can
the substance of the programs that are being delivered. The effects may
be somewhat more subtle than those of expenditures or services, but the
effects are still real.
122 ADVANCED INTRODUCTION TO PUBLIC POLICY

The role of interests in tool selection presents something of a public man-

agement paradox. The tools that are the easiest to adopt, meaning primar-
ily that the tools are favored by the affected interests, often are the most
difficult to administer, and perhaps also the least effective in the longer
run. This can be seen clearly in the Affordable Care Act – “Obamacare”.
The complex mechanisms selected for expanding health insurance in the
United States involve the insurance industry directly, and in many ways
constitute a subsidy to that industry. This choice reduced opposition to
the program (at least at the onset) but produced a mechanism that is very
complicated and difficult to navigate for citizens.

The role of interests can also be seen in the creation of instruments that
are in essence building coalitions among affected groups and beneficiar-
ies. For example, nutrition programs, such as SNAP4 in the United States,
are in effect serving both farmers and the economically deprived. It might
be more efficient simply to give the less affluent citizens a cash transfer
and let them buy the food, but the program may not have been adopted
without the involvement of the farm lobby. This particular instrument
has the additional advantage of controlling the consumption of the poor,
reflecting the beliefs among conservatives that these individuals cannot
be trusted to make the correct decisions when given cash. This instru-
ment thus serves three interests – the needy, farmers and conservative
ideologues.

Summary

Trying to understand how policy instruments are chosen involves under-

standing at least four possible sets of explanations. The first is the role
that individuals play, and the various conceptions that policymakers have
of the best instruments to match particular circumstances. The second
factor is the role of institutions, and especially the dependence on organ-
izations in the public sector on routine and familiarity in the selection
of instruments. Third, we can see that policy instruments reflect ideas,
including ideas of the professions and other expert bodies with clear ideas
about what the best solutions for problems are likely to be. And finally,
we can see that social and economic interests are concerned with the
 POLICY INSTRUMENTS 123

selection of policy instruments, just as they are with the selection of other
aspects of public programs.

While this plethora of possible explanations for instrument choice is gen-

erally useful, it also raises the question of which of these explanations is
the most effective. The answer, as with so many aspects of policy studies,
is that “It depends”. For example, a policy which is visible to a number
of interests is likely to evoke their involvement and the exercise of their
political power in the selection. Likewise, a highly technical policy area is
likely to be dominated by ideas and the role of experts in advocating their
ideas for the solution will be crucial.

The final, and perhaps most important, question about tool selection is
not really answered by this array of explanations. In a design perspective
on instrument selection we want to know how to relate policy problems
to the tools that will attempt to solve them. These explanations for choice
as yet do not have such a capacity. We can get some inklings from this dis-
cussion, for example, that it is not totally irrational to utilize instruments
that an organization or individual responsible for intervening in the
policy area finds familiar, but there is no algorithm that links problems
and instruments with any probability of predicting the most appropriate
choice.

Evaluating instruments

Having now enumerated the array of tools available to governments, and

discussed some of the politics of choice, we must ask the daunting ques-
tion of what is a good tool? This question is daunting for several reasons,
not least of which being that it requires specifying the situation in which
the tool is used before any evaluation can reasonably be made. Further,
there are a number of dimensions along which tools can be evaluated, and
these are likely to provide contradictory assessments of the utility of the
tool, in general and in specific situations. This multiplicity of criteria is
a problem for the analysis of public policy in general (see Chapter 7) but
is certainly apparent for the assessment of instruments.
124 ADVANCED INTRODUCTION TO PUBLIC POLICY

In this section I will be providing some assessment criteria for instru-

ments coming from economic, political administrative and ethical foun-
dations. To some extent these criteria can be used in a general manner
– some instruments are likely to be more efficient than others – but that
also have to be understood in context. For example, at times governments
may want their interventions in the economy to be relatively invisible, but
at other times (crisis, for example) they may want the intervention to be
very visible so that citizens will know that it is actively taking measures to
address the problem.

These evaluations of instruments are closely related to the evaluations of

policies in general that will be discussed in Chapters 7, 8 and 9. To some
extent the nature of the instruments involved in a policy will produce the
overall effects of the program, although the content of an instrument – the
type of tax or the type of regulation – must also be considered. Further,
instruments that are preferred by citizens, for example, those that permit
greater choice, may make even other unpalatable programs politically
acceptable.

Political features of instruments

In the world of government perhaps the first set of criteria that must
be considered about policy instruments is their political characteristics.
In government politics generally is trumps so it is important to be able
to select instruments that will provoke the least negative reaction from
groups in society or from other actors in government. Those two political
characteristics, however, may themselves not be the same, and those
tools that the general public likes (or is most willing to accept) may not
be especially favored by actors within the public sector (political and
administrative).

Perhaps most fundamentally the congruence of an instrument changes

with the values of the general political and social values of the country.5
If an instrument does not have such congruence then it is less likely
to be effective, and more likely to provoke resistance. For example,
market-based instruments are relatively acceptable in the United States
but may be less acceptable in European societies that have substantially
 POLICY INSTRUMENTS 125

less devotion to the private sector. Similarly, the United States would find
more intrusive instruments involving direct public sector involvement
less acceptable while they are normal for much of Europe and Latin
America. Even demands to wear a face mask in the midst of a pandemic
were considered intrusive by many Americans.

The visibility of an instrument often has an effect on its political accepta-

bility. That is, some instruments are readily apparent to the public while
others may be well hidden. The contrast between the value-added tax
which does not appear as a separate item in the price and the general
sales tax in the United States or Canada that is added at the point of sale
demonstrates different levels of visibility of a common instrument (a tax
on expenditures). For conservatives higher visibility is generally positive,
making the costs and effects of government more evident to citizens. On
the other hand, liberals (and policymakers) might be more pleased with
less visible instruments that might minimize public resistance.

Finally, the accountability of policy instruments is a crucial political

criterion. As noted in the preceding chapter and in reference to some
extent in the discussion above, the use of non-governmental actors and
more complex forms of service delivery makes accountability more dif-
ficult. Thus, while these instruments and forms of intervention may be
less expensive, and in some ways more effective, the difficulties involved
in holding the participants accountable for the use of public money and
public authority make them potentially suspect politically (Considine,
2002).

Economic features of instruments

Policy instruments have economic characteristics that also must be

considered when making decisions about how to design public poli-
cies, Perhaps the most obvious of these is the cost of the instrument.
Some policy instruments, especially those that involve private actors
or “prayers”, may impose little costs on the public sector, and in some
cases they may be effective in producing the desired outcomes. This then
quickly leads on to a discussion of the efficiency of the instrument, with
126 ADVANCED INTRODUCTION TO PUBLIC POLICY

those producing greater ratios of benefits to costs being more desirable

even if they involve more total resources (see Chapter 8).

Another important economic characteristic of instruments is the extent

to which they tend to distort the market. In the best of all economic
worlds instruments (and public sector programs more generally) would
not disturb the efficient (sic) functioning of the market in allocating
resources. This market distortion can be seen perhaps most easily in
tax expenditures such as that given to housing in most political systems
(O’Sullivan and Gibb, 2003). These programs tend to divert investment
away from potentially more productive uses of the money into the con-
sumption of better and bigger houses than might otherwise be possible.
While citizens may like this effect of the policy instrument in supporting
home ownership, economists may not be as supportive of this instrument.

Administrative criteria of policy instruments

To be effective policy instruments must be administered, and different

instruments are more or less capable of being administered effectively.
While the economic costs of the instruments certainly affect their admin-
istration, there are other features of instruments that must also be consid-
ered. In administering instruments there must be economic efficiency but
there are also other conditions that should be met. As in so many other
aspects of public policy, there are multiple criteria to be considered and
there may well be contradictions and conflicts. For example, economic
efficiency may be in conflict with desires to target particular segments of
the population effectively.

The capacity to target sectors of the population is a crucial administra-

tive criterion of policy instruments (see Ingram and Schneider, 1990;
Schneider, 2013) . A good policy instrument will deliver the service to
those members of the society who are meant to receive it but not to those
who do not. This is a difficult standard to meet for any instrument. On
the one hand, instruments may deliver the program to those who were
not the intended targets, or they may miss individuals or organizations
that were intended to receive the benefits or punishments. Conservatives,
for example, oppose social programs that may deliver the benefits to
 POLICY INSTRUMENTS 127

individuals who are not legally eligible. On the other hand, an instru-
ment that does not identify and vaccinate all potential victims of a dread
disease must be considered ineffective even if it costs little per individual
vaccinated. Thus, targeting is an administrative criterion but like almost
everything else in public policy analysis there is also a political dimension
which must be considered.

The enforceability of an instrument, and therefore its effectiveness, is

also relevant for its administration. For example, many states in the
United States now require citizens to pay the same sales taxes on goods
bought on the internet as they would if the products were purchased at
a store within the state. The problem is that the states have no reasonable
means of monitoring those purchases and as yet have not been able to get
national legislation to force on-line merchants to collect the tax for them.
Therefore this tax is rarely paid, and may make the state government
appear ineffective.

The lack of enforcement of poorly designed or conceptualized pieces of

legislation not only has short-term policy consequences but it may have
longer-term consequences for citizens’ respect for government. If govern-
ments persist in passing laws that cannot be enforced then it exacerbates
the image held by many citizens that it is ineffective and rather inept.
That image will, in turn, make the enforcement of even more reasonable
laws more difficult. There is increasing evidence that governments are
now legitimated as much through their effectiveness as through proce-
dural mechanisms such as voting (Gilley, 2009) so government should
be cautious about undermining the perception (and the reality) of its
enforcement capacity.

Time plays an important role in the assessment of instruments and their

administration. All political leaders would like to create benefits imme-
diately, and deter costs as long as possible. That ideal world is probably
not accessible, but policymakers do need to consider time not only for
its potential political benefits and costs. Programs that produce their
intended benefits quickly are more likely to be welcomed than those that
may produce even greater benefits further in the future.6 And it is not just
politicians who prefer benefits in the short run. Citizens also prefer imme-
diate benefits and have a very high discount rate for benefits produced in
the future (Frederick, Loewenstein and O’Donoghue, 2002).
128 ADVANCED INTRODUCTION TO PUBLIC POLICY

While some of the administrative impact of time may be political and/

or psychological, there may be more tangible factors to consider as well.
The world, and perhaps especially the world of policymaking, is uncertain
and changing. Therefore the social and political conditions on which
a program is premised may disappear quickly, whether through economic
crisis, changes in government or through technological change. While
immediate benefits are too much to be hoped for, designing programs
for a very long-term payoff may be both politically and administratively
risky.7

The availability of the internet and other digital mechanisms has to some
extent altered the way in which policy instruments are administered
(Hood and Margetts, 2009). There are much greater opportunities for
individual citizens to implement programs for themselves. We can pay
taxes, pay for parking, apply for a passport and perform a host of other
tasks without actually having to encounter a public servant. But the
digital divide based on income, residence, race and other factors may then
produce greater inequalities in the services delivered to citizens. And the
digital divide between wealthy and poor countries may be even greater
than that within individual countries.

Finally, as already discussed, the familiarity of an implementing organiza-

tion or individuals with an instrument must be considered when design-
ing programs. An instrument that is familiar to the implementers, even if
those implementers are ordinary citizens, is more likely to be successful
than more creative forms of intervention. Over time the implementer will
learn about how to manage a new instrument and the virtues of familiar-
ity will be restored, but the transaction costs are potentially substantial.

Ethical criteria

Finally, instruments affect other values in society than just the utilitarian
values captured in the economic assessment. Most of our discussion of
policies and policy instruments calculates the costs and benefits and then
compares those two totals. But the tools also raise normative and ethical
questions that need to be included in an overall assessment of the tools.
 POLICY INSTRUMENTS 129

While there are a number of criteria of this sort, the following list provides
some insights into the possible impacts of tool selection.

The most important of the values affected by the selection of tools is the
extent to which the autonomy of individuals is preserved. We can assume
that everything else being equal, policies, and the tools used to implement
them, should maintain the autonomy of individuals. In democratic soci-
eties we assume that individual citizens should be able to make as many
decisions as possible about their own lives. For example, pensions for the
elderly are paid in cash and the recipients can do anything they want with
that money. Most pension recipients will use that money for food, heating
and all the other necessities of life, although there is nothing to prevent
their spending it on alcohol, gambling and tobacco. The state could force
pensioners to live in supervised housing and eat three healthy meals each
day, but instead lets adults make their own decisions. And in some cases,
for example, “food stamps” in the United States, government does ensure
that the benefits are used only in certain ways.

The maintenance of the autonomy of the individual can extend beyond

simply letting people spend their government checks as they see fit. The
use of “nudge” and similar policy instruments to some extent removes
the autonomy of individuals by manipulating them to respond in certain
ways. The designers of the program assume they know better than
ordinary citizens, and exercise their “liberal paternalism” to produce the
desired responses (Thaler and Sunstein, 2003, p. 179).

The administrative criterion of targeting mentioned already should be

related directly to an ethical requirement of equity. That is, in an ethical
framework for policymaking individuals who are similarly situated rela-
tive to the targets of the program, for example, who are equally in poverty,
should be treated equally by public programs. But an instrument that
depends heavily on the individual initiative of the potential recipients
for receiving benefits is likely to miss individuals who are illiterate, lack
transportation or who are simply shy. Depending upon the individual,
to apply for the benefit may be conceptualized as a reasonable rationing
method administratively but it does violate a sense of equitable treatment
of citizens.

The role of administration in producing equity can also be considered

from the perspective of the discretion being exercised by the public serv-
130 ADVANCED INTRODUCTION TO PUBLIC POLICY

ants (and non-governmental actors) in the implementation of a program

(see Hupe, 2013). Programs that provide their civil servants – especially
street-level bureaucrats – with a great deal of discretion are less likely to
produce the equitable outcomes that should be desired on ethical (and
legal) grounds. The reforms of the public sector under the rubric of New
Public Management have emphasized granting civil servants greater
autonomy and discretion, so that the possibilities of these outcomes are
increased. Somewhat paradoxically, the more civil servants are empow-
ered, the less their clients may be.

Summary

I have now presented four alternative dimensions along which to evaluate

policy instruments. These criteria range from the seemingly hard-headed
economic analysis of costs and benefits to much softer and less quantifia-
ble criteria such as normative and ethical standards. Each of these sets of
criteria is important, has relevance for the success of public programs, but
there are internal differences within each of them. For example, some eco-
nomic criteria may conflict with each other, so that total costs may have to
be considered along with the relative costs and benefits of the instrument.

As well as the conflicts within the four broad categories of criteria, there
may more likely be conflicts across those categories. As already noted,
the utilitarian nature of economic criteria are likely to conflict with other
criteria based more on equity. The difficulty in all these criteria, and the
conflicts among them, is that there is no clear way in which to rank and
to weight these criteria in designing programs. When faced with choices
among instruments the decision-maker must exercise judgment about the
relative virtues of the instruments. The criteria will therefore be applied
differently in different situations and by different organizations. This
judgment is perhaps inevitable and having this list of criteria may help
by at least making the choices more apparent to those decision-makers.
 POLICY INSTRUMENTS 131

The Swiss Army Knife of government

Roderick Macdonald (2005) has provided one of the most interesting

examinations of the tools literature in policy studies, likening the tools
available to government to a Swiss Army Knife, containing an often dizzy-
ing array of possibilities. Those possibilities themselves, however, present
a problem of choice for governments. Indeed, simply knowing that all
these possibilities exist raises the question of what to do with them, and in
what circumstances. And even with the several dozen tools on the largest
of the knives (or the largest tool chests of any government organization)
the right tool may not be there.

Macdonald’s full paper provides a number of insights into the nature and
selection of policy tools, but several points of the analysis stand out and
can serve as a useful summary for this chapter. The first point is there
is rarely if ever a single best response (tool choice) for a situation. That
selection is cultural and situational and the effects of tool choice may
only be understood as the program is being implemented. Therefore, we
should perhaps consider that tools that can have multiple uses (the hook
disgorger) as opposed to a single use (the corkscrew) are potentially more
valuable.

A second point worth remembering in tool selection is that policymak-

ers are not writing on a tabula rasa but rather are intervening after any
number of other previous attempts to solve a problem (see Hogwood
and Peters, 1983). This persistence of responses means in part that there
are preconceptions about what tools are useful and how they should be
administered that may be difficult to overcome. Just as the makers of the
Swiss Army Knife have assumptions about which tools we need and how
we will use them, so too do organizations in government fall back on their
comfortable routines when they select and implement policy instruments.

Third, there is a danger that the availability of multiple tools overwhelms

judgment. The more sophisticated the tools of government become, the
less room there may be for public servants and other implementers of
the programs to adapt and adjust the programs as conditions change
or unforeseen circumstances arise. In short, overly detailed instrument
choice may in the long run produce rigid and ineffective implementation,
whereas a simpler tool (the knife blade?) may have generated just as much
132 ADVANCED INTRODUCTION TO PUBLIC POLICY

positive output with less cost. But how do we know what the limitations of
the knife blade may be, even if it is our favorite instrument?

I do not want to belabor the point about the Swiss Army Knife but the
analogy to government instruments is interesting and does help to illus-
trate some of the dimensions of choice that are involved in designing
public policy. Instrument choice appears as easy as reaching into a knap-
sack, pulling out the knife, and then choosing the obvious instrument for
the task. But the obvious instrument may not really be the best, even if
we were able to define “best” in any unambiguous manner. Further, the
particular set of instruments that are most available may not be the best,
or even adequate, to achieve the tasks.

Summary

Instruments are crucial for implementing public policies. They some-

times tend to be considered as things unto themselves, but they are
primarily means of delivering public programs. Therefore, the successful
program designer must consider carefully the match between the goals of
the program and the availability of instruments. The best tool may involve
treasure, but in the midst of a government financial crisis the only real
alternative may be information. So then how do we make information
work in that context?

And not only is the temporal context important, so too is the cultural
and social context. Understanding the norms and values of the society
into which a program is being implemented is important not only for the
success of the particular policy but also for the general political success
of the political system. Using coercive instruments in societies that value
autonomy and individual choice may ultimately generate compliance, but
at some cost. The skillful designer must always remember that he or she
is functioning in a political environment that will influence success and
failure.
 POLICY INSTRUMENTS 133

Notes
1. For Europe see Héritier and Rhodes (2012) for a similar analysis of contem-
porary policy instruments.
2. The logic of reframing (see Chapter 4) attempts to address these divergent
views and produce effective responses.
3. Of course, institutions also have an interest in their preservation and in min-
imizing disruption to their established routines. These may not be as visible
as the interests of social groups, but they are interests nonetheless.
4. Supplemental Nutrition Assistance Program – the program that used to be
called Food Stamps.
5. On congruence theory see Eckstein (1980).
6. The logic of discounting in cost-benefit analysis places this criterion in
a more economic context. See Chapter 8.
7. The perception of time in policy may also be a cultural factor, with some
societies having a longer-term perspective and being more willing to undergo
short-term deprivations for that future gain.
PART III

Evaluating policy
7 Evaluating public policy:
an introduction

After governments formulate and implement policies, those policies must

be evaluated. The need for evaluation arises from at least two demands
about action in the public sector. The first of these is the need to assess
how well programs are performing in order to make that performance
better. While the usual image among the public is that government pro-
grams do not work very well, the evidence from many countries is more
positive (Schwartz, 1987; Baggott, 2012; Compton and ‘t Hart, 2019).
Even then, however, for programs that are performing reasonably well
there is a still an opportunity to learn from past interventions in order to
improve those programs.

As we will point out below, that assessment of performance can be done

in a variety of ways. As well as the distinction between economic and
ethical evaluations of policy, there is a difference between short-term and
longer-term assessments of policies. Much of the contemporary empha-
sis on performance measurement and management tends to empha-
size shorter-term assessments as contrasted to conventional evaluation
research (Rossi, Lipsey and Freeman, 2004; Vedung, 2007) that empha-
sizes longer-term and more fundamental evaluations of programs. These
styles of evaluation serve different purposes, and reflect different political
priorities, but both can make contributions to understanding public
policies.

The assessment of performance is the principal component of this funda-

mental purpose of evaluation and this analysis can provide opportunities
for learning. If we know a program does not work as intended then that
information can be used to determine what works in public programs
(see Marier, 2013). In addition, the information coming from evaluations
can be shared across countries or levels of government, contributing
to the “evidence-based policymaking” that has become popular among

135
136 ADVANCED INTRODUCTION TO PUBLIC POLICY

decision-makers in the public sector (Pawson, 2006). Although there

are multiple barriers to effective learning from evaluations, these studies
provide a foundation for understanding how and why programs work
and also what can be done to improve them.

The second use of evaluation, in addition to understanding which pro-

grams work and which do not, is for purposes of accountability. Holding
government accountable for its actions is a fundamental value for
democracies, and even for non-democratic governments (Hadenius and
Toerell, 2007). In order to know if the remainder of the political system
(especially the public bureaucracy) is delivering the programs having
been formulated and resourced by the political actors in the system, those
services must be evaluated. As well as identifying success and failure of
the program, evaluation may also be used to identify the sources of any
failures in the system.

The accountability dimension of policy evaluation is especially important

given how much of accountability is conducted almost exclusively as
a political exercise. That is, instruments of accountability, such as ques-
tion time in parliament, are used to embarrass a government, often over
rather trivial matters, while major policy issues are shunted aside. These
more political forms of accountability make good fodder for the media
but may not address underlying public problems. Therefore, emphasizing
the role of evaluation in accountability, even in the diminished form of
performance management (Peters, 2007), may improve performance in
the public sector.

The process of evaluation

The Swedish scholar Evert Vedung is one of the leading experts in policy
evaluation. He has argued (2013) that there are intellectually, and in prac-
tice, six models of evaluation. Each of these models has a set of governing
assumptions that direct the evaluator to examine a set of outcomes and
processes that require assessment, and which also raise different ques-
tions about those outcomes. While the goal attainment model Vedung
discusses is almost certainly the typical model for evaluation, the other
 EVALUATING PUBLIC POLICY: AN INTRODUCTION 137

models also provide insight into evaluation as a technical exercise and

perhaps more importantly as a political exercise.

What Vedung refers to as the goal attainment model is the default option
for policy evaluation. The basic question for this model is whether the
outputs intended by the framers of the program were produced or not,
and to what extent, and at what cost. While this model appears quite fun-
damental, it still depends heavily on the capacity to measure goal attain-
ment effectively. And it also depends heavily on the capacity to identify
program goals in some unambiguous manner while, as already noted for
implementation, that can be very difficult, and then to identify the extent
to which programs have been able to attain those goals.

A second model of evaluation has been referred to by Vedung as the

side-effects model. This model begins with some of the questions of
the first model but then expands the scope of coverage of the analysis
(Merton, 1936; see below). The question in this model then becomes not
only whether the stated goals of the program are attained but also if there
were negative side-effects that would undermine the overall efficacy of
the program. The difficulty with the side-effects model then is how much
negative side-effect offsets how much positive goal attainment? And
again, there are the fundamental questions about measurement that must
be extended to include negative side-effects, and perhaps also positive
side-effects that must also be considered.

Relevance is the third model discussed by Vedung. In this model the

target of the evaluation is not goal attainment of any particular program,
but rather the effects on the underlying problem in the policy area. In the
default goal attainment model the assessment is of reaching the stated
targets of the program, but in this model the question is whether the
underlying social or economic problem is being addressed successfully.
For example, a medical program may reach its stated targets but not
be very effective in actually improving health.1 Using this model makes
the assessment of a policy intervention all the more difficult, given that
it is not always clear what the problem is that is being addressed by the
program (see Chapter 2) and the extent to which the program may con-
tribute to overall social well-being.

At one point in the development of program evaluation there was

a movement for “goal-free evaluation” (Scriven, 1991). The logic of this
138 ADVANCED INTRODUCTION TO PUBLIC POLICY

form of evaluation is that evaluators should not be told what the formal
and explicit goals of the program are, but rather should attempt to assess
the benefits and the costs of the program in the abstract. Thus, even if
the program may be reaching some stated goal its general outcomes for
the society, as in the relevance model, may be insufficient to consider it
a success.

A fourth version of evaluation, the client model, permits the clients of

a program to make their own evaluations of the program. The previous
models of evaluation have been largely objective, depending upon meas-
uring the attainment of goals, or perhaps effects on some underlying
policy problem. In this model, however, the question becomes whether
the targets of the program are satisfied with the results, and perhaps
also if they are satisfied with the manner in which the program is being
delivered. These subjective evaluations of the program are made more
difficult because there may be multiple beneficiaries of the program with
different assessments, for example, farmers and the less affluent evaluat-
ing a hunger program.

It is quite possible that the objective and the subjective dimensions of

evaluation will not correspond (see Bovens, ‘t Hart and Peters, 2000).
Both forms of evaluation have their difficulties. The objective forms of
measurement may be driven by professional and political standards,
which are important but may not really satisfy the clients of the programs.
Indeed, for some programs that are controlled politically, for example,
social welfare controlled by conservative governments, the evaluations
may be almost inherently related negatively. On the other hand, if regu-
lated industries evaluate a program positively it may not be accomplish-
ing its goals.

The stakeholder model is to some extent the client model writ large. That
is, instead of just considering the views of the clients alone, this model
considers the views and interests of a whole range of actors connected to,
and affected by, the program (Vedung, 2010). The stakeholder model of
evaluation is more participatory than the others, and therefore provides
a range of alternative perceptions of the performance of the program. At
the same time, however, the multiple assessors involved only exaggerates
the underlying problem of obtaining something approaching the unam-
biguous answers about performance that is inherent in policy evaluation.
 EVALUATING PUBLIC POLICY: AN INTRODUCTION 139

This model can be expanded even further to think of generalized public

involvement in evaluation. Democratic governments have used mech-
anisms such as public hearings and open public meetings as means of
evaluation and policy formulation for some years, but the drive for public
participation in these systems has expanded opportunities (Bingham,
Nabatchi and O’Leary, 2005). Increased public involvement has definite
political advantages but may not be able to provide the detailed assess-
ment of performance that is desirable for evaluation.

Finally, a collegial model of evaluation allows the actors involved in the

delivery of the service to evaluate themselves. While this would appear to
violate an assumption that evaluation should be done by some independ-
ent organization or actor, there are some good reasons to consider this
internal model of evaluation. The most important is that who would know
better how the program functions and what the potential sources of failure
are than the people actually involved. Further, this style of evaluation may
be less disruptive to the organization than the uncertainty created by an
external evaluation. Finally, this form of internal evaluation has become
one component of the general move toward empowering workers within
public sector organizations, with the belief that this empowerment will
ultimately improve performance (Denhardt and Denhardt, 2007).

These six models can all provide insights into the performance of public
programs, and all can be justified as valid means of understanding how
government works. Therefore, strategic questions arise concerning which
model is appropriate for which settings, and how governments (and their
partners) should choose among them. The answers to those questions
may depend in part upon the design of the programs and the intentions
of the formulators. For example, if a program is intended to be a relatively
technical intervention into a well-known policy area then the goal attain-
ment model appears appropriate. If, however, the program has a broader
perspective and affects numerous actors in society then the client or the
stakeholder model may provide more useful information.

The vast majority of the discussion of policy evaluation focuses on the

assessment of individual programs. This is understandable and reflects the
need for program managers, and their political sponsors, to understand
better how the program is performing. That said, however, there is some
need to consider policies in a broader perspective and to understand their
interactions with other programs. For example, attempting to understand
140 ADVANCED INTRODUCTION TO PUBLIC POLICY

whether a single health program works may also involve understanding

how it fits with a range of other health programs, as well as with nutrition,
recreation and housing programs. Even more broadly, subjective evalua-
tions of well-being by citizens can be used as a comprehensive measure of
the success of public governance (Bache, 2013).

Barriers to effective evaluation

Vedung’s discussion of evaluation models above makes it clear that

performing effective evaluations is not easy, and involves a number of
elements that must be aligned, or even whole alternative conceptions of
what the evaluation should be. There are, in addition, a number of specific
problems that confound the evaluator, most of these reflecting problems
in measuring the effects of programs. Paradoxically, the more meaningful
the intended measure, the less likely it is to be reliable and valid. In par-
ticular, governments and citizens want to know the ultimate impact of
programs on the society and economy, rather than the measurement of
outputs such as expenditures or outcomes, such as the number of services
being delivered.

Attempts to measure impacts, which in general occur in the future,

and often the distant future, lead to thinking about some specific prob-
lems with measuring across time. For example, there are sleeper effects,
meaning that programs that appear to produce little or no effects in the
short term may produce significant benefits in the longer term (Salamon,
1979). On the other hand, the initial effects of programs may decay, so
that programs that appear effective in the short term may not have any
sustained effects.2 Unfortunately for policymakers they must make and
evaluate decisions in the present, so must estimate what the longer-term
consequences of their actions may be.

The above discussion is based on the assumption, often incorrect, that the
goals of the program are clearly stated, or even identifiable. The political
process is such that it tends to produce legislation, and therefore pro-
grams, that have vague or even contradictory goals. Therefore, evaluators
may have little meaningful guidance about what to measure, and what
outcomes would indicate success. The evaluators therefore may have to
 EVALUATING PUBLIC POLICY: AN INTRODUCTION 141

make their own decisions about whether to assess what they consider to
be the relevance model (see above) or to posit more proximate goals and
associated measures.

Another problem often encountered in evaluation research is that even

seemingly successful programs produce side-effects and unintended con-
sequences. For example, creating the Interstate Highway System in the
United States did facilitate automobile travel but it also made it easier for
people to commute into the cities for work, and therefore made it easier
to move out of the cities. Thus, some portion of the decline of inner cities
can be attributed to the development of an improved highway system.
Further, the greater ease of travel on these highways produced more miles
driven with more pollution and more energy imported. Was this program
really a success, and if so, to what extent?

Sam Sieber (1980) developed a classification of the unintended conse-

quences of public programs, using the terms “fatal remedies” and “regres-
sive interventions” for the most serious cases that produced results exactly
the opposite of these intended. For example, studies of manpower train-
ing programs have shown that in some instances individuals who spent
the greatest time in the program were less likely to gain employment.
These were often individuals who required more training, but in many
cases potential employers had the perception that they were difficult cases
because they had spent so much time training. Even here, however, it is
difficult to call the program a complete failure because many trainees who
spent less time in the program were successful in getting jobs.

There are numerous barriers to effective evaluation but one that is not
discussed adequately is the role of the intended recipients of the evalua-
tion – typically political leaders and/or the general public. The literature
on policy advice has emphasized the need to equip political leaders with
the capacity to absorb and assess the advice they are being given (Landry,
Lamari and Amara, 2003). The information coming from evaluations also
must be understood by policymakers so that they use that information for
a subsequent round of policymaking. And ordinary citizens may need to
be able to understand the evaluations as well if they are to be capable of
holding their governments accountable.

When we consider the barriers to effective evaluation, we also need to

think about context (Pollitt, 2013). Many evaluation studies, especially
142 ADVANCED INTRODUCTION TO PUBLIC POLICY

those based on experimental designs, assume that their findings are

valid because they worked in one setting. But the school of “realistic
evaluation” scholars (Pawson and Tilly, 1997) has been pointing out that
a natural experiment in policy is still heavily dependent on context (see
also Tilly, 2000; Peters, 2020). As a simple example, a study of the effects
of labor market policy will depend upon local economic conditions.

Realistic evaluation then is analogous to Donald Campbell’s conception of

quasi-experimental design and the threats to validity that inevitably affect
such designs (see Shadish, Cook and Campbell, 2001). Evaluations are
always being conducted in time and space and those external factors will
influence the findings. The role of these contextual factors in affecting the
outcomes also must be considered when thinking about “evidence-based
policymaking”. If the evidence is so contingent upon a particular setting
then how useful is it?

Performance management as an alternative to

evaluation

The performance management models discussed by Vedung are all

based on a conception of evaluation that involves relatively detailed and
systematic research to generate answers about the underlying features of
programs and government policy. This is expensive and time-consuming
for the public sector, and often provides answers only after the program
has been operating for some time. In political environments that are
increasingly concerned with short-term results and sound bites rather
than more detailed explanations of programs, that style of evaluation is
increasingly unpopular. Further, as government resources are perceived
to be more scarce, governments find that reducing evaluation costs saves
money without reducing public services (although of course that may
reduce quality).

In part as a function of the spread of the New Public Management,

performance measurement and management has become a replacement
for evaluation research (Van Dooren, Bouckaert and Halligan, 2010).
Although this managerial instrument shares a common purpose of
assessing what is happening in the public sector, it is also significantly
 EVALUATING PUBLIC POLICY: AN INTRODUCTION 143

different from evaluation research. Most fundamentally, performance

management is a short-term perspective on performance, being con-
ducted every three or six months in most cases. While this may provide
some immediate feedback for administrators and politicians, the empha-
sis on producing demonstrable benefits in a short period may undermine
longer-term success.

Following from the point above, the emphasis of performance manage-

ment is on just that – management. The assumption, implicit or explicit,
is that if a program is managed well it will be successful in producing
the types of results intended. This approach therefore addresses only
indirectly the extent to which the initial design of a program is adequate
and may permit it to achieve those goals. There is little doubt that good
management can contribute to the achievement of goals, but there is also
little doubt that the best management available cannot overcome faulty
formulation.

Further, even if the program is designed well and still fails, is it necessarily
fair, or productive in the long run, to place all the blame for that failure
on the management of the organization delivering the program? Clearly
managers bear some responsibility for any failures, but there may be other
elements within the organization, and in the relationships with other
programs and non-governmental actors, that prevent program success.
From this perspective performance management can be an exceedingly
blunt instrument for assessing public programs and their success or
failure. This bluntness is especially important when the success of the
program depends on co-production – students being actively involved,
or participants in labor market programs actively seeking employment
(Alford, 2009).

Finally, the short-term focus of the program also enables managers to

“game” the outcomes and to undermine the exercise. Measurement is
an Achilles heel of performance management (see Bouckaert and Peters,
2002) and permits managers to engage in strategies that will enhance the
appearance, if not the reality, of their performance. This is sometimes
referred to as Goodhart’s Law, or “when a measure becomes a target it
ceases to be a good measure” (see Muller, 2018). For example, a training
organization may only accept clients who are well educated and well
motivated, rather than attempting to train those potential clients who
144 ADVANCED INTRODUCTION TO PUBLIC POLICY

require the greatest assistance. This makes the program look good while
actually defeating its purpose.

Crime statistics and police performance statistics provided an excellent

example of the problems with gaming performance information. Police
officials could classify crimes as lesser offences to make it appear that
more serious crime was declining. They would also bargain with crimi-
nals, getting them to accept blame for crimes other than those for which
they had been apprehended in exchange for lighter sentences – so-called
“nodding”. This increased the clearance rate of the local police and
made them appear more efficient. Officials in local police forces blamed
these tactics on the extreme pressures to meet performance targets (see
Rawlinson, 2013).

Thus, although measurement is an inherent problem for evaluation, it

may be even more so for performance management. There are the same
questions about the capacity to measure outcomes and impacts of public
programs, and link changes to specific features of a program, and the
short-term nature of performance measurement makes that linkage even
more difficult. Further, although there is some external monitoring,
a good deal of performance management is done through the organi-
zation administering the program. This characteristic of performance
management is analogous to the collegial model of evaluation discussed
above, and that internal monitoring may further compromise the objec-
tivity of performance management.

Leaving aside the problems with measurement of performance evalu-

ation, just as we discussed above, these measures can still provide an
important starting point for understanding, perhaps especially for the
public. Performance indicators in a number of policy areas such as
education and health are being distributed publicly, enabling citizens to
assess the quality of the schools that their children attend, or the hospitals
where they may go for surgery. These indicators may be imperfect when
compared against some absolute scale of measurement, but may still be
useful for citizens attempting to compare different service providers. And
these “league tables” provide a form of accountability for the performance
of public institutions (Hood et al., 2004).

Auditing has become another form of evaluation for public sector pro-
grams, and it falls somewhere between the intensity of evaluation research
 EVALUATING PUBLIC POLICY: AN INTRODUCTION 145

and the immediacy of performance management. Public audit organiza-

tions, for example, the National Audit Office in the United Kingdom and
the Government Accountability Office in the United States, do more than
financial audits, they also assess the efficacy and efficiency of programs
(see Pierre, Peters and de Fine Licht, 2018). The increasing role of these
organizations has made financial accountability more important in evalu-
ation, but also improved communication with lawmakers.

In many ways the results of performance management, and those league

tables, are more useful for consumers than they are within government.
In particular, much of the literature assumes that in the best of all worlds
performance information will be linked to budget, but how? If an organ-
ization performs poorly should it be punished and lose budget funds, or
would that merely punish citizens as much as the organization? Or should
the organization be given more funds to produce better performance,
or would that only reward failure? There are very difficult problems for
making evaluations and utilizing those evaluations of performance.

Experiments and evaluation

Most policy evaluation has involved conducting research on existing pro-

grams and determining the extent to which they achieve their stated goals,
and/or the extent to which they satisfy important stakeholder groups
(including clients of the program). This style of evaluation research has
yielded valuable information about programs and the outcomes they gen-
erate for citizens, but also can have significant flaws as social research. In
particular, they are implemented with few controls on extraneous factors
that may influence findings, and without considering possible alternative
forms for providing those services.

The “quasi-experimental” designs characteristic of most evaluation

research (see Shadish, Cook and Campbell, 2001) can be supplemented by
more strictly experimental designs that use randomized assignment and
control groups to gain a better assessment of the effectiveness of public
sector interventions. For example, in the United Kingdom and else-
where there have been experiments with sentencing convicted criminals,
attempting to understand what types of sentences may reduce recidivism
146 ADVANCED INTRODUCTION TO PUBLIC POLICY

(Farrington, 2003). The New Jersey Income Maintenance Experiment

in the United States provided evidence about the capacity of patterns of
income support to both encourage individuals to find employment and
to provide adequate support for the family involved, and Finland in 2019
performed a major experiment of income maintenance and employment
(see pp. 23–4; on Canadian experiments see Forget, 2011).

There is a growing body of experimental evidence used in evaluation,

and this is becoming slowly the gold standard for evaluation. While the
experimental method has many virtues methodologically, there are also
some questions (John, 2012) that must be raised when used as a means
of policy evaluation. The first is the extent to which findings from an
experiment will be replicated when a program goes into effect in reality.
Individuals will generally know that they are part of an experiment and
therefore may behave differently than if they thought the program being
implemented was “The Program”.3 Further, there are ethical concerns
as to whether it is appropriate not to give a control group the benefit of
a program that may in fact benefit them. Similarly, an experiment dealing
with real people may lead implementers to avoid the experimental proto-
col in evaluation in order to do their jobs in what they deem to be the best
manner possible.4

Evaluation as feedback

In policy terms evaluation is important to improving public policies.

Although that policy dimension is central to evaluation, there is also
a more political use as feedback into the political system considered more
broadly. That is, policymaking it about policy, but it is also about the
role of the state in society and its legitimate position within that society.
If governments adopt successful policies they will be able to build their
legitimacy, which in turn will increase their capacities to make even more
effective policies. And the effects of policy may extend to private institu-
tions as well as to the public sector.

Daniel Béland (2010) has discussed six different streams of policy feed-
back into the state and the policy process. Although some aspects of these
six streams may be closely related, there are a range of significant impacts
 EVALUATING PUBLIC POLICY: AN INTRODUCTION 147

that do need to be considered. The first of these feedbacks is into the state
itself. For example, as Béland points out, programs such as Social Security
in the United States, and the social programs more generally in European
countries, have been useful for building the modern welfare state. And
further these programs have encouraged the creation of a host of interest
groups that advocate for benefits from the more active state.5 And indeed
the creation of public sector programs has generated analogous programs
in the private sector, with private pension schemes developing to supple-
ment the sometimes modest benefits created through the public program.

Perhaps the most familiar of the feedback effects of policy is the lock-in
associated with path dependency and historical institutionalism (Sydow
et al., 2009). The logic has been that if a program is successful it creates
positive feedback that helps to institutionalize that program and prevent
external challenges. The political aspects of lock-in can be reinforced by
programs such as pensions that create stocks of benefits based on prior
contributions, making reform difficult. Thus, attempting to produce
change in programs after periods of success runs up against both political
and administrative barriers.

Finally, and to some extent most importantly, feedback from public pro-
grams create the ideational foundation for other programs, and further
for the role of the state in providing types of public programs. Again, for
Social Security, the idea of security being provided through the public
sector provided a foundation for a range of other programs attempting
to provide income security for citizens. Other programs, such as guar-
antees of “full employment” in a number of countries following World
War II created the foundation for a range of other programs supporting
employment.

All these versions of feedback to some extent depend upon evaluations of

public programs, albeit more political than technical evaluations. More
technical evaluations may also be involved in the process as means of
justifying the programs and building a foundation for the political assess-
ment of the program. The examples above are largely positive, but the
reverse dynamics are also available if programs do fail, or do not live up to
popular expectations. Indeed, there may even be forms of negative lock-in
in which areas of intervention become defined as virtually impossible for
governments to intervene effectively. For example, if “Obamacare” were
148 ADVANCED INTRODUCTION TO PUBLIC POLICY

to fail it might be generations before American government would be

prepared to undertake health care reform again.

Although the majority of the discussion of policymaking has been of

formal political institutions and official actors, evaluation reminds us
that this is public policy. Therefore, evaluators and political officials
need to find means of making evaluation information available through
the media or on-line. This feedback to citizens is a crucial aspect of
democratic politics, whether it is done through the media or directly by
governments. The public may not be interested in all the technical detail
from sources like the Government Accountability Office in the United
States, or Riksrevisionsverket in Sweden, but they do want to know if the
schools are performing well, and the water is safe to drink. The pursuit
of accountability needs to be more than politicians and administrators
discussing programs and should allow the public into the conversation.

Evaluation and learning

We noted above that one of the major purposes of evaluation is for policy-
makers to be able to learn from past mistakes, or even form past successes,
in order to be able to make subsequent policies better. Although learning
should be a crucial outcome of the evaluation process, actually learning
or drawing lessons from policy (see Rose, 1993) is rather difficult. As
we have been noting throughout this book, the process of making and
implementing policies involves numerous variables and numerous actors,
so that determining what the lessons of success or failure are is difficult.

Although the difficulties associated with learning are manifest, the con-
temporary politics of policymaking emphasizes “evidence-based policy”
(Pawson, 2006; but see Hammersley, 2013). The assumption is that
evidence generated from policy initiatives in one setting can be used to
inform decision-making in others. Those difficulties are increased if there
is an attempt to transfer the lessons learned in one political or cultural
setting and to another. The extensive literature on policy diffusion (Braun
and Gilardi, 2006) points to these difficulties but also demonstrates that
with care learning is possible. Longer-term success in learning depends
in part on “deutro-learning”, or learning about learning (Visser, 2007).
 EVALUATING PUBLIC POLICY: AN INTRODUCTION 149

The question of learning about policy raises questions about the utili-
zation of evaluations. There is an optimistic assumption embedded in
the evaluation literature that bureaucrats and politicians will read the
evaluation reports, learn from them, internalize the findings and improve
the policy. As already noted, there are numerous barriers to effective eval-
uation and many of these depend on the willingness of the actors involved
in the process to utilize the information available. All the evaluation and
policy advice in the world will be of little utility if the decision-makers
involved in policymaking have ideological blinders, or other blinders, that
inhibit their use of the information generated. Indeed, given the commit-
ment that some politicians have to policies, there may more “policy-based
evidence making” than genuine use of evaluation results in policy.

Conclusions

It should be clear from the above that evaluation is both difficult and
necessary for public policymakers. The necessity arises from the need
for accountability, as well as from the opportunities to use evaluation
research to improve the quality of the programs being delivered. The dif-
ficulty arises in that public policies are themselves complex undertakings
involving multiple actors and also involve equally complex interventions
in the economy and society. Therefore determining success and failure is
difficult enough, and determining the sources of that success and failure
may be even more difficult.

This chapter has raised some of the general issues associated with eval-
uation, as well as identifying the numerous difficulties associated with
effective evaluation. The following two chapters will present two alterna-
tive directions for evaluation of policies. One, the more commonly used
approach, utilizes economic criteria to assess policies. The utilitarian
assumptions behind this model are that more is better, and that “more”
can be measured in dollars, pounds, euros or whatever. Therefore, if the
economic benefits of a program exceed the economic costs then it can be
said to be a good program.

The alternative approach to evaluation considers normative criteria

for policy, asking non-utilitarian questions about the choices made by
150 ADVANCED INTRODUCTION TO PUBLIC POLICY

governments in the formulation and delivery of public programs. This

approach is perhaps less satisfying for policymakers because it provides
less clear answers about the success and failure of programs. The nor-
mative approach to evaluation may depend upon the particular ethical
perspective of the evaluator, or perhaps on the normative priorities of the
formualtors and the clients of the program. Therefore, the answer to what
is good policy is often unclear.

Each of these approaches to evaluation can provide its own answers but
those answers may well be contradictory. Policymakers and evaluators
therefore may be put into the position of choosing between those two
approaches, or of finding some means of using the two together. The
superior strategy would be to attempt to put the two forms of evaluation
together, but that is easier said than done. The standards and even the
logic of evaluation are different, but both approaches can provide insights
into the quality of policy and its capacity to improve the lives of citizens.

Notes
1. The Commonwealth Fund (2014), for example, rates the National Health
Service in the United Kingdom as the best health care system in the world,
although it notes a deficiency in “keeping people alive”. This seems a cri-
terion that is more than a little relevant for the evaluation of a health care
system.
2. Headstart is an American program to prepare children from low-income
families for entering kindergarten. The evidence has been that students who
have been through this program are equal to their middle-class counterparts
but that the effects of the early enrichment decayed almost completely by the
sixth grade.
3. If they do not know that they are part of an experiment there is an ethical
issue for government. In most experimental situations participants would
have the option not to participate, but cannot do so if they are not informed.
4. For example, an experiment in means of addressing spousal abuse was
undermined (as a research exercise) by the tendency of the police to arrest
perpetrators, rather than to offer counseling. Not only is this the usual
pattern of intervention for the police but they felt that they needed to protect
the women involved.
5. Skocpol (1992) demonstrated how pensions for veterans provided after the
American Civil War helped to build demand, and support, for veterans
organizations.
8 Evaluating public policy:
the utilitarian dimension

The conventional mechanism for evaluating public policy is to assess

the economic consequences of the programs. That statement does not
mean so much attempting to understand the effects of the programs on
economic growth, inflation or other macro-economic variables, although
those effects may certainly be one component of the evaluation. Rather,
the most commonly used approaches to policy evaluation attempt to
measure the costs and benefits of programs in economic terms, usually
at a micro level. The next step is to compare the net of costs and benefits
against not only other programs but also in some absolute terms (Farrow
and Zerbe, 2013). Unlike some of the other forms of evaluation used in
the public sector, cost-benefit analysis tends to be prospective rather than
retrospective. This method is used to decide ex ante whether a program
is worth pursuing or not. Therefore, along with several other issues to be
detailed below, there is a very fundamental issue that much of this evalu-
ation is based upon speculation about the real costs and benefits that are
likely to accrue in the future. In this approach to policy it is important to
estimate not only how much benefit will be created but also when it will
be created, with the standard economic assumption that a dollar today is
worth more than a dollar next year. This prospective method of assessing
the effects of policy may have substantial inaccuracy, but it is perhaps
more useful for people considering policy than a more accurate analysis
five years after the program has begun.

In addition to the inaccuracy that may easily arise in this prospective anal-
ysis, this type of analysis may be easier to manipulate in order to produce
desired political outcomes. That manipulation need not be dishonest but
may arise out of sincere commitments to particular policies and styles of
making policy. This ability to manipulate, or perhaps more politely this
possible inaccuracy, results in part from attempting to value outcomes
that may be difficult to evaluate in market terms, from uncertainties about

151
152 ADVANCED INTRODUCTION TO PUBLIC POLICY

the timing of the costs and benefits, and generalized uncertainties about
the future.

The prospective nature of cost-benefit analysis is important because

this methodology is used to advise policymakers about what programs
to undertake and which to avoid. Most policy advice is inherently pro-
spective, but much of that depends on the experience and judgment of
advisors. Cost-benefit, and the associated methods (see below), depends
more on a specific methodology to provide that prospective information
about policy. It may still need to be filtered through the judgment of
experienced advisors, but this is a means of providing policymakers with
(what appear to be) hard numbers about policy choices.

The logic of cost-benefit analysis

Cost-benefit analysis is based on welfare economics and logic of the

consumer surplus. One part of the intellectual justification of cost-benefit
analysis and its emphasis on the creation of “more” comes from welfare
economics. The initial criterion for choice coming from welfare econom-
ics was Pareto optimality (Blaug, 2011). This criterion was that a move (in
this case a policy) would be justified if at least one person was benefitted
and no one else was harmed. The difficulty was that this is an extremely
conservative criterion, given that very few public programs can be imple-
mented without some losses for some citizens – notably through taxes.
This criterion therefore can be a recipe for doing nothing.

The so-called Kaldor-Hicks criterion for selecting moves (again policies)

provides an alternative to the Pareto criterion that is more conducive to
action in the public sector. This principle argues that a policy is socially
justifiable if it creates sufficient benefits that the winners can compensate
the losers, and still have some benefits for themselves. This does not mean
that they will provide that compensation, only that they could. Further,
this principle is blind as to who the beneficiaries of the policy may be, so
that one that creates large benefits for those who are already affluent is as
justifiable as one that creates benefits for the poor.
 EVALUATING PUBLIC POLICY: THE UTILITARIAN DIMENSION 153

A second fundamental idea underlying cost-benefit analysis is the con-

sumer’s surplus (Mishan, 1967). Stated simply, this surplus represents the
amount of money a consumer would be willing to pay for a given product,
minus the amount he or she must actually pay. Consumers tend to value
the first unit of a product or service they receive more highly than the
second, and the second more highly than the third: The first quart of milk
where there has been none is more valuable than the second. But units of
a product are not priced marginally, but sold at an average price, which
means increased production will give consumers surplus value. Thus
any investment that reduces the cost of the product or service produces
a benefit in savings that increases the consumer surplus. The govern-
ment’s investment in a new superhighway that reduces the cost to con-
sumers of driving the same number of miles – in time, in gasoline, and in
potential loss of life and property – creates a consumer surplus. And as the
time, gasoline and lives saved by the new highway may be used for other
increased production, the actual savings represent a minimum definition
of the social benefits created by the construction of the new highway.

The concept of opportunity costs is also important in understanding

cost-benefit analysis. This concept is based on the rather obvious fact that
any resource used in one project cannot be used in another. For example,
the concrete, steel and labor used to build a superhighway cannot be
used to build a new dam. Consequently, all projects must be evaluated
against other possible projects to determine the most appropriate way to
use resources, especially financial resources. Projects are also compared,
implicitly if not explicitly, with taking no action and allowing the money
to remain in the hands of individual citizens. Again, the basic idea of
getting the most “bang for the buck” is fundamental to understanding
cost-benefit analysis.

When identifying and assessing costs and benefits, the analyst must also
be concerned with the range of effects of the proposed program and the
point at which the analysis disregards effects as being too remote for
consideration. This concept is analogous to the concept of standing in
courts that gives individuals the right to sue for damages (Whittington
and MacRae, 1996).1 For example, building a municipal waste incinerator
in Detroit, Michigan, will have pronounced effects in Windsor, Ontario,
Canada, that must be considered – even though that city is outside the
United States. The prevailing air currents may mean that some ash and
acid from the incinerator also reaches Norway, but those effects may
154 ADVANCED INTRODUCTION TO PUBLIC POLICY

be so minimal that they can safely be disregarded. This form of analysis

requires making judgments about what effects are sufficiently proximate
and important to be included in the calculations. That judgement is also
to some extent political, so choosing which effects (positive or negative)
can determine the viability of a project.

Finally, time is important in evaluating costs and benefits. The costs and
benefits of most projects do not occur at once but accrue over a number
of years. If a highway is built, it will be serviceable for 50 years and will be
financed over 20 years through government bonds. Policymakers must
be certain that the long-term costs and benefits, as well as the short-term
consequences, are positive. This, of course, requires some estimation of
what the future will be like. We may estimate that our new superhighway
will be useful for 50 years, but oil shortages may reduce driving during that
period so that the real benefits will be much less. Or, conversely, the price
of gasoline may increase so much that the savings produced are more
valuable than assumed at present. These assumptions about the future
must be built into the model of valuation if it is to aid a decision-maker.

In part because of the uncertainty concerning future costs and benefits,

and in part because of the general principle that people prefer a dollar
today to a dollar next year, the costs and benefits of projects must be
converted to present values before useful cost-benefit calculations can
be made. That is to say, the benefits that accrue to the society in the
future have their value discounted and are consequently worth less than
benefits produced in the first year of the project. Costs that occur in
future years are likewise valued less than costs that occur in the first few
years. Cost-benefit analysis thus appears to favor projects that offer quick
payoffs rather than greater long-term benefits but perhaps also higher
maintenance and operation costs. Although there may be good logical
justification for these biases, they influence the kinds of programs that
will be selected, and that has definite social implications, not least for
future generations. Other analytic aids for government decision-makers,
such as “decision trees”, include probabilities of outcomes, to cope with
the uncertainties of the future, but cost-benefit analysis tends to rely on
discounting future costs and benefits.
 EVALUATING PUBLIC POLICY: THE UTILITARIAN DIMENSION 155

Measuring costs and benefits

The most fundamental problem in the actual application of cost-benefit

analysis, leaving aside the question of the ranking of outcomes, is the
measurement of costs and benefits. To do this requires placing many
non-economic outcomes on the common measuring rod of money. It
is (relatively) easy to assign a monetary value to outcomes such as less
energy used when there are better highways, or even the time that com-
muters may save each day from those better roads. It is, however, much
more difficult to assign a monetary value to the lives saved because of
a safer road, or the environmental losses from building that road. But to
make the method work we have to be able to assign such values.

The critics of cost-benefit analysis and allied methods have argued that
attempting to assign values to non-marketed factors, especially those such
as environmental qualities, is “nonsense on stilts” (Self, 1975). By their
very nature, it is argued, these items are not marketable and therefore
assigning economic values is logically incorrect and tends to provide the
illusion of precision. But the idea is not only common in cost-benefit
analysis, it has become enshrined in laws concerning the justification of
new public programs (Miller, 2005).

The above being said, there are several standard ways of thinking about
assigning values to seemingly non-economic costs and benefits. One is
willingness to pay. How much would the average citizen be willing to pay
for living in an area with clean air as opposed to one with high levels of
pollution? This method, referred to as “hedonic pricing”, can be deter-
mined simply by looking at real estate values in different parts of a city,
or perhaps in roughly comparable cities with different levels of pollution.
Awards made by courts for loss of amenity values and damages also
provide a measure of the value of environmental damage. Finally, these
prices can be determined through surveys, asking respondents what they
might be willing to pay for public goods such as clean air or low levels of
noise (Mitchell and Carson, 2013).

Valuing life is perhaps even more difficult, although many common

public sector projects do involve both the loss and the preservation of lives
(Frakt, 2020). As already noted, building a new road may save lives, but
it may also involve the loss of life for some construction workers. What
156 ADVANCED INTRODUCTION TO PUBLIC POLICY

are those lives worth? The courts also can provide a partial answer here,
when the legal system permits claims for wrongful death and judges or
juries make awards to survivors.2 Even these methods, as fair as they are
intended to be, appear to many to represent rather heartless choices. That
said, as we will point out in the following chapter, simply saying that pre-
serving life is an absolute value tends to vastly oversimplify the difficult
choices that must be made when designing public policies.

Problems in cost-benefit analysis

Although cost-benefit analysis is the standard means of evaluating public

policies, it is not without its problems. In addition to the technical
issues already mentioned, there are more fundamental issues with this
methodology. Indeed, for some critics the utilitarian logic at the heart of
this model is itself a problem, given that this logical foundation tends to
privilege programs with demonstrable economic benefits, and tends to
devalue programs that may have less visible benefits.

Following from the above, another of the fundamental issues in this

methodology is that the distributional consequences of the programs are
largely irrelevant to the evaluation of programs (Loomis, 2011). As already
noted, the basic logic of the method is that more is better, but it does not
matter who gets the more. Further, when the benefits are produced does
matter, and the welfare of future generations may be discounted in this
model. Therefore we need to consider some of the basic issues that affect
the utility of cost-benefit analysis, despite its appeal on many grounds.

Efficiency as the dominant criterion

Like most economic analyses, cost-benefit analysis relies on efficiency as
the dominant goal of policy. The assumption is that society will be better
off if programs are efficient and produce greater net benefits for society.
While that is certainly an important value, it is by no means the only
concern of the public sector and other values need to be weighed against
this economic criterion. For example, for some programs effectiveness
may be more important than efficiency. The program may be of sufficient
concern to citizens to deliver even if it cannot be justified readily on these
 EVALUATING PUBLIC POLICY: THE UTILITARIAN DIMENSION 157

economic grounds. Analysts may say, for example, that it is cost-effective

to close small firehouses in neighborhoods but citizens rarely will accept
that economic logic. And equity as a criterion is also crucial for programs
being implemented through the public sector.

Although there are numerous criteria that can be applied to public pro-
grams, the diffusion of neo-liberalism and its associated programs of the
New Public Management have made program efficiency seemingly more
acceptable. Whereas cost-benefit analysis focuses on the efficiency of the
design of a program, the New Public Management focuses on the effi-
ciency of service delivery. Taken together, these approaches to policy and
governance emphasize the one value at the expense of many others when
thinking about the public sector.

The discount rate

As noted above, the costs and benefits of programs accrue over time, and
there must be some means of placing these on a common measuring rod.
To do this both benefits and costs must be discounted to present value.
And to do that we need a discount rate – in essence an interest rate – that
can be applied to future values. The question then is what rate to apply.
The simplest answer is simply to apply the market interest rate existing at
the time, treating the public program just as one would any other type of
economic activity.

There is also an argument to be made for a social discount rate that

is different from the market rate. After all, the public sector is not the
market, and there are a range of concerns that may alter the way in which
we evaluate future costs and benefits . For example, it could be argued
that the appropriate social discount rate should be zero. After all, public
benefits are beneficial no matter when they are produced, and reducing
them to present value may undervalue them. And imposing market values
on public projects may prevent programs that can create some benefits for
society from being implemented, especially if those benefits accrue far in
the future (see below).

The selection of a discount rate then is both a political and an economic

activity and it may not be appropriate to permit market logic to dominate
the choice (Moore et al., 2004). The zero discount rate mentioned above
may make a great deal of social sense even if it certainly does not conform
158 ADVANCED INTRODUCTION TO PUBLIC POLICY

to market logic. There are real opportunity costs for capital that could be
used in a public project, and a discount rate that is artificially low may not
in the long run produce the greatest benefit for the society.

Inequality
As I noted above, the fundamental logic of cost-benefit analysis is that
more is better. There is some logic in that position, given that if there are
more resources available in society then all everything else being equal the
society will be better off. But cost-benefit analysis is largely blind to the
distribution of the resources being created through the programs it anal-
yses. If all the benefits created in the program are concentrated in a few
people, and even a few affluent people, the logic of the evaluation system
is that of the Kaldor-Hicks criterion – there is enough that those winners
could compensate the losers.

The distributional blindness of cost-benefit analysis is especially unfor-

tunate as a method for assessing public programs. If, as has been argued
above, one of the justifications for the intervention of the public sector
into the economy and society is to rectify the skewed distributions of
income and other assets produced by the market, then certainly the
distributional consequences of a program should be considered. Further,
although a fairer distribution may be an end in itself it may also have
important secondary consequences that could justify its inclusion in the
assessment of the impacts of programs.

There are ways in which the distributional consequences might be

included in the analysis of a program (see Brent, 2006, chapter 10). For
example, if one begins with the assumptions of Keynesian economics that
money going to less affluent people will have a higher multiplier because
of a higher propensity to consume, then some measure of differential
economic consequences could be gained. The assumption would be that
this higher marginal propensity to consume is indicative of the greater
value that the income has to a low-income individual. Rather than being
a luxury that can be saved, for a low-income citizen the marginal dollar or
euro is important for their livelihood.

Even if the assumed positive benefits of redistribution, or at least more

equitable distribution, cannot be measured in this manner it might still be
included by monetizing the benefits in the same way as other non-traded
 EVALUATING PUBLIC POLICY: THE UTILITARIAN DIMENSION 159

goods and services. For example, we know that reducing poverty also
tends to reduce crime in a society. We also know that lower levels of
poverty tend to be associated with more stable families, that in turn is
associated with increased probabilities that children will attend school
and graduate. Therefore, producing a more equal distribution of income
may have real benefits for society as a whole as well as for the individuals
who are made better off by the policies.

Inter-generational equity
Most of the discussion of equity and the distribution of benefits focuses on
the distribution of benefits across social classes or perhaps geographical
regions. There are also important questions of inter-generational equity
and externalities (Hepburn, 2007). As already noted, the standard model
of cost-benefit analysis discounts future costs and benefits, and therefore
emphasizes what happens for the generation that is making the decisions,
rather than future generations. This model is perhaps good economics but
it raises a number of other issues that are important for policy analysis.

The most obvious issue arising is how to protect future generations from
a number of negative outcomes from contemporary decisions. Some of
those outcomes may be absolutes rather than relative levels of cost and
benefit. Depletion of resources such as fossil fuels, for example, may be
difficult to cost but has tremendous implications for the well-being of
future generations. And climate change implies the extinction of species
and perhaps submerging cities or even whole countries. Any attempt to
assess outcomes of those magnitudes in simple economic terms would
indeed be “nonsense on stilts”, in the words of Peter Self.

While the potential negative outcomes of contemporary action are

perhaps more obvious, there can also be positive outcomes in the future
that might be extremely undervalued utilizing a simple economic assess-
ment. For example, investment in research on sustainable energy could
produce a bonanza for future generations, with current generations
reaping few if any of the benefits of the investment. Again, assessing pol-
icies may require more explicit concern about the future, and about suc-
cessive generations, than is conventionally done when using cost-benefit
analysis.
160 ADVANCED INTRODUCTION TO PUBLIC POLICY

There is ample evidence that most political systems are not good at
thinking about the future, and the long-range implications of their policy
choices (Jacobs, 2011). This short-term perspective on policymaking is
perhaps to be expected in democratic governments, given that political
leaders, or at least their parties, want to be re-elected. Bureaucrats have
been able to take a longer-term perspective on policy but the increasing
use of performance management tends to shorten their time perspective
as well.

One of the more interesting approaches to the longer-term implications

of policy choices is the “precautionary principle” that has been at the
heart of a good deal of European Union policymaking about issues such
as genetically modified organisms (GMOs). The logic of this principle, as
the name implies, is that decision-makers should exercise caution when
adopting programs given that there may be long-term and irreversible
implications (O’Riordan and Cameron, 2013). While this principle does
indeed take the future into account it may also be a barrier to innovation,
requiring some balancing of important principles.

Extensions of cost-benefit analysis

While the basics of cost-benefit analysis are useful for evaluating the
utility of policy interventions into the economy and society, there are
several extensions of the basic model that also merit some consideration.
These methods are all based on the same basic utilitarian logic but apply
that logic in somewhat different ways.

Risk-benefit analysis
The framing of cost-benefit analysis is typically in terms of seeming
certitude. The costs and benefits are assumed to be identifiable and there
is an implicit assumption that they will indeed occur if the program is
adopted. The real world of policy, however, is rarely that exact and there
are real probabilities of unplanned or unanticipated events. Therefore, in
addition to coping with planned interventions and planned outcomes,
the unforeseen risks that face any public program need to be considered.
For example, when building a dam or a flood control system what level
 EVALUATING PUBLIC POLICY: THE UTILITARIAN DIMENSION 161

of rain and flooding does an agency plan for? Even if the agency plans for
the flood that occurs once every 100 years – a seemingly safe assumption
– there is still a real chance for the 500-year flood.

Adding other dimensions of analysis

Cost-benefit analysis is based on economic criteria and fundamental
economic assumptions. It is difficult to argue that economic variables are
not important for any policy, but it is also difficult to argue that these are
the only criteria of relevance. Therefore, there have been suggestions that
the methods should be extended to include other important considera-
tions. For example, Simpson and Walker (1987) argued for adding both
technical, environmental and risk factors to the analysis to provide a more
complete picture of the benefits and costs of programs. These factors
would enable the analyst to understand not only the economic impacts
of the program but also take into account the risks of other outcomes and
especially the impact of the program on the environment.

Matthew Adler (2019) has argued for measuring the welfare benefits of
a public policy in a more comprehensive manner. Rather than focusing
on a simple additive approach to costs and benefits, this approach would
attempt to take into account the preferences of the recipients as well. The
approach would also attempt to include some means of assessing out-
comes on an ethical as well as a monetary basis (see below).

While adding a range of other factors to the analysis of a possible public

program is in many ways appealing, it raises other questions. The first
is how does the analyst put together information from these alternative
dimensions with the economic results of cost-benefit analysis. The prob-
abilities are that these different forms of analysis will provide different
answers expressed in different terms, leaving the policymaker with the
need to exercise his or her judgment. The probabilities are then that the
seemingly “hard” answers from cost-benefit analysis will dominate the
other forms of analysis and in the end little will have changed.

Regulatory impact analysis

Cost-benefit analysis appears particularly applicable to public programs
involving capital investments or other large-scale interventions by gov-
ernment into the private. sector. While programs of these nature do
162 ADVANCED INTRODUCTION TO PUBLIC POLICY

have major impacts on the society, there are other forms of public sector
activity that also affect the economy and society. While the fundamental
logic of analysing these interventions may be the same as cost-benefit the
impacts may be more difficult to estimate and may be confined more to
the economy and society than to the public sector itself (Arrow, 1996).

Regulatory impact analysis has been implemented in order to assess the

effects of public sector regulation. This approach to regulatory account-
ability was initiated in the Reagan administration in the United States,
and then was elaborated by subsequent administrations (Hahn and
Sunstein, 2002). This concern with the impacts of regulation has been
adopted in a very significant manner by the Organisation for Economic
Co-operation and Development (OECD) in a series of regulatory reviews
in the member countries (OECD, 2010). These analyses, as well as some
by other international organizations, do tend to operate from an assump-
tion that less regulation is better, an assumption that favors businesses but
not necessarily citizens or the environment.

Environmental and social impact analysis

Following on from the point above about adding different dimensions
to the analysis of policy, there are alternative tools that can be applied to
policy choices that can supplement, or substitute for, economic analysis.
The most commonly used of these alternatives is environmental impact
analysis. The National Environmental Protection Act (NEPA), passed
in the United States in1969, mandated that for many federal programs
government agencies would have to perform an assessment of the envi-
ronmental impacts of that project (Taylor, 1984; Jay et al., 2007). The
purpose was to attempt to be sure that federal programs conformed to
NEPA, and that decision-making was done with an awareness of impacts
of a program – adverse environmental impacts would not necessarily veto
a project but the decision-makers did have to justify continuing with the
program. The Trump administration has, however, seriously weakened
these reviews (Friedman, 2020).

Although the environmental impact statement was an American inven-

tion the methodology, or at least the fundamental idea, has been diffused
widely. For example, the European Union requires an Environmental
Impact Assessment (EIA) or a Strategic Environmental Assessment
(SEA) for at least as broad a range of public sector activities as does the US
 EVALUATING PUBLIC POLICY: THE UTILITARIAN DIMENSION 163

government. This is true for projects funded by the European Union itself
as well as for some programs initiated by the member states. Although
there are some differences in the details of these assessment as compared
to the EIS in the United States the basic idea of some ex ante analysis of
programs for their effects on the environment is the same.

Social Impact Analysis (SIA) is an analog of Environmental Impact

Statements and became part of law soon after the EIS was developed. The
purpose of the SIA is to examine the unintended consequences of pur-
poseful interventions, that is, the unanticipated effects of public policies
(see above). In particular, this method focuses on communities, and on
issues of distributive justice in those communities (Interorganizational
Committee, 2003). The method also contains a strong emphasis on data
collection and analysis in order to provide relatively objective analyses of
the effects. And like EIAs this approach has been diffused widely, espe-
cially to the European Union.

Summary

Cost-benefit analysis and its various off-shoots are the principal means
for assessing ex ante the value of public programs. The methods use an
economic logic as the foundation for that assessment, assuming first that
“more” is always desirable, and further that everything that governments
do can be reduced to a common metric of money. Further, this model of
evaluation has a decided preference for benefits produced in the short
run, and for costs that accrue in the long run.

These basic characteristics of cost-benefit analysis tend to make some

programs appear more desirable than others, based on that economic
logic. The capacity of this method to give decision-makers a clear answer
in pounds and pence, however, gives the method a great deal of appeal
for decision-makers as a means of justification as well as an aid to making
decisions. As the following chapter will point out, however, there are
alternative ways of evaluating policies that need to be considered. These
alternatives do not provide the neat answers that cost-benefit anlysis does,
but they do require decision-makers to think more expansively about the
choices they are making in the name of the public.
164 ADVANCED INTRODUCTION TO PUBLIC POLICY

Notes
1. R.O. Zerbe, “A place to stand for environmental law and economic analysis”,
http://​www​.cserge​.ucl​.ac​.uk/​Zerbe​.pdf (accessed 8 August 2014).
2. This is, of course, most developed in the United States but in other legal
systems there may be administrative rulings that can substitute for judicial
decisions.
9 Normative and ethical
analysis of policy

The previous chapter has provided one means of deciding on what is

a good public policy. This chapter will provide an alternative to the
utilitarian logic inherent in the cost-benefit model. Indeed, the search in
this chapter is for mechanisms for assessing policies that depend upon
ethical standards, but which can still provide decision-makers with usable
advice about what decisions to make. One principal virtue of cost-benefit
analysis is that it provides those decision-makers with a definitive answer
about whether a policy is viable (within the assumptions of that model of
course). Unfortunately, the answers coming from the normative analysis
will be somewhat more ambiguous.

In this chapter I will present a series of criteria that are important

for assessing policies from the normative perspective. The logic of
cost-benefit analysis and other utilitarian analyses of policies is that we
should assess programs based on who gets what, and the value of those
benefits – the consequences of the programs. This chapter will be con-
cerned more with the “logic of appropriateness” of the policies than with
the “logic of consequentiality” (March and Olsen, 1989). Phrased differ-
ently, there is a difference between a deontological approach that posits
certain absolute prohibitions and requirement and an approach that is
just concerned with the consequences of those actions (see Adler, 2019).
That is, some decisions about policy may need to be made simply because
they are the correct thing to do, regardless of the consequences.

As already implied, there is no clear and agreed upon means of selecting

among these various principles, and each does provide a certain impor-
tant perspective for policy choices. These are all important criteria for
making policy choices but they are not necessarily consistent and may
even be contradictory. Unlike cost-benefit analysis this form of evaluation
does not provide a single bottom-line but rather requires policymakers,

165
166 ADVANCED INTRODUCTION TO PUBLIC POLICY

and perhaps also citizens, to make judgments of their own about the
appropriateness of policy decisions. The various normative criteria that
can be applied to policy action could be summarized as the “public
interest”, but even that criterion requires a great deal of elaboration (Ho,
2011). What follows here is a catalog of several of the more important
normative criteria for policy choices, as well as some discussion of the
difficulties in applying these criteria.

Fairness

Perhaps the fundamental normative criterion for public policy is fairness.

What the public typically want from their government are policies that
are fair, and treat citizens in an equitable manner (Sen, 2009). Citizens
may be willing accept some deprivations, such as higher taxes or loss
of services in times of crisis, provided they believe the policy treats all
citizens equally, and is being implemented fairly. For example, rates of tax
evasion appear lessened if citizens believe that the taxes and their enforce-
ment is fair (Slemrod, 2007). Further, people will cheat on their taxes
if they believe other people are cheating – that is only fair. In addition,
governments are responsible in the mind of most people for maintaining
individual rights and fairness among individuals through law, and doing
so through equal enforcement.

While citizens demand fairness from their governments, defining what

constitutes fairness is much more difficult. Being human we often think
that what benefits us is fair, but programs that benefit other people, and
especially people whom we may not particularly like, are not fair. For
example, there is a good deal of evidence that people believe that the tax
expenditures that benefit them are fair, but others are not (Fancy, 2011).
Other people may think that the distributions produced by the market are
inherently fair,1 so that if government intervenes those actions produce
unfairness.

If fairness is to be at all useful as a criterion of evaluating public policies

it is important to develop some standards that go well beyond these per-
sonal conceptions of fairness. Removing that individual element from the
definition of fairness is central to one of the most important standards
 NORMATIVE AND ETHICAL ANALYSIS OF POLICY 167

of justice developed in contemporary political philosophy. John Rawls

(1971) developed a conception of justice based on the “veil of ignorance”.
The central question in this approach is what distribution of benefits and
costs would individuals select if they were able to make that choice not
knowing where they would actually be in the final distributions of goods
and services produced by the market and by government.

The assumption behind the Rawls approach to fairness is that individuals

making decisions behind this veil of ignorance would choose more equal
distributions than those generally produced by the market. While an
individual may enjoy being at the top of such a distribution of goods and
services, they in general would be unwilling to take the risk of being at the
bottom. Selecting a more equal distribution ensures that the individual
would not encounter the risk of being deprived severely. Of course, more
risk accepting individuals might be willing to take the chance of “winning
big” in the distribution of rewards, but for most citizens avoiding serious
deprivation is a preferable alternative.

Of course, we can never really make decisions behind such a veil of igno-
rance, and governments with complex decision-making processes gener-
ally attempt to calculate in advance just who will benefit and who will lose
from any policy choice (see Chapter 8). Therefore, while this definition
of fairness is an interesting intellectual exercise, it is perhaps not of much
utility to practical policymakers. Indeed, much practical politics is about
using various forms of political power to shift the burdens of government
onto others, and to capture the benefits of government for oneself.2

Although the veil of ignorance is a useful intellectual exercise, Rawls

has supplied another means of considering fairness in policymaking.
Although he ranks equality of opportunity as the primary criterion
for achieving justice, his “difference principle” is also closely linked to
the pursuit of justice in the public sector (Rawls, 2001). The difference
principle argues that in the absence of any other information or criteria,
policy choices that minimize differences within the society are preferable.
Given the assumption that individuals would choose a more equal set of
outcomes in the absence of other information about their own position,
pressing for such a distribution of outcomes therefore should be consid-
ered fair.
168 ADVANCED INTRODUCTION TO PUBLIC POLICY

In political terms there are debates about the equality of opportunity as

opposed to equality of outcomes. More conservative political regimes,
and actors, stress creating opportunities for individuals rather than
demanding that all citizens necessarily receive equal benefits in all policy
areas. That said, there are some areas of government activity, for example,
the enforcement of civil liberties that do require more strict equality.
Thus, saying that fairness and equality are important values for public
policy is only the beginning of political debates about the domain of
policies to which those values apply, and how more equal societies can be
produced (see, for example, Jacobs and Skocpol, 2005).

Autonomy and freedom of choice

As noted above, John Rawls ranked equality of opportunity and the

capacity for choice as the most important of the values for justice. He
argues that individuals should have the right to be treated as individuals
and as adults, and that a just society will provide “fair opportunities” for
all citizens. There are always instances in which government must act in
loco parentis to protect children and other vulnerable members of society
from their own decisions, as well as from others. Political and admin-
istrative leaders may also believe that ordinary citizens are incapable of
processing complex policy issues and therefore justify a lack of transpar-
ency about those issues. And administrators and policy advisors may have
some of the same negative evaluations of politicians and feed them only
a limited about of information as decision premises.

The statement from Rawls indicates clearly that one important ethical
concern for designing public policies is that individual citizens should be
able to make up their own minds about matters that affect their own lives.
In some areas that exercise of choice is very easy to permit. For example,
citizens do not have to claim their pensions or their unemployment bene-
fits. Most do, but they do not have to. Likewise, farmers could choose not
to accept payments that subsidize their farms. Even under programs such
as “Obamacare” that are considered by many to be impositions, citizens
may select one of a number of possible programs of health insurance.
 NORMATIVE AND ETHICAL ANALYSIS OF POLICY 169

At the same time that there are numerous choices available to citizens,
in many other cases those choices will be limited. An individual citizen
will be able to select from among a number of health insurance programs
under the Affordable Care Act, but before the Supreme Court removed
the mandate he or she had to make a selection or be fined. Citizens gen-
erally cannot avoid (legally) paying their taxes, although the clever use of
exemptions and deductions will significantly lessen the obligations (Alm
and Togler, 2011). And in time of war citizens may not be able to escape
military or other national service.3

The ethical, and political, question concerning maximizing individual

choice therefore becomes where to draw the line in terms of individual
choices versus use of the power of the state to ensure at least minimal
consumption of some public programs. For example, while we may allow
pension recipients not to accept their monthly payments from govern-
ment, we become more skeptical when parents refuse benefits for their
children. And those parental behaviors become even more suspect when,
as in refusing vaccinations, their choices may have societal consequences
as well as consequences for the children (Isaacs et al., 2009).

In economic terms the concept of “merit goods” is used to describe goods

and services to which individuals or the society should have access regard-
less of their capacity, or willingness, to pay (Musgrave, 1957). Obvious
examples would be public education, nutrition, public health protections
and a range of other benefits. That range may depend upon the political
and ideological perspectives of the individual preparing it. Similarly,
“demerit goods” might be used to describe social and economic outcomes
from which members of society should be protected, whether through
regulation or prohibition.4

Even in the case of merit goods that most of the society assume should
be consumed by all citizens, there may still be individual objections. For
example, parents may not want to vaccinate their children even though
this has value for the society as well as individual children. Should they be
allowed not to consume this merit good? Some of the answer may depend
on the basis of the objection. Some religious groups do not believe in
vaccination and may have an exemption but individuals who have other
objections may be forced to vaccinate. The “Hobby Lobby” case in the
United States (Burwell v. Hobby Lobby Stores, Inc) on contraception may
170 ADVANCED INTRODUCTION TO PUBLIC POLICY

be the extreme version of the use of these exemptions from law (Carroll,
2014).

The design of public policies may be based on individual autonomy and

choice, or it may be more controlling. Some instruments used to deliver
public policies restrict the choices of individuals while others permit
more choice, At the extreme, instruments such as education vouchers
permit individual citizens to make choices very much as if they were in
a marketplace, albeit with some restrictions such as a mandate that a child
must receive some form of approved education. In other policy areas
such as safety or public health the instruments used may be “command
and control” given the dangers to individuals or the society more broadly.

The increasing use of policy instruments based on “nudges” or other

subtle psychological mechanisms (see above; Hausman and Welch, 2010)
raises additional questions about individual choice and autonomy. When
confronting a command and control policy instrument the citizen is
aware that he or she is being coerced, and can react accordingly.5 When
the instrument employed is more manipulative and is largely hidden from
the casual observer then the citizen is no less coerced but may have little
opportunity to even evaluate the nature of the coercion being applied.
Ethically this covert coercion is perhaps even worse than other coercive
instruments because it undermines autonomy more completely.

Preservation of life

Preserving life itself is obviously as important as preserving the autonomy

of individuals while they are alive (Sunstein, 2014). We saw above that
one of the challenges for cost-benefit analysis is valuing lives so that these
extreme consequences of policy choices could be added into the net of
costs and benefits of the program. The problem for ethical analysis is in
some ways more difficult, given that taking lives has to be justified on
some more fundamental moral basis (Glover, 1977), rather than being
simply a byproduct of other actions. Likewise, although we tend to
consider life as invaluable individuals and governments are constantly
making decisions that take lives, or fail to save lives, and we should con-
sider the basis on which those decisions can be justified.
 NORMATIVE AND ETHICAL ANALYSIS OF POLICY 171

One of the most common reasons that is used to justify taking lives is
the preservation of the society. So, members of the armed forces may
be asked, or required, to sacrifice their lives in order to preserve their
country and their fellow citizens, and policemen and firemen also risk
their lives regularly to protect the society as a whole. Likewise, countries
have justified the death penalty as a means of preserving society (often
with the assumption that this extreme punishment would prevent future
violence). Similarly, police officers may be sanctioned to use deadly force
in the defense of the society, as well as in self-defense.

In some societies taking lives has been justified in terms of maintaining

the autonomy of the individual. For example, policies permitting indi-
viduals to take their own lives, or permitting physicians to take the lives
of terminally ill patients, may be justified on the grounds of autonomy;
they permit citizens to make decisions about their own lives (Rachels,
1986; MacMahan, 2006). The precept of maintaining autonomy and the
defense of life may obviously be in conflict at times. Laws against suicide,
and especially assisted suicide, emphasize the value of life. In other cases
such as Terri Schiavo in the United States, government may attempt to
preserve life against the explicit requests of the family of an individual
who is brain dead (Quill, 2005).

Another common question concerning life versus autonomy concerns

using safety devices such as seatbelts or helmets for motorcyclists.
Individuals argue that they should have the right to assume the risks of
not using those devices, while governments generally attempt to save their
lives by mandating their use. In addition to maintaining life and health,
government may also justify their impositions through the costs that inju-
ries and death may impose on society, as well as lessening the suffering of
family members of those sufficiently reckless to ignore the safety devices
available.

Another increasingly important issue posing conflicts between the values

of preserving life and individual autonomy concerns the rights of individ-
uals to control the use of their own organs. The general practice in most
countries has been that individuals have to sign donor cards of some sort
in order for their organs to be available for transplant in case of their death
or imminent death (Truog, 2008). The need for organs to preserve life,
however, is pushing some governments toward rules that assume that the
172 ADVANCED INTRODUCTION TO PUBLIC POLICY

individual gives consent to using organs unless the individual explicitly

opts out.

Another of the most contentious issues in the preservation of life depends

upon the definition of when life begins. The intense debate over abortion
in any number of countries has pitted those claiming that life begins at
conception against those claiming that life only begins with the viabil-
ity of the fetus (Halliburton, 2014). And these debates may extend to
techniques such as in vitro fertilization that are argued to interfere with
natural processes. For some participants in the debate, however, these
technical details are irrelevant, with the right of a woman to control her
own body and reproductive life as paramount values. These values cannot
be expressed in the financial values of cost-benefit analysis but depend on
more fundamental values.

Maintaining life as a political issue also confronts the conflicting values

placed on visible lives and less visible but equally valuable lives. This may
appear when a woman’s life is threatened by continuing a pregnancy,
but may arise even more obviously when a mountain climber is stranded
on the side of a remote mountain.6 Governments will spend significant
amounts of money attempting to save that individual, even if he or she
got into the position largely by their own volition. The money used in the
rescue would, if used to improve highways or for other safety programs,
could potentially save hundreds of lives. What is the correct use of this
money, based on a principle of preserving life?

Although the preservation of life is a fundamental value its application to

public policy issues is by no means simple nor unambiguous. This value
cannot be considered in isolation from other important values, especially
the preservation of individual autonomy. Therefore, decisions about pre-
serving life become clouded with numerous different values involved. It
is easy to say that life is absolute and sacred but in reality individuals and
governments make decisions that take or threaten life. As with so much
of policymaking, the choices being made are complex, contradictory and
often incomplete.
 NORMATIVE AND ETHICAL ANALYSIS OF POLICY 173

Stigma

Equality is one of the underlying principles of fairness, but societies

sometimes find maintaining equality difficult. In some societies people
may be stigmatized because of race, or religion, or characteristics such
as physical deformity or mental illness. But the actions of governments
and their public policies may also be the source of stigmatization and
poor treatment by the members of society. Almost all citizens in modern
governments are the recipients of some form of government benefits.
These may be well disguised, as when more affluent citizens receive their
benefits in the forms of tax expenditures. Although citizens are all benefi-
ciaries, programs are not delivered in ways that are necessarily equal, and
the recipients of some programs are treated poorly at the point of delivery,
and simply receiving some social benefits may make some citizens appear
less deserving, and even less real citizens, in the eyes of the public and
especially in the eyes of some political elites.7

Being the recipient of means-tested benefits often means that other

members of the society will consider the recipients as less worthy
because of those benefits. The assumption, especially in societies such
as the United States and to some extent the United Kingdom, is that the
individual is somehow less capable and worthy of respect than are those
who do not receive benefits. There is a long tradition of characterizing
the “deserving” and the “undeserving” poor in society (Handler and
Hasenfeld, 2006; Erler, 2013), with the elderly and children being consid-
ered deserving of concern while individuals of working age who were not
supporting themselves being considered undeserving.

Everything else being equal, policies should be designed in ways that

minimize the stigmatization of the recipients of the benefits. This means
primarily not using means-tested programs to deliver social assistance,
or if means-tested programs are used they should be administered in
ways that do not make the recipients visible to the remainder of society.
The problem is that social assistance programs are almost by definition
means tested, meaning that designing programs to hit the target popu-
lation effectively (Schneider and Ingram, 1993) without stigmatization is
difficult.
174 ADVANCED INTRODUCTION TO PUBLIC POLICY

In addition to the stigmatization that may be produced simply by the

nature of the program, there may be additional difficulties produced
for the beneficiaries of programs through the manner in which they are
delivered. A number of studies from a variety of cultures (Dubois, 2010;
Auyero, 2012) have demonstrated that the recipients of social programs
are not treated as well as customers in the marketplace or the recipients of
other public programs. T.H. Marshall (1964 ) once commented that the
welfare state would not be achieved until people standing in line for the
dole were treated in the same manner as people standing in line for opera
tickets. If that is true, then there is a very long way to go before the type of
equality foreseen by Marshall is achieved.

There may be ways of delivering programs to the less affluent, or to any

other group that may be stigmatized, without creating that stigma. One
mechanism for reducing stigmata is through using universal benefits
rather than means-tested benefits. One example of a program of this
nature is the child benefit used in most European countries (Alesina and
Glaser, 2004), with parents8 of children regardless of their income receiv-
ing a monthly allowance. For the more affluent this becomes taxable
income while for the less affluent it is a direct benefit. The other option is
tax-based programs such the earned-income tax credit (Alsott, 1995) that
provides benefits through the tax system, just as many affluent citizens
receive benefits through tax expenditures.9

Stigma is a particular problem for social assistance programs, but it is part

of a more general problem in policy design of labeling. That is, by creating
categories within which individual cases are slotted, public policies may
create not only stigma but also self-fulfilling prophecies about the behav-
ior of individuals. For example, labeling individuals as “disabled” at once
can create rights for those individuals and also label them as incapable
(Bagenstos, 2000). Likewise, placing children in “remedial” classes may
provide them with a suitable education but also label them as less capable
than their peers, and may reduce the probabilities that they will ever
succeed in the educational system.

Thus, stigma is about the formal criteria required to receive benefits, but it
is also about the manner in which people are treated by their government,
and by their fellow citizens. Governments can remove formal barriers to
programs and make them more universal, but at times attempts to assist
citizens may produce negative social and psychological outcomes for
 NORMATIVE AND ETHICAL ANALYSIS OF POLICY 175

those citizens. If benefits can be monetized fairly and made universal that
is beneficial but governments have difficulties in controlling the often
inadvertent behavior of citizens as they relate to other citizens.

Thus, as already noted, the problem of stigma and labeling presents

a dilemma for public policy designers. On the one hand, targeting pro-
grams to specific population represents good design, providing a benefit
to only those eligible but not wasting resources on those who are not. In
doing this, the programs may inadvertently, or perhaps intentionally,
stigmatize some groups within the society. Weighing these two criteria
for good policy – one administrative and more ethical – is but one of
the numerous challenges confronting policymakers. The policy solution
selected is dependent on ideology, culture and politics but those choices
will still have to be made and the consequences accepted.

Truth-telling, many hands and dirty hands

Policies may be designed in ways to enhance their fairness and the auton-
omy of citizens, but they also have to be implemented. The implementa-
tion process also involves a number of challenges to an ethical delivery
of public policies. Some of the above values, for example, stigma, may in
reality be as much a function of the manner in which programs are deliv-
ered as they are of the underlying conception of the programs. Policy and
administration may be treated as separate academic enterprises, but are in
reality intertwined, and administration affects the success of programs as
well as the impact of programs on citizens.

One of the crucial ethical issues for the delivery of public programs is
“dirty hands” (Walzer, 1973; Coady, 1991). Although governments and
individual public servants would like to be able to make decisions openly
and ethically, there are times at which more difficult choices must be
made. The problems of having to act in less than a fully ethical or truthful
manner is usually considered to be a particular issue for international
politics, as when diplomats must prevaricate on behalf of their country
and national security reasons are used to restrict access to information
(Thompson, 1999). Even at the domestic level governments must some-
176 ADVANCED INTRODUCTION TO PUBLIC POLICY

times lie or shade the truth in order to make programs more effective, or
perhaps to preserve other important values, including the lives of citizens.

Truth-telling is a fundamental value in most ethical and religious systems

(Bok, 1978). Telling the truth is valued for its own sake, but it is also
valued for more consequentialist reasons. If one individual does not tell
the other the truth then that second individual will be making decisions
with false information and hence may make poor decisions. The first, less
than truthful, individual will have exercised power over the second. While
this lying may be acceptable in social situations – the familiar little white
lies that may facilitate ordinary life – it becomes much more problematic
when dealing with public policies.

Even in the public sector there may be instances in which lying may be
acceptable, or even necessary. As already noted, in foreign affairs it is gen-
erally accepted that diplomats or national leaders may have to lie on behalf
of their countries. The argument justifying this is that representatives of
one state have obligations of truthfulness to the members of their own
society but not necessarily to members of other societies, especially when
those societies are in adversarial or even competitive relationships. Thus
these officials accept the burden of “dirty hands” by lying, or perhaps by
doing even more serious wrongs than lying to others as a consequence of
accepting positions of responsibility.

In policymaking secrecy, and a lack of transparency, by not disclosing the

full truth may be a form of lying. As with other forms of lying on behalf of
one’s country, this behavior may be considered justifiable in international
affairs, but becomes less viable in domestic politics. The assumption, as
least among citizens in democratic regimes, is that their leaders have an
obligation to be truthful to the members of their political community.
Indeed, the obligation for truthfulness may be greater in government
because it represents a collective undertaking for the society and the
leaders have duties to the members of the society.

The question then becomes when, if ever, is lying and secrecy justified
when dealing with the citizens of one’s own country. Given that infor-
mation on national security issues given to the public would easily be
dispersed to potential adversaries then maintaining secrecy in those areas
appears necessary. Likewise, secrecy can be justified for information that
could be embarrassing to individual citizens or which might harm indi-
 NORMATIVE AND ETHICAL ANALYSIS OF POLICY 177

viduals economically (patent applications, for example). The assumption

is that all other information should be widely available to the public so
that they may too be involved in the policy process.

Making government fully transparent may have a strong moral claim, but
it may not enhance the quality of the policies being made (see Torfing et
al., 2012, chapter 7). Indeed, some countries have tended to shield their
policymaking from public view in order to ensure that the process will not
be unduly pressured by external forces.10 If policy advisors are to provide
the policymakers with “frank and fearless” advice, then they may require
some protections from external scrutiny. This is especially true for civil
servants who expect to be able to work anonymously and whose continu-
ing effectiveness in their positions may depend upon protections.

The commitment to telling the truth may appear almost quaint in a period
of “post-truth politics” (Fischer, 2019; Braun, 2020). A number of polit-
ical leaders in the populist era have created their own versions of truth
about important policy issues such as climate change and the COVID-19
pandemic. The internet has become a source of alternative facts about
those and many other policy issues, and what is telling the truth becomes
contestable.

In summary, the value of truth-telling and transparency in government

like so many of these other values affecting public policy are difficult to
make absolute. These values may be phrased in absolute terms – “Thou
Shall Not Bear False Witness” – but in reality there are compromises and
even direct denials of those absolutes. The answers to that question are
usually phrased in utilitarian terms – the benefits produced for society are
greater than if the truth were told. There may be other ways of justifying
the choice, such as the preservation of life – but most answers come down
to utilitarian reasons. We began this chapter looking for non-utilitarian
approaches to evaluating policies, but these appear difficult to maintain
once they confront tangible losses for individuals, and especially losses for
the society as a whole.
178 ADVANCED INTRODUCTION TO PUBLIC POLICY

Summary

Governments have to make difficult choices when they make and imple-
ment their policies. Even policies that may be effective, and even efficient,
in economic terms may entail other social and political effects that may
be undesirable. But even if there are attempts to make policy choices
congruent with ethical values, the decision-making problem is not over.
There are a number of important ethical values that can be contradictory,
or at least not consistent. Therefore, policy choices may involve balancing
a number of important values and attempting to make a choice that min-
imizes the adverse effects.

Like the discussion of cost-benefit analysis and allied methods these con-
siderations may be part of the formulation of the policy rather than an ex
post evaluation of the success or failure of a program. That said, however,
some of the effects of programs on individuals may not become apparent
until they are implemented. And given that these programs are being
constructed and assessed in a political environment, the choices among
values inherent in these decisions will include much more than ethical
or philosophical values. In the end political or even values may trump
ethical concerns as governments attempt to produce a package of ethical
and affordable policies. But the full range of utilitarian and non-utilitarian
values should be considered when designing policy.

Notes
1. These people are, rather naturally, those who benefit from the distributions
produced by the market.
2. James Q. Wilson’s (1980) analysis of politics focused on the concentration of
costs and benefits.
3. In most societies the military draft is a thing of the past but in some countries
such as Switzerland and Singapore there is universal military service. In prin-
ciple, in others a draft could be restored if military personnel requirements
increased significantly. See Levi (1997).
4. This concept is very similar to that of externalities discussed previously.
In both cases economic or social action produces “bads” that affect other
citizens, and the role of government is either to prevent those bads or com-
pensate those who are affected.
 NORMATIVE AND ETHICAL ANALYSIS OF POLICY 179

5. That reaction generally is to comply with the regulation, but perhaps then to
engage in political activity to modify the policy.
6. I have nothing against mountain climbers – this is only an example of risky
individual behaviors that have public policy implications.
7. Mitt Romney’s famous rant about the 47 percent of the population depend-
ent upon government is a now an infamous example of stigmatization of
citizens who receive public benefits.
8. In most cases the mother is designated as the recipient, assuming she will
bear primary responsibility for caring for children.
9. This is one of the few effective social programs that has been primarily an
American invention.
10. Much of the justification for the Official Secrets Act in the United Kingdom,
and analogs in other Westminster systems, was to preserve the internal
openness of the discussions. The assumption was having to make decisions
in public would produce policy choices that were more politically palatable
but not so good as policies.
10 Conclusion: policy success
and failure

Policymaking is the central activity of government. Governments do not

always make and implement policies on their own but rather often involve
partners coming from the private sector – both market and non-market
actors. No matter who is involved in the policy process, they face serious
challenges to reach their policy goals, and their political goals associated
with the policy. The design perspective I have adopted in this book
assumes that policymakers attempt to shape programs in ways that can
generate success, but that achieving success is a difficult and sometimes
impossible process.

But what is success? It has seemed that identifying policy failure has
been easier for both citizens and policy analysts. For example, there have
been studies of failures in planning (Hall, 1980) and implementation
(Pressman and Wildavsky, 1974), as well as more general discussions of
“policy fiascoes” (Bovens and ‘t Hart, 1996). Other works have attempted
to contrast success and failure in different national settings and different
policy areas (Bovens, ‘t Hart and Peters, 2000). Further, there have been
several discussions of the failures of specific policy instruments and styles
of intervention (Sieber, 1980; Birkland and Waterman, 2008).

The evaluation sections above developed several versions of policy success,

but these definitions were perhaps excessively restricted. In particular,
cost-benefit analysis equates success with policies that produce greater
economic benefits than they require in costs. But most cost-benefit
analysis is ex ante, and we may not know if even that minimalist goal is
achieved for some time. For most citizens the idea of success in public
policies is more related to effectiveness than to efficiency, and are gen-
erally more concerned that services are actually delivered rather the
relationships to costs.

180
 CONCLUSION: POLICY SUCCESS AND FAILURE 181

Process success
Allan McConnell (2010) has provided an extensive discussion of policy
success. He argues that this success has three dimensions. The first is the
process dimension of policy success, or failure. At the most basic level pol-
icymaking processes need to be able to generate policy decisions, and to
do so in a timely manner. If we consider the gridlock and the inability of
some governments to make decisions, this seemingly simple dimension of
success may not be so simple. While we might think of these deadlocked
processes as typical of less-developed political systems, the failure of the
US federal government to make a budget in most years of the past decade
demonstrates that this can be a more generic phenomenon.

As well as simply making decisions, the policymaking process must be

able to maintain legitimacy for the political system. Most political systems
have clearly defined legal and constitutional procedures for making
policy, so any attempt to circumvent those rules may bring the legitimacy
of the policy, and the system more generally, into question. This having
been said, however, the formal decision-making process may be slow,
and some difficult issues may not be easily resolved through processes
with numerous veto players (Tsebelis, 2000). Further, non-majoritarian
institutions (Majone, 2002a) may be able to make those difficult decisions
that more political institutions cannot.

Program success
Program success represents the second dimension of policy success in
McConnell’s analysis. This dimension is what most people might think
of policy success more generally, meaning actually reaching the intended
goals of the policy. A successful policy in this perspective delivers benefits
to the intended targets of the program but, as the evaluation chapters have
indicated, ascertaining the extent to which that success has been achieved
may not be easy. Although we will discuss political success below, the
definition of program success may also be politically defined.

First, there is the question of which goals, and whose goals. Few policies
are so simple or so unambiguous that the analyst can say that goals have
been reached. For example, is building a new highway about energy effi-
ciency, traffic safety, reducing time lost for commuters or providing jobs
for construction workers? The answer is usually that it is about all four,
182 ADVANCED INTRODUCTION TO PUBLIC POLICY

and in different economic periods the priority given to each of the goals
may change And if the program does well on most goals, but fails on one
or more then is this program a success or not? That answer may depend
primarily on the organization responsible, valuing its core policy goals
– reducing commuter time and traffic safety in the case of the highway
department – significantly higher than the ancillary goals.

In addition, the extent to which the program is seeking appropriate goals,

and appropriate levels of response to problems, can generate questions
about success. Moshe Maor has demonstrated the existence of “policy
bubbles”, meaning patterns of inadequate or excessive responses to policy
problems. The reaction of some politicians such as Trump and Bolsonaro
to the COVID-19 pandemic is a clear example of a negative policy bubble.
Thus, a program may succeed in its own terms but still represent a failure
in more objective terms.

Success and failure of programs may also be affected by unintended

consequences of programs. Is a program that reaches its principal goals
but also has a number of negative side-effects really a success? Once again
this may depend heavily upon who is asked to evaluate the program. And
for some analysts any program that is very expensive and is perceived
as wasteful may not be considered a success even if it hits the intended
targets. Different ideological perspectives also will produce different
perspectives on success and failure of programs. For example, for liberals
(in the American sense of the term) having more people receive social
benefits may be a success, while for conservatives that would be failure.

Political success
The political nature of some of the answers that can be given to the ques-
tions above lead rather naturally to the third dimension of success and
failure, which is political. Public policymaking is operating in an inher-
ently political environment, and the effects of policies on the political for-
tunes of the actors involved will, of course, be of great concern for those
participants in the process. Perhaps most obviously the impact of policy
choices on the electoral fortunes of the participants will be central in the
minds of the political class. If, as Anthony Downs (1957) argued some
years ago, politicians make policy to be elected rather than vice versa then
this is a crucial dimension of success.
 CONCLUSION: POLICY SUCCESS AND FAILURE 183

But the political dimension of policy success extends well beyond the
electoral success of specific politicians. It is becoming increasingly appar-
ent that the legitimacy of political systems depends upon the outputs of
the system as much or more than the process by which policies are made
(Gilley, 2009). This is especially true for the European Union in which
the electoral connections of citizens to much of the leadership is tenuous
at best (Scharpf, 2009). But many national states are legitimated by their
capacity to maintain economic growth and generous welfare state benefits
for the public.

Summary
Policymakers want to be able to say that they have produced effec-
tive and successful policies for their societies. But that claim is not as
simple as it might appear and policy success and failure is complex
and multi-dimensional. Further, success and failure may be perceptual
and constructed (see Stone, 2002), and those perceptions differ across
individuals. Given that politicians and citizens often have very different
perceptions of the political and policy world it is only natural that they
will evaluate policies differently.

Obamacare in the United States represents a clear example of this

multi-dimensionality and complexity in assessing success and failure.
While there is little doubt that the launch of the program was a fiasco,
by spring 2014 the program had reached and then exceeded its target for
enrolling citizens (Somashekar and Millman, 2014). But there did not
appear to be as many young enrollees as hoped, given that younger people
with lower health care costs were important for the financial viability of
the program. And many Republicans and even independents continued
to rail against the program. Is this a success?

Going beyond individual policies

The discussion of policymaking to this point has focused on the success

or failure of individual programs, but governments make and implement
hundreds, if not thousands, of programs. This crowded policy envi-
ronment involves a number of policies that conflict with one another,
184 ADVANCED INTRODUCTION TO PUBLIC POLICY

and others that could support one another but do not work together.
Therefore, one crucial element for improving the success of policies is to
improve their coordination, and to foster more integrated policies across
subsystems (Peters, 2014a). In some cases programs may actually con-
tradict one another, as when the US Department of Agriculture supports
tobacco farming, while the Department of Health and Human Services
discourages smoking.

It is very easy for analysts and practical policymakers to say that there
should be more coordination across policies, but producing that coordi-
nation is much more difficult. There are a number of organizational and
political barriers to producing coordination. At one level the protection
of “turf”, meaning the budgets, personnel and policy space allocated to an
organization in government (Beale, 1995; Blatter, 2003), and the control
of information to preserve organizational power hinder coordination.
Differences among political parties in government and vertical coordina-
tion issues in federalism contribute further to continuing separation and
conflict among policies. And finally professionalism, while perhaps con-
tributing to success of individual programs, may also make cooperation
with other programs more difficult.

Although governments tend to be divided into a large number of special-

ized programs, most of the important policy problems facing government
are not so specialized. For example, improving the health of the popula-
tion is a major policy issue but cannot be handled just using the programs
within a ministry of health. Improving health also involves food and
nutrition programs, sports and recreation, social assistance, education,
housing, and perhaps others policy areas. Therefore, getting all those
programs to work together cooperatively presents a major opportunity
for governments, but also a major challenge. This challenge typically is
addressed through the hierarchical power of executives or central agen-
cies, but network or even market devices may be effective in producing
needed cooperation among organizations and individuals.
 CONCLUSION: POLICY SUCCESS AND FAILURE 185

Reactions to policy failure: governing in the shadows

We have already argued that market failure is a compelling reason for

the intervention of the public sector into the economy and society. Even
anti-government conservatives may recognize that markets can be imper-
fect and some means of compensating for those failures must be found.
But even étatiste supporters of government interventions must realize
that governments can, and do, also fail to deliver the programs that they
have created as effectively as citizens, and most of the program’s person-
nel, would like.

The preceding section of this chapter has demonstrated the difficulty of

defining and assessing policy success and failure. When there is a real
or perceived failure then there is an impetus for change. Government
organizations can learn from their past successes and failures and modify
programs during the next round of decision-making, adapting to changes
in the environment and the effects of their own interventions (Peters,
2014c). But public programs may in some cases be incapable of actually
delivering effective policy, either because of their own weakness or the
nature of the policy problem being addressed. What then are the options
for policymaking?

Fritz Scharpf famously argued (1997) that when governments delegate to

social and market actors that delegation is occurring within a “shadow
of hierarchy”. This means that governments retain the legal authority
to withdraw the delegation whenever the agents fail to act properly.
Policymaking is also occurring in a shadow, or indeed a set of shadows, so
that when government fails to deliver the policies there are other alterna-
tives. In the contemporary political environment the obvious alternative
to the public sector has been the market. Whether through privatization,
the use of market-based instruments such as effluent charges or vouchers,
or contracting to private sector actors markets are a major alternative to
public sector provision. And market actors may provide basic govern-
ment services in some settings where governments are incapable (Hönke,
2010).

The market is not the only alternative source of policy and service pro-
vision for citizens. Social actors such as not-for-profits and faith-based
organizations are also alternative sources for policy and policy delivery.
186 ADVANCED INTRODUCTION TO PUBLIC POLICY

These actors are especially important in areas such as social policy and
education where these organizations often are major or even dominant
actors. Further, with the financial strains in the welfare state as a function
of demography and tax resistance (Ferrara and Rhodes, 2013) these social
actors are likely to become even more significant as the providers of
public services.

Finally, expertise can function as an alternative source of policy. In

a number of policy areas such as economic policy, or areas of energy
policy, governments may want to delegate to expert organizations such as
central banks, regulatory organizations or scientists. These organizations
also have the capacity to make “credible commitments” to individuals and
firms so that those actors can make long-term decisions (North, 1993).
While expertise and credible commitment can be effective in making and
delivering public policies, there are also problems for democracy when
policy is so dependent upon expertise with little democratic control.

These four alternative sources for policymaking and governance exist in

any society, and to some extent exist for any policy area. Simply saying
that, however, is only the beginning of understanding the manner in
which choices of styles of intervention are made. One may argue that this
represents an evolutionary approach to policymaking, but it also must
involve agency (Capano, 2009). How do actors involved make decisions
about how to move away from rather conventional formats for public
policy toward using other actors and institutions in order to provide more
effective policymaking? And when will they want to move back to those
conventional formats for policymaking?

Not only is there a question of agency and decision-making in the

movement among broad approaches to policy expressed here as a set of
“shadows”, but there is also a question of mixtures. Few policies in con-
temporary governments can be seen as being delivered purely by one or
another institution. Rather, most policy delivery represents hybrid forms
of action and interactions among numerous actors (Torfing et al., 2012;
see also Chapters 5 and 6). Therefore, policymakers will need to consider
the nature of these hybrids and the way in which the multiple actors can
combine to provide the most effective public services.

The study of policymaking “in the shadows” also links with the study of
informal means of governance (Helmke and Levitsky, 2004). Most of our
 CONCLUSION: POLICY SUCCESS AND FAILURE 187

discussions of public policy focus on formal rule-making and the formal

actors who make those rules, but a good deal of governing is informal and
depends on social institutions such as clientelism and clans as much as on
government as we usually conceptualize it. Therefore, as we continue to
develop this field of study we must understand better the informal as well
as the formal.

Continuing challenges to studying public policy

Although the study of public policy has made numerous advances since
the initial formulation of the policy sciences by Lasswell and his col-
leagues, there are a number of continuing issues about how best to con-
tinue that advance. To some extent the problem in studying public policy
is that there is an embarrassment of riches. As several of these chapters
have pointed out, there are multiple models of policymaking, and multi-
ple approaches to particular aspects of the policy process such as imple-
mentation. While this is a rich analytical palette from which to choose
when painting a picture about policy, the problem is choosing under what
conditions to employ one approach or another. This question is largely
empirical, because there are no meta-theories that permit making choices
among the available options (see Alexander, 1984).

Although there are many theoretical options for understanding public

policy, there are still some significant issues that are not addressed ade-
quately. One may be explaining innovation in policy. There is a large
literature on policy change that certainly touches on issues of innovation,
but much of that literature has rather incremental assumptions. While
that may reflect reality, as indeed much of policy change in governments
is incremental, it does not provide an effective understanding of the crea-
tion of genuinely new and innovative approaches. Nor do we understand
the structures and procedures that could foster greater innovation.

Public policy in comparative context

The necessity of making choices about policymaking strategies outlined
above also points to the need to consider public policy in a comparative
context. The discussion in this book largely has been generic, assuming
that policymaking is virtually the same everywhere. While that generality
is certainly true at one level, there are also important institutional and
cultural features of individual policymaking systems that demand more
188 ADVANCED INTRODUCTION TO PUBLIC POLICY

nuanced understandings. These general understandings of policy may be

a good place to begin our discussions but they are not always a good place
to end them.

The need to think comparatively is at least as true for different policy

areas as it is for different political systems. Gary Freeman has argued
(1985) that the differences across policy areas are more significant than
are those across countries, and that health policy (as one example) may
be more similar in several countries than are education and health within
the same country. Factors such as the nature of the interests involved, the
level of technical and professional content, the nature of the instruments
available, and a host of other factors may all differentiate different policy
areas and differentiate their politics, as well as more technical aspects of
their delivery.

If we consider some of the general models of policymaking above in light

of different policy areas some of these differences become apparent. For
example, if we consider health policy through the lens of the advocacy
coalition framework the shift in emphasis in health policy from more
technical, scientific forms to the “science of delivery” (Mulley et al., 2013)
can be seen as creating a new vision for health care. Whereas this discus-
sion appears largely to be conducted on the basis of science, proposed
shifts in policy areas such as education or social policy have tended to
be more politicized. Similarly, the scientific nature of health policy and
energy policy may be less amenable to the multiple streams approach than
other policy areas with less clearly defined foundations.

Leaving aside comparisons among policy areas, there are also important
comparisons across political systems that influence policy and policy-
making. For example, the complexity of policymaking systems, defined
perhaps as the number of veto players in the system, will influence the
capacity to make effective policy decisions. Weaver and Rockman (1993)
assessed the capacities of parliamentary versus more complex presidential
systems to make policy choices, finding no clearly defined differences.
Scharpf (1988), on the other hand, argued that systems with multiple veto
players, such as the European Union, find it difficult to make other than
minor adjustments from the status quo. The multiple checks and balances
in American government also inhibit effective policymaking, especially
when complicated by sharp ideological differences between the political
parties.
 CONCLUSION: POLICY SUCCESS AND FAILURE 189

The above being said, complex political systems may also be able to foster
innovation, or at least provide openings for innovation. Federal systems,
for example, allow innovation in sub-national units that might be dif-
ficult or impossible at the central government level. And systems with
multiple active policymaking institutions, including perhaps an active
court system, will provide more loci for action than would more linear
parliamentary systems. Thus, while polycentric political systems may
find it difficult to be decisive, they may have broader policy agendas that
ultimately will produce more innovative programs.

As well as structural differences across regimes, there are also important

cultural differences. One classic example of these differences may be in
Esping-Andersen’s (1990) discussion of types of welfare states. Despite
the numerous critiques of this work (see Hemerijck, 2012), it does
demonstrate the manner in which social and cultural factors influence
policy choices. These cultural influences are often quite subtle, and may
lead to stereotyping, but they are real. And they may be more difficult
to change than policy regimes based more on political or ideological
foundations.

As implied above, the difficulty in comparative analysis of public policy

may be in making causal links between presumed independent variables
and policy choices. And further, making the link between those independ-
ent variables and the success and failure of programs. Policymaking is
complex and involves numerous decisions. As Pressman and Wildavsky
(1974) pointed out, even the implementation stage of the process may
involve a hundred or more independent decisions. Therefore, attempt-
ing to track the process and to determine when and how certain factors
influenced final choices is at best difficult. This is especially true when
attempting to understand these choices in a comparative context.

We are therefore left with what may be another paradox for studying
public policy. On the one hand, understanding the dynamics of policy,
and perhaps even its content, is best done comparatively; without the
comparative framing of policy outcomes it is difficult to understand
their broader relevance. On the other hand, however, that comparative
frame introduces an even greater array of factors that can influence the
outcomes, so that explanations may become more convoluted and more
difficult to understand.
190 ADVANCED INTRODUCTION TO PUBLIC POLICY

As much as I favor comparative analysis for developing more nuanced

understandings of policy, bringing in the comparative dimension may
pose extremely difficult challenges, and those challenges may not be worth
surmounting for decision-makers faced with an immediate problem
within their own context. Comparative analysis may introduce some
factors for explanation that, while useful for the explanation, may not be
manipulable by those decision-makers. While we are interested in public
policy for scientific reasons, we are also interested in comparative analysis
for more practical reasons – we want to be able to make and implement
better policies. Given that practical concern, having explanatory variables
that can be manipulated readily enables the policy analyst to provide
useful advice to decision-makers.

Linking with institutions

Much of policy analysis appears to be almost institution free. Although
mentioning numerous actors, mostly the bureaucracy, the models of the
policy process tend to assume that those institutions are rather uniform
across countries, and therefore the models developed in one setting will
be applicable to all. This problem tends to be rife within models of policy
developed by American and British scholars but also appears for other
nationalities as well. While the discussion of comparison above focused
primarily on differences in political and policy cultures, institutional
analysis also needs to be strengthened.

Some years ago Fritz Scharpf (1986) argued that policy design performed
without references to the institutions within which that design is occur-
ring is not likely to be successful. This argument and other institutionalist
arguments are that design involves building institutional structures along
with the policy (see Araral, 2014). That said, simple structural changes
cannot be sufficient to guarantee policy successes, but rather these two
aspects of policy need to be linked and constructed together.

The linkage between institutions and policy can be identified within

a number of areas of the policy process. For example, the agenda-setting
process is influenced by the number of available venues for getting issues
in to the process. And implementation is shaped not only by the number
of clearance points that may inhibit a policy being put into effect as
planned, but also the implementation structures that integrate public
and private sector actors who affect the outcomes of the implementation
 CONCLUSION: POLICY SUCCESS AND FAILURE 191

process. We could extend the list of institutional linkages involved in

making public policy function but the fundamental point is that policy
can be considered in a structural or institutional vacuum.

Where do we go from here?

This book has provided a broad perspective on public policy. Although

it has been written primarily from the perspective of political science
and the political dimension of policy, I have attempted to include some
perspectives from other disciplines. The book provides an overview of
contemporary public policy analysis but there is still a great deal that
needs to be done for policy analysis to fulfill the promise of improving the
performance of governments and their allies in the private sector.

The design framework that I have been using for this book is meant to
provide a means of integrating a number of dimensions of policy analysis.
The assumption is that by considering the cause of the perceived problem,
the means of addressing that problem, and the means of evaluating
whether the interventions by the public sector have been successful it may
be more possible to provide a clear understanding of how policies get
made, and then remade. Further, by thinking of policymaking as a design
process there is always a conception of what the final product should, or
at least could, look like.

As committed as I may be to the general logic of policy design, this does

not provide the detailed answers to many important questions. Those
questions are in part academic but they are also practical. Scholars are
interested in the theoretical and analytic questions associated with policy
design, but a technology for effective policy design does not exist. Men
and women involved in making policies for real governments with real
people are less concerned with theory and more with whether a program
will actually work. I have attempted to provide some insights for both of
these communities, but the real answers for design come from blending
the two strands of analysis and concern.
References

Adler, M. (2019) Measuring Social Welfare: An Introduction (Oxford: Oxford

University Press).
Alesina, A. and E.L. Glaser (2004) Fighting Poverty in the U.S. and Europe:
A World of Difference (Oxford: Oxford University Press).
Alexander, E. (1984) “After rationality, what?: a review of responses to paradigm
breakdown”, Journal of the American Planning Association 50, 62–9.
Alexander, E.R. (1989) “Improbable implementation: the Pressman-Wildavsky
paradox revisited”, Journal of Public Policy 9, 451–65.
Alford, J. (2009) Engaging Public Sector Clients: From Service-Delivery to
Co-Production (Basingstoke: Palgrave Macmillan).
Aligica, P.D. and P.J. Boetkke (2009) Challenging Institutional Analysis and
Development: The Bloomington School (New York: Routledge).
Allison, G.T. and P. Zelikow (1999) The Essence of Decision, rev. edn (New York:
Longmans).
Alm, J. and B. Togler (2011) “Do ethics matter?: tax compliance and morality”,
Journal of Business Ethics 101, 635–61.
Alsott, A.L. (1995) “The earned income tax credit and the limitations of tax-based
welfare policies”, National Tax Journal 108, 533–92.
Altmann, D.M., D C. Douek and R.J. Boynton (2020) “What policymakers need to
know about COVID-19 protective immunity”, The Lancet 395, 1527–9.
Ansell, C. and A. Gash (2007), “Collaborative governance in theory and practice”,
Journal of Public Administration and Theory 18, 543–71.
Araral, E. (2014) “Policy and regulatory design for developing countries: a mecha-
nism design and transaction cost approach”, Policy Sciences 47, 289–303.
Army Corps of Engineers (1996) Risk-Based Analysis for Corps Flood Project
Studies (Davis, CA: Hydrological Engineering Center).
Arrow, K.J. (1996) Benefit-Cost Analysis in Environmental Health and Safety
Regulation (Washington, DC: Office of Management and Budget).
Auyero, J. (2012) Patients of the State: The Politics of Waiting in Argentina
(Durham, NC: Duke University Press).
Axelrod, R. (2007) The Evolution of Cooperation, rev. edn (New York: Basic
Books).
Bache, I. (2013) “Measuring quality of life for public policy”, European Journal of
Public Policy, 20, 21–38.
Bachrach, P. and M.S. Baratz (1962) “Two faces of power”. American Political
Science Review 56 (4), 947–52.

192
 REFERENCES 193

Bagenstos, S.R. (2000) “Subordination, stigma and ‘disability’”, Virginia Law

Review 86, 398–486.
Baggott, R. (2012) “Policy success and public health: the case of public health in
England”, Journal of Social Policy 41, 391–408.
Bardach, E. (1977) The Implementation Game: What Happens After a Bill Becomes
a Law (Cambridge, MA: MIT Press).
Bardach, E. (1998) Getting Agencies to Work Together: The Practice and Theory
of Managerial Craftsmanship (Washington, DC: The Brookings Institution).
Bardach, E. and R. Kagan (1982) Going by the Book: The Problem of Regulatory
Unreasonableness (Philadelphia, PA: Temple University Press).
Bardach, E. and R. Kagan (2009) Adversarial Legalism: The American Way of Law
(Cambridge, MA: Harvard University Press).
Barrett, D. (2014) “Watchdog strips official status from police crime statistics”,
The Telegraph, 15 January.
Bates, R.H., A. Greif, M. Levi, J.-L. Rosenthal and B.R. Weingast (1988) Analytic
Narratives (Princeton, NJ: Princeton University Press).
Baumgartner, F.R. and B.D. Jones (2010) Agendas and Instability in American
Politics, 2nd edn (Chicago: University of Chicago Press).
Baumgartner, F.R., D.A. Epp and K. Shoub (2018) Suspect Citizens: What 20
Million Traffic Stops Tell Us About Policing and Race (Cambridge: Cambridge
University Press).
Beach, D. and R.B. Pedersen (2013) Process-Tracing Methods: Foundations and
Guidelines (Ann Arbor: University of Michigan Press).
Beale, R. (1995). “Turf protection: conflict between authorities”, Australian
Journal of Public Administration 54, 143–7.
Béland, D. (2010) “Reconsidering policy feedback: how policies affect politics”,
Administration and Society 42, 568–90.
Béland, D. and R.H. Cox (2011) Ideas and Politics in Social Science Research (New
York: Oxford University Press).
Béland, D. and M. Howlett (2016) “How solutions chase problems: instrument
constituencies in the policy process”, Governance 39, 393–409.
Béland, D., P. Rocco and A. Waddan (2016) “Assessing policy drift: social policy
change in the United States”, Social Policy & Administration 50, 201–18.
Bellé, N., P. Cantarelli and P. Belardinelli (2018) “Prospect theory goes public:
experimental evidence on cognitive biases in public policy and management
decisions”, Public Administration Review 78, 828–40.
Bemelmans-Videc, M.-L., R.C. Rist and E. Vedung (1998) Carrots, Sticks and
Sermons: Policy Instruments and their Evaluation (New Brunswick, NJ:
Transaction Publishers).
Bennett, J.T. (2014) “Congressional conundrum: cut weapons programs now or
later”, Defense News, 11 May.
Benz, A. and J. Broschek (2012) Federal Dynamics: Continuity, Change and
Varieties of Federalism (Oxford: Oxford University Press).
Bernstein, J. and S. Parrott (2014) Proposal to Strengthen Minimum Wage Would
Help Low-Wage Workers, with Little Effect on Employment (Washington, DC:
Center for Budget and Policy Priorities).
Best, R. (2010) “Situation or social problem: the influence of events on media
coverage of the homeless”, Social Problems 57, 74–91.
194 ADVANCED INTRODUCTION TO PUBLIC POLICY

Bingham, L.B., T. Nabatchi and R. O’Leary (2005) “The new governance: practices
and processes for stakeholder and citizen participation in the work of govern-
ment”, Public Administration Review 65 (5), 547–58.
Birkland, T.A. (1998) “Focusing events, mobilization and agenda setting”, Journal
of Public Policy 18, 53–74.
Birkland, T.A. (2006) Lessons from Disaster (Washington, DC; Georgetown
University Press).
Birkland, T.A. and S.E. de Young (2013) “Focusing events and policy windows”,
in E. Araral, S. Fritzen, M. Howlett. M. Ramesh and X. Wu (eds), Routledge
Handbook of Public Policy (London: Routledge).
Birkland, T.A. and S. Waterman (2008) “Is federalism the reason for policy failure
in Hurricane Katrina”, Publius 38, 692–714.
Blatter, J. (2003) “Beyond hierarchies and networks: institutional logics and
change in transboundary space”, Governance 16, 503–26.
Blaug, M. (2011) “Welfare economics”, in R. Towse (ed.), Handbook of Cultural
Economics (Cheltenham, UK and Northampton, MA, USA: Edward Elgar).
Blyth, M. (2007) “Puzzling, powering or persuading: the mechanisms for building
institutional orders”, International Studies Quarterly 51, 761–77.
Bobrow, D. and J.S. Dryzek (1987) Policy Analysis by Design (Pittsburgh:
University of Pittsburgh Press).
Boin, A. (2004) “Lessons from Crisis Research”, International Studies Quarterly
6, 165–94.
Boin, A., P. ‘t Hart and B. Sundelius (2005) The Politics of Crisis Management:
Public Leadership Under Pressure (Cambridge: Cambridge University Press).
Bok, S. (1978) Lying: Moral Choice in Public and Private Life (New York:
Pantheon).
Boscarino, J. (2019) “From Three Mile Island to Fukushima: the impact of analogy
on attitudes toward nuclear power”, Policy Sciences 52, 21–42.
Boston, J., A. Braddock and D.L. Eng (2010) Public Policy: Why Ethics Matter
(Canberra: ANU Press).
Botterill, L. (2013) “The promise and challenge of evidence-based policymaking”,
in L. Botterill (ed.) Drought, Risk Management and Policy Decision-Making
Under Uncertainty (Boce Raton: CRC Press).
Botterill, L.C. and A. Fenna (2019) Interrogating Public Policy Theory: A Political
Values Perspective (Cheltenham, UK and Northampton, MA, USA: Edward
Elgar).
Bouckaert, G. and J.A. Halligan (2007) Managing Performance: International
Comparisons (London: Routledge).
Bouckaert, G. and B.G. Peters (2002) “Performance measurement and manage-
ment: the achilles’ heel of administrative modernization”, Public Performance
& Management Review 25, 359–62.
Bovens, M.A.P. and P. ‘t Hart (1996) Understanding Policy Fiascoes (New
Brunswick, NJ: Transaction Publishers).
Bovens, M.A.P., P. ‘t Hart and B.G. Peters (2000) Success and Failure in Public
Governance (Cheltenham, UK and Northampton, MA, USA: Edward Elgar).
Bowen, E.B. (1982) “The Pressman-Wildavsky paradox: four addenda, or why
models based on probability theory can predict implementation success and
 REFERENCES 195

suggest useful tactical advice for implementers”, Journal of Public Policy 2,

1–21.
Braun, D. and F. Gilardi (2006) “Taking ‘Galton’s Problem’ seriously: towards
a theory of policy diffusion”, Journal of Theoretical Politics 18, 298–322.
Braun, K. (2020) “Unpacking post-truth”, Critical Policy Studies 13, 432–6.
Breeden, A. (2020) “I’m suffocating: details of chokehold death in Paris renew
scrutiny of police”, New York Times, 23 June.
Brehm, J. and S. Gates (2002) Working, Shirking, and Sabotage: Bureaucratic
Response to a Democratic Public, 4th edn (Ann Arbor: University of Michigan
Press).
Brent, R.J. (2006) Applied Cost-Benefit Analysis, 2nd edn (Cheltenham, UK and
Northampton, MA, USA: Edward Elgar).
Bressers, H. and P.J. Klok (1988) “Fundamentals for a theory of policy instru-
ments”, International Journal of Social Economics 15, 22–41.
Brodkin, E.Z. (2011) “Policy work: street-level bureaucracy under the new man-
agerialism”, Journal of Public Administration Research and Theory 21, 253–77.
Brook, Y and D. Watkins (2010) Free Market Revolution (New York: Palgrave
Macmillan).
Brown, C. (2007) “Tragedy, tragic choices and contemporary international politi-
cal theory”, International Relations 21, 5–13.
Brown, R. (2004) “Considerations of the trigins of Herbert Simon’s Theory of
‘Satisficing’ (1933–1947)”, Management Decision 42, 1240–56.
Browne, A. and A. Wildavsky (1984) “Implementation as adaptation”, in J.L.
Pressman and A. Wildavsky (eds), Implementation, 3rd edn (Berkeley:
University of California Press).
Calabrese, G. and P. Bobbitt (1978) Tragic Choices (New York: W.W. Norton).
Calista, D. (1994) “Policy implementation”, in S. Nagel (ed.), Encyclopedia of
Policy Studies, 2nd edn (New York: Marcel Dekker).
Capano, G. (2009) “Understanding policy change as an epistemological and theo-
retical challenge”, Journal of Comparative Policy Analysis 11, 7–31.
Carnes, N. (2013) White-Collar Government: The Hidden Role of Class in Economic
Policy-Making (Chicago: University of Chicago Press).
Carroll, A.E. (2014) “How hobby lobby rule could limit access to birth control,
New York Times, 30 June.
Carter, P. (2012) “Policy as palimpset”, Policy & Politics 40, 423–43.
Challis, L. (1988) Joint Approaches to Social Policy: Rationality and Practice
(Cambridge: Cambridge University Press).
Chappell, L. (2002) “Federalism and social policy: the case of domestic violence”,
Australian Journal of Public Administration 60, 59–69.
Chong, D. and J.N. Druckman (2007) “Framing theory”, Annual Review of
Political Science 10, 103–26.
Christensen, T. and P. Laegreid (2007) Transcending New Public Management:
The Transformation of Public Sector Reforms (Aldershot: Ashgate).
Clemmitt, M. (2014) “Can Obama’s health insurance expansion recover from its
shaky rollout?” CQ Researcher, 17 April.
Coady, C.A. (1991) Politics and the Problem of Dirty Hands: A Companion to
Ethics (Oxford: Blackwell).
196 ADVANCED INTRODUCTION TO PUBLIC POLICY

Cobb, R.W. and C.D. Elder (1972) Participation in American Politics: The
Dynamics of Agenda-Building (Baltimore, MD: Johns Hopkins University
Press).
Cohen, M.D., J.G. March and J.P. Olsen (1971) “A garbage can model of organiza-
tional choice”, Administrative Science Quarterly 17, 1–25.
Commonwealth Fund (2014) Mirror, Mirror on the Wall 2014 Update; How the US
Health Care System Compares Internationally (New York: The Commonwealth
Fund).
Compton, M. and P. ‘t Hart (2019) Great Policy Successes (Oxford: Oxford
University Press).
Considine, M. (2002) “The end of the line?: accountable governance in an age of
networks, partnerships, and joined-up government”, Governance 15 (1), 21–40.
Culpepper, P.D. (2002) “Powering, puzzling and ‘pacting’: the informational logic
of negotiated reforms”, European Journal of Public Policy 9, 774–90.
D’Adda, G., V. Caprarp and M. Tavoni (2017) “Push, don’t nudge: behavioral
spillovers and policy instruments”, Economic Letters 154, 92–5.
Dahl, P.S. and K.M. Hansen (2006) “Diffusion of standards: the importance of size,
region and external pressures in diffusion processes”, Public Administration 84,
441–59.
Dahl, R.A. and C.E. Lindblom (1953) Politics, Economics and Welfare: Planning
and Politico-Economic Systems Resolved into Basic Processes (New York:
Harper).
Dasgupta, P.S. and P.R. Ehrlich (2013) “Pervasive externalities at the population,
consumption and environment nexus”, Science 340, 324–38.
Davis, O.A., M.A.H. Dempster and A. Wildavsky (1966) “A theory of the budget-
ary process”, American Political Science Review 60 (3), 529–47.
Dempster, M.A.H., and A. Wildavsky (1979) “On change: or, there is no magic
size for an increment”, Political Studies 27, 371–89.
Denhardt, R.B. and J.V. Denhardt (2007) The New Public Service: Serving Not
Steering (Armonk, NY: M.E. Sharpe).
Dery, D. (2000) “Agenda-setting and problem definition”, Policy Studies 21 (1),
27–47.
Donahue, J.D. and R.G. Zeckhauser (2011), Collaborative Governance: Private
Roles for Public Goals in Turbulent Times (Princeton, NJ: Princeton University
Press).
Doonan, M. (2013) American Federalism in Practice: The Formulation and
Implementation of Contemporary Health Policy (Washington, DC: The
Brookings Institution).
Downs, A. (1957) An Economic Theory of Democracy (New York: Harper).
Downs, A. (1972) “Up and down with ecology – the ‘issue-attention cycle’”, The
Public Interest, 38–50.
Dror, Y. (1986) Policymaking Under Adversity (New Brunswick, NJ: Transaction
Publishers).
Dryzek, J.S. and C. List (2003) “Social choice and deliberative democracy: a recon-
ciliation”, British Journal of Political Science 33, 1–28.
Dubois, V. (2010) The Bureaucrat and the Poor: Encounters in a French Welfare
Office (Aldershot: Ashgate).
 REFERENCES 197

Duit, A. and V. Galaz (2008), “Governance and complexity – new issues in gov-
ernance theory”, Governance 21, 311–35.
Dunlop, C. (2013) “Epistemic communities”, in E. Araral, S. Fritzen, M. Howlett,
M. Ramesh and X. Wu (eds), Routledge Handbook of Public Policy (London:
Routledge).
Dunlop, C.A. and C.M. Radaelli (2018) “the lessons of policy learning: types,
triggers, hindrances and pathologies”, Policy & Politics 46, 255–72.
Dunn, W.N (2012) Public Policy Analysis: An Introduction (Boston: Pearson).
Eckstein, H. (1980) The Natural History of Congruence Theory (Denver: Graduate
School of International Studies, University of Denver).
Elmore, R.F. (1985) “Forward and backward mapping: reversible logic in the analy-
sis of public policy”, in K. Hanf and T.A.J. Toonen (eds), Policy Implementation
in Federal and Unitary States (Dordrecht: Martinus Nijhoff).
Elster, J. (1991) “The puzzling scope of rationality”, European Journal of Sociology
32, 109–29.
Erler, H.A. (2013) “A new face of poverty: economic crises and poverty dis-
courses”, Poverty and Public Policy 4, 183–204.
Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism (Princeton,
NJ: Princeton University Press).
Eterno, J.A. and E.B. Silverman (2012) The Crime Numbers Game: Management
by Manipulation (Boca Raton, FL: CRC Press).
Exworthy, M. and M. Powell (2004) “Big windows and little windows: implemen-
tation in the ‘congested state’”, Public Administration 82, 263–81.
Fancy, C. (2011) “The politics of social policy in America: the causes and effects of
indirect versus direct social spending”, Journal of Politics 73, 73–83.
Farrington, D.P. (2003) “British randomized experiments on crime and justice”,
The Annals of the American Academy of Political and Social Sciences 589,
150–67.
Farrow, S. and R.O. Zerbe (2013) Principles and Standards for Cost-Benefit
Analysis (Cheltenham, UK and Northampton, MA, USA: Edward Elgar).
Fenna, A. and T. Hueglin (2010) Comparative Federalism: A Systematic Inquiry
(New York: University of Toronto Press).
Fenner, F., D.A. Henderson, I. Arita, Z. Jezek and I.D. Ladnyi (1999) Smallpox and
Its Eradication (Geneva: World Health Organization).
Fernandez, S. and T. Moldogaziev (2010) “Empowering public sector employ-
ees: to improve performance; does it work?”, American Review of Public
Administration 4 (1), 23–47.
Ferrara, M. and M. Rhodes (2013) Recasting European Welfare States (London:
Routledge).
Fischer, F. (2019) “Knowledge politics and post-truth in climate denial: on the
social construction of alternate facts”, Critical Political Studies 13, 133–52.
Fischer, F. and H. Gottweiss (2012) The Argumentative Turn Revisited: Public
Policy as Communicative Practice (Durham, NC: Duke University Press).
Forget, E.L. (2011) “The town with no poverty”, Canadian Public Policy 37,
283–305.
Frakt, A. (2020) “What is the value of human life? Governments already tally it”,
New York Times, 12 May.
198 ADVANCED INTRODUCTION TO PUBLIC POLICY

Frederick, S., G. Loewenstein and T. O’Donoghue (2002) “Time discounting and

time preference: a critical review”, Journal of Economic Literature 40, 351–401.
Freeman, G.P. (1985) “National styles and policy sectors: explaining structured
variation”, Journal of Public Policy 5 (4), 467–96.
Freeman, J.L. and J.P. Stevens (1987) “A theoretical and conceptual reexamination
of subsystem politics”, Public Policy and Administration 2, 9–24.
Friedman, L. (2020) “Trump weakens major conservation law to speed construc-
tion permits”, New York Times, 15 July.
Fung, A., M. Graham and D. Weil (2007) Full Disclosure: The Perils and Promise
of Transparency (Cambridge: Cambridge University Press).
Ganghof, S. (2003) “Promises and pitfalls of veto player analysis”, Swiss Political
Science Review 9, 1–25.
Genieys, W. (2010) The New Custodians of the State: Programmatic Elites in
French Society (New Bruswick, NJ: Transaction Publishers).
Genieys, W. (2017) The New Custodians of the State: Programmatic Elites in
French Society (London: Routledge).
Geva-May, I. (2004) “Riding the wave of opportunity: termination in public
policy”, Journal of Public Policy Research and Theory 14, 309–33.
Gilens, M. and B.I. Page (2014) Testing theories of American politics: elites, inter-
est groups and average citizens”, Perspectives on Politics 12, 564–81.
Gilley, B. (2009) The Right to Rule: How States Win and Lose Legitimacy (New
York: Columbia University Press).
Gingrich, J.R. (2011) Making Markets in the Welfare State: The Politics of Varying
Market Reforms (Cambridge: Cambridge University Press).
Glover, J. (1977) Causing Death and Saving Lives (London: Penguin).
Goodsell, C.T. (2011) Mission Mystique: Belief Systems in Public Agencies
(Washington, DC: CQ Press).
Goodwin, P. and R.B. Noland (2003) “Building new roads really does create extra
traffic”, Applied Economics 35, 1451–7.
Gormley, W.T. (1989) Taming the Bureaucracy: Muscles, Prayers and Other
Strategies (Princeton, NJ: Princeton University Press).
Gottweiss, H. (2007) “Rhetoric in policymaking: between logos, ethos and
pathos”, in F. Fischer and G. Miller (eds), Handbook of Policy Analysis (Boca
Raton: CRC Press).
Guedes-Neto, J.V. and B.G. Peters (2021) “Working, shirking and sabotage in
times of democratic backsliding: an experimental study”, in Brazil, in M.
Becker, B.G. Peters, J. Pierre, K. Yesilkagit and S. Becker (eds), Democratic
Backsliding, Populism and Public Administration (Cambridge: Cambridge
University Press).
Hadenius, A. and J. Toerell (2007) “Pathways from authoritarianism”, Journal of
Democracy 18, 143–57.
Hahn, R.W. and C.R. Sunstein (2002) “A new Executive Order for improving
federal regulation? Deeper and wider cost-benefit analysis”, University of
Pennsylvania Law Review 50, 1489–552.
Hall, P.G. (1980) Great Planning Disasters (Berkeley: University of California
Press).
Halliburton, R. (2014) Autonomy and the Situated Self: A Challenge to Bioethics
(Lanham, MD: Lexington Books).
 REFERENCES 199

Hammerschmid, G. and R. Meyer (2004) “Public management dynamics in

a federal legalistic Rechtstaat: results from an executive survey in Austria”,
International Journal of Public Sector Management 18, 629–40.
Hammersley, M. (2013) The Myth of Research-Based Policy and Practice (London:
Sage).
Handler, J.F. and Y. Hasenfeld (2006) Blame Welfare: Ignore Poverty and Inequality
(Cambridge: Cambridge University Press).
Hanf, K. and T.A.J. Toonen (1985) Policy Implementation in Federal and Unitary
States (Dordrecht: Martinus Nijhoff).
Hardin, G.J. (1968) “Tragedy of the commons”, Science 162 (3859), 1243–8.
Hausman, D.M. and B. Welch (2010) “Debate; to nudge or not to nudge”, Journal
of Political Philosophy 18, 123–36.
Hayes, M. (2017) “Incrementalism and public policy”, in Oxford Research
Encyclopedia of Politics (New York: Oxford University Press).
Head, B. (2021) “Coping with wicked problems in policy design”, in B.G. Peters
and G. Fontaine (eds), Handbook of Policy Design (Cheltenham, UK and
Northampton, MA, USA: Edward Elgar).
Heclo, H. (1974) Modern Social Politics in Britain and Sweden (New Haven, CT:
Yale University Press).
Helmke, G. and S. Levitsky (2004), “Informal institutions and comparative poli-
tics: a research agenda”, Perspectives on Politics 2, 725–40.
Helms, L. (2012) Poor Leadership and Bad Governance (Cheltenham, UK and
Northampton, MA, USA: Edward Elgar).
Hemenway, D. and M. Miller (2013) “Public health approach to the prevention of
gun violence”, New England Journal of Medicine, 12 April.
Hemerijck, A. (2012) Changing Welfare States (Oxford: Oxford University Press).
Henry, A.D. (2011a) “Information, networks and complexity of trust in commons
governance”, International Journal of the Commons 5, 188–212.
Henry, A.D. (2011b) “Ideology, power and the structure of policy networks”,
Policy Studies Journal 39, 361–83.
Hepburn, C. (2007) “Valuing the far-off future: discounting and its alternatives”,
in G. Atkinson, S. Dietz and E. Neumayer (eds), Handbook of Sustainable
Development (Cheltenham, UK and Northampton, MA, USA: Edward Elgar).
Héritier, A. and M. Rhodes (eds) (2012), New Modes of Governance in Europe:
Governing in the Shadow of Hierarchy (Basingstoke: Palgrave Macmillan).
Herweg, N., N. Zahariadis and R. Zohlnhöfer (2018) “The multiple streams frame-
work: foundations, refinements and empirical application”, in C.M. Weible
and P.A. Sabatier (eds), Theories of the Policy Process, 4th edn (Boulder, CO:
Westview Press).
Hill, M. and P. Hupe (2014) Implementing Public Policy, 3rd edn (London: Sage).
Hisschemöller, M. and R. Hoppe (1995) “Coping with intractable controversies:
the case of problem structuring in policy design and analysis”, Knowledge and
Policy 8, 40–60.
Hjern, B. and D.O. Porter (1981) “Implementation structures: a new unit of
administrative analysis”, Organization Studies, 2 (3), 211–27.
Ho, L-S. (2011) Public Policy and the Public Interest (London: Routledge).
Hogwood, B.W. and B.G. Peters (1983) Policy Dynamics (Brighton: Wheatsheaf).
200 ADVANCED INTRODUCTION TO PUBLIC POLICY

Hönke, J. (2010) “New political topographies: mining companies and indirect

discharge in Southern Katanga (DRC)”, Politique africane 120, 105–27.
Hood, C. (1986) The Tools of Government (Chatham, NJ: Chatham House).
Hood, C. (2011) The Blame Game: Spin, Bureaucracy and Self-Preservation in
Government (Princeton, NJ: Princeton University Press).
Hood, C. and H. Margetts (2009) Tools of Government in a Digital Age (Basingstoke:
Palgrave Macmillan).
Hood, C., O. James, B.G. Peters and C. Scott (2004) Controlling Modern
Government: Variety, Commonality and Change (Cheltenham, UK and
Northampton, MA, USA: Edward Elgar).
Hooghe, L. and G. Marks (2003) “Unraveling the central state, but how? Types of
multilevel governance”, American Political Science Review 103, 233–43.
Hoornbeek, J.A. and B.G. Peters (2017) “Understanding policy problems: a refine-
ment of previous work”, Policy & Society 36, 365–84.
Hoppe, R. (2010) The Governance of Problems: Puzzling, Powering and
Participation (Bristol: Policy Press).
Houghton, D.P. (1998) “Analogical reasoning and policymaking: where and when
is it used?”, Policy Sciences 31, 151–76.
House, R.S and E. Araral (2013) “The institutional analysis and development
framework”, in E. Araral, S. Fritzen, M. Howlett, M. Ramesh and X. Wu (eds),
Routledge Handbook of Public Policy (London: Routledge).
Howlett, M. (2001) Designing Public Policy: Principles and Instruments (London:
Routledge).
Howlett, M. (2019) “Procedural policy tools and the temporal dimensions of
policy design”, International Review of Public Policy 1, 27–45.
Howlett, M. and J. Rayner (2007) “Design principles for policy mixes: cohesion
and coherence in ‘new governance arrangements’”, Policy and Society 26, 1–18.
Hupe, P. (2007) “The policy cycle”, in B.G. Peters and J. Pierre (eds), Handbook of
Public Policy (London: Sage).
Hupe, P. (2013) “Dimensions of discretion: specifying the object of street-level
bureaucracy research”, Der Moderne Staat, https://​budrich​-journals​.de/​index​
.php/​dms/​article/​view/​15317 (accessed November 2020).
Hupe, P. (2019) Research Handbook on Street Level Bureaucracy (Cheltenham,
UK and Northampton, MA, USA: Edward Elgar).
Ingram, H.B. and A.L. Schneider (1990) “Improving implementation through
framing smarter statutes”, Journal of Public Policy 10, 61–87.
Interorganizational Committee (2003) “Principles and guidelines for social impact
assessment in the USA”, Impact Assessment and Project Appraisal 21, 231–50.
Isaacs, D., H. Kilham, J. Leask and B. Tobin (2009) “Ethical issues in vaccination”,
Vaccine 27, 615–18.
Jacobs, A.M. (2011) Governing in the Long Term: Democracy and the Politics of
Investment (Cambridge: Cambridge University Press).
Jacobs, L.R. and T. Skocpol (2005) Inequality and American Democracy: What
We Know and What We Need to Learn (New York; Russell Sage Foundation).
James, T.E. and P.D. Jorgensen (2009) “Policy knowledge, policy formulation, and
change: revisiting a foundational question”, Policy Studies Journal 37, 141–62.
 REFERENCES 201

Jann, W. and K. Weigrich (2007) “Theories of the policy process”, in F. Fischer,

G.J. Miller and M.S. Sidney (eds), Handbook of Public Policy Analysis (Boca
Raton, FL: CRC Press).
Jay, S., C. Jones, P. Slinn and C. Wood (2007) “Environmental impact assess-
ment: retrospect and prospect”, Environmental Impact Assessment Review 27,
287–300.
Jenkins-Smith, H.C. and P.A. Sabatier (1999) “The Advocacy Coalition
Framework: an assessment”, in P.A. Sabatier (ed.), Theories of the Policy Process
(Boulder, CO: Westview Press).
John, P. (2012) Analyzing Public Policy (London: Routledge).
John, P. (2013) “All tools are informational: how information and persuasion
define the tools of government”, Policy & Politics 41, 605–20.
Jones, B.D. (2001) Politics and the Architecture of Choice: Bounded Rationality and
Governance (Chicago: University of Chicago Press).
Jones, B.D. and F.R. Baumgartner (2012) “From there to here: punctuated equilib-
rium to the general punctuation thesis to a theory of government information
processing”, Policy Studies Journal 40, 1–20.
Jones, C.O. (1984) An Introduction to the Study of Public Policy, 2nd edn
(Monterey, CA: Brooks-Cole).
Kerwin, C.M. and S. Furlong (2017) Rule-Making: How Government Agencies
Write Law and Make Policy, 5th edn (Washington, DC: CQ Press).
Kingdon, J.W. (1985) Agendas, Alternatives and Public Policies (New York:
Longman), reprinted in 2003.
Kirschen, E.S. (1964) Economic Policy in Our Time (Amsterdam: North-Holland).
Kiser, L.L. and E. Ostrom (2000) “The three worlds of action: a metatheoretical
synthesis of institutional approaches”, in M.D. McGinnis (ed.), Polycentric
Games and Institutions: Readings from the Workshop in Political Theory (Ann
Arbor: University of Michigan Press).
Knoepfel, P. (2018) Public Policy Resources (Bristol: Policy Press).
Krugman, P. (2014) “The myth of German austerity”, New York Times, 23
February.
Laegreid, P. and K. Verhoest (2010) Governance of Public Sector Organizations:
Proliferation, Autonomy, and Performance (Basingstoke: Palgrave Macmillan).
Landry, R., M. Lamari and N. Amara (2003) “The extent and determinants
of the utilization of university research in government agencies”, Public
Administration Review 63, 192–205.
Lane, J.-E. (1983) “The concept of implementation”, Statsvetenskapliga Tidskrift,
17–40.
Lascoumbes, P. and P. Le Gales (2007) “Understanding public policy through its
instruments – from the nature of instruments to the sociology of public policy
instrumentation”, Governance 20, 1–21.
Lasswell, H.D. (1936) Politics: Who Gets What When, How (New York:
McGraw-Hill).
Lasswell, H.D. (1956) The Decision Process; Seven Categories of Functional
Analysis (College Park, MD: College of Business and Public Administration,
University of Maryland).
Lefebvre, A. (2019) Basic Income: The Finnish Experiment 2017–2018 (Paris:
Librinova).
202 ADVANCED INTRODUCTION TO PUBLIC POLICY

Leppo, K., T. Stahl, M. Wismar, E. Ollila and E. Lahtinen (2013) Health in All
Policies: Prospects and Potentials (Helsinki: Ministry of Social Affairs and
Health).
Lerner, D. and H.D. Lasswell (1951) The Policy Sciences (Palo Alto: Stanford
University Press).
Levi, M. (1997) Consent, Dissent and Patriotism (Cambridge: Cambridge
University Press).
Levin, K., B. Cashore, S. Bernstein and G. Auld (2010) “Playing it forward: path
dependency, progressive incrementalism and the ‘super wicked’ problem of
climate change”, Unpublished paper.
Levin, K., B. Cashore, S. Bernstein and G. Auld (2012) “Overcoming the tragedy
of super wicked problems: constraining our future selves to ameliorate global
climate change”, Policy Sciences 45, 121–52.
Ley, A.J. and E.P. Weber (2015) “The adaptive venue shopping framework: how
emergent groups choose environmental policymaking venues”, Journal of
Environmental Policy 24, 703–22.
Lijphart, A. (1984), Democracies: Patterns of Majoritarian and Consensus
Government in Twenty-One Countries (New Haven, CT: Yale University Press).
Lindblom, C.E. (1965) The Intelligence of Democracy: Decision Making Through
Mutual Adjustment (New York: Free Press).
Linder, S.H. and B.G. Peters (1984) “From social theory to policy design”, Journal
of Public Policy 4, 237–59.
Linder, S.H. and B.G. Peters (1987) “A design perspective on policy implementa-
tion: the fallacy of misplaced precision”, Review of Policy Research 6, 459–75.
Linder, S.H. and B.G. Peters (1989) “Instruments of government: perceptions and
contexts”, Journal of Public Policy 9, 35–58.
Lipsky, M. (2010) Street-Level Bureaucracy: Dilemmas of the Individual in Public
Services, Updated edn (New York: Russell Sage Foundation).
Loomis, J.B. (2011) “Incorporating distributional issues into benefit cost analysis:
why, how and two empirical examples using non-market valuation, Journal of
Benefit-Cost Analysis 2, 1–24.
Lowi, T.J. (1964) “American business, public policy, case studies and political
theory”, World Politics 16, 677–715.
Lowi, T.J. (1967) The End of Liberalism: Ideology, Policy and the Crisis of Public
Authority (New York: W.W. Norton).
Lowi, T.J. (1972) “Four systems of politics, policy and choice”, Public
Administration Review 32, 298–310.
Lukes, S. (2004) Power: A Radical View, 2nd edn (Basingstoke: Palgrave
Macmillan).
Lundqvist, L. (1980) The Hare and the Tortoise: Clean Air Policies in the United
States and Sweden (Ann Arbor: University of Michigan Press).
Macdonald, D. (2001) “Coerciveness and the selection of environmental policy
instruments”, Canadian Public Administration 44, 161–87.
Macdonald, R. (2005) “The Swiss army knife of government”, in P. Eliadis, M.
Hill and M. Howlett (eds), Designing Government (Montreal: McGill/Queens
University Press).
MacMahan, J. (2006) The Ethics of Killing: Problems at the Margin of Life (Oxford:
Oxford University Press).
 REFERENCES 203

Madrick, J. (2009) The Case for Big Government (Princeton, NJ: Princeton
University Press).
Mahoney, J. and K. Thelen (2010) Explaining Institutional Change: Ambiguity,
Agency and Power (Cambridge: Cambridge University Press).
Majone, G. (2002a) “The precautionary principle and its policy implications”,
Journal of Common Market Studies 40, 89–109.
Majone, G. (2002b) Regulating Europe (London: Routledge).
Maor, M. (2012) “Policy overreaction”, Journal of Public Policy 32, 231–59.
Maor, M. (2019) “Overreaction and bubbles in politics and policy”, in A. Mintz
and L. Terris (eds), Oxford Handbook on Behavioral Political Science. Oxford
Handbooks Online (Oxford: Oxford University Press).
March, J.G. and J.P. Olsen (1989) Rediscovering Institutions: The Organizational
Basis of Politics (New York: The Free Press).
Marier, P. (2013) “Policy learning”, in E. Araral, S. Fritzen, M. Howlett. M. Ramesh
and X. Wu (eds), Routledge Handbook of Public Policy (London: Routledge).
Marshall, T.H. (1964) Class, Citizenship and Social Development (Garden City,
NY: Doubleday).
May, P.J. (1992) “Policy failure and learning”, Journal of Public Policy 12, 187–206.
May, P.J. and A.E. Jochim (2013) “Policy regime perspective: policies, politics and
governing”, Policy Studies Journal 39, 285–305.
May, P.J., A.E. Jochim and B. Pump (2010) “Boundary-spanning policy problems:
politics and policymaking”, Paper presented at the Annual Meeting of the
American Political Science Association, Washington, DC.
Maynard-Mooney, S. and M. Musheno (2003) Cops, Teachers, Counselors:
Narratives on Street-Level Judgment (Ann Arbor: University of Michigan
Press).
McCain, R.A. (2009) Game Theory and Public Policy (Cheltenham, UK and
Northampton, MA, USA: Edward Elgar).
McConnell, A. (2010) Understanding Policy Success: Rethinking Public Policy
(Basingstoke: Palgrave Macmillan).
McConnell, A. (2019) “The use of placebo ideas to escape policy traps”, European
Journal of Public Policy 27, 957–6.
McCool, D. (1998) “The subsytem family of concepts: a critique and a proposal”,
Political Research Quarterly 51, 551–70.
McCubbins, M.D., R. Noll and B. Weingast (1989) “Structure and process, politics
and policy: administrative arrangements and the political control of agencies”,
Virginia Law Review 75, 431–82.
McDonnell, L.M. and R.F. Elmore (1987) “Getting the job done: alternative policy
instruments”, Educational Evaluation and Policy Analysis 9, 133–52.
McElroy, D. (2013) “UN warning over legal ‘designer drugs”, The Telegraph, 16
December.
McFarland, A.S. (2004) Neopluralism: The Evolution of Political Process Theory
(Lawrence: University Press of Kansas).
McGee, Z.A. and B.D. Jones (2019) “Reconceptualizing the policy subsystem:
integration with complexity theory and social network analysis”, Policy Studies
Journal 47, S138–58.
McKibbin, W.J. (2007) “The economics of international policy coordination”,
Economic Record 64, 241–53.
204 ADVANCED INTRODUCTION TO PUBLIC POLICY

Medina, L.F. (2007). A Unified Theory of Collective Action and Social Change (Ann
Arbor: University of Michigan Press).
Meltsner, A.J. (1972) “Political feasibility and policy analysis”, Public
Administration Review 32, 859–67.
Merton, R.K. (1936) “The unanticipated consequences of purposive social action”,
American Sociological Review 1, 894–904.
Miller, D. (2005) “Methods for evaluating environmental justice – approaches to
implementing Executive Order 12898”, in D. Miller and D. Pattassini (eds),
Beyond Cost Benefit Analysis (Aldershot: Ashgate).
Mintrom, M. and P. Norman (2009) “Policy entrepreneurship and policy change”,
Policy Studies Journal 37, 649–67.
Mishan, E.J. (1967) Cost-Benefit Analysis (London: Allen & Unwin).
Mishan, E.J. and E. Quah (2007) Cost-Benefit Analysis, 5th edn (London:
Routledge).
Molina, O. and M. Rhodes (2002) “Corporatism: the past, present and future of
a concept”, Annual Review of Political Science 5, 305–31.
Moore, M.A., A.E. Boardman, A.R. Vining, D.L. Weimer and D H. Greenberg
(2004) “‘Just give me the number!’: practical values for the social discount rate”,
Journal of Public Policy Analysis and Management 23, 789–812.
Mitchell, R.C. and R.T. Carson (2013) Using Surveys to Value Public Goods
(London: Routledge).
Morag-Levine, N. (2009) Chasing the Wind: Regulating Air Pollution in the
Common Law State (Princeton, NJ: Princeton University Press).
Mörth, U. (2004) Soft Law in Governance and Regulation: An Interdisciplinary
Analysis (Cheltenham, UK and Northampton, MA, USA: Edward Elgar).
Mucciaroni, G. (2013) “The garbage can model and the politics of policymaking”,
in E. Araral, S. Fritzen, M. Howlett. M. Ramesh and X. Wu (eds), Routledge
Handbook of Public Policy (London: Routledge).
Muller, J.Z. (2018) The Tyranny of Metrics (Princeton, NJ: Princeton University
Press).
Mulley, A., J. Wennberg, J. Weinstein, E. Fisher and A. Binagawho (2013)
“Embracing delivery science for universal health care”, The Lancet 382, 25–6.
Musgrave, R.A. (1957) “A multiple theory of budget determination”, FinanzArchiv
25, 33–43.
Nelson, B.J. (1984) Making an Issue of Child Abuse: Political Agenda-Setting for
Social Problems (Chicago: University of Chicago Press).
Nelson, R.R. (1977) The Moon and the Ghetto (New York: W.W. Norton).
Nelson, R.R. (2011) “The Moon and the Ghetto revisited”, Science and Public
Policy 38, 681–90.
New York Times (2019) “Finland’s basic income trial boost happiness, but not
employment”, 9 February.
North, D.C. (1993) “Institutions and credible commitment”, Journal of
Institutional and Theoretical Economics 149, 11–23.
Nowrasteh, A. and S. Cole (2014) Building a Wall Around the Welfare State, Not
the Country (Washington, DC: Cato Institute).
O’Riordan, T. and J. Cameron (2013) Interpreting the Precautionary Principle
(London: Routledge).
 REFERENCES 205

O’Sullivan, A. and K. Gibb (2003) Housing Economics and Public Policy (Oxford:
Blackwell Scientific).
O’Toole, L.J. (2000) “Research on policy implementation: assessments and pros-
pects”, Journal of Public Administration Research and Theory 10, 263–88.
O’Toole, L.J. (2011) “Interorganization relations in implementation”, in B.G.
Peters and J. Pierre (eds), Handbook of Public Administration, 2nd edn
(London: Sage).
Olson, M. (1965) The Logic of Collective Action: Public Goods and the Theory of
Groups (Cambridge, MA: Harvard University Press).
Organisation for Economic Co-operation and Development (2010) Regulatory
Policy and Sustainable Growth (Paris: OECD).
Ostrom, E. (1990) Governing the Commons: The Evolution of Institutions of
Collective Action (Cambridge: Cambridge University Press).
Ostrom, E. and X. Basurto (2011) “Crafting analytical tools to study institutional
change”, Journal of Institutional Economics 7, 317–43.
Otsuki, T., J.S. Wilson and M. Sawedeh (2001) “Saving two in a billion: quanti-
fying the trade effects of European food safety standards on African exports”,
Food Policy 26, 496–514.
Pawson, R. (2006) Evidence-Based Policy: A Realist Perspective (London: Sage).
Pawson, R. and N. Tilly (1997) Realistic Evaluation (London: Sage).
Payan, T. (2006) Cops, Soldiers and Diplomats: Explaining Agency Behavior in the
War on Drugs (Lanham, MD: Lexington Books).
Perrow, C. (1984) Normal Accidents: Living with High Risk Technologies (New
York: Basic Basic Books).
Peters, B.G. (2001a) The Future of Governing: Four Emerging Models, rev. edn
(Lawrence: University Press of Kansas).
Peters, B.G. (2001b) “The politics of policy instruments”, in L.M. Salamon (ed.),
Handbook of Policy Instruments (New York: Oxford University Press).
Peters, B.G. (2007) “Performance-based accountability”, in A. Shah (ed.),
Performance Accountability and Combating Corruption (Washington, DC:
World Bank).
Peters, B.G. (2013) American Public Policy: Promise and Performance, 11th edn
(Washington, DC; CQ Press).
Peters, B.G. (2014a) Pursuing Horizontal Management: The Politics of Public
Sector Coordination (Lawrence: University Press of Kansas).
Peters, B.G. (2014b) “Implementation structures as institutions”, Public Policy
and Administration 29, 84–105.
Peters, B.G. (2014c) “Cybernetic models of governance”, in D. Levi-Faur (ed.),
The Oxford Handbook of Governance (Oxford: Oxford University Press).
Peters, B.G. (2015) “State failure, governance failure, and policy failure: exploring
the linkages”, Public Policy and Administration 30, 261–76.
Peters, B.G. (2018) Policy Problems and Policy Design (Cheltenham, UK and
Northampton, MA, USA: Edward Elgar).
Peters, B.G. (2020) “Can we be so casual about being causal?”, Journal of
Comparative Policy Analysis (forthcoming).
Peters, B.G. (2021) American Public Policy: Promise and Performance, 12th edn
(Washington, DC: CQ Press).
206 ADVANCED INTRODUCTION TO PUBLIC POLICY

Peters, B.G. and G. Fontaine (2021) Handbook of Policy Design (Cheltenham, UK

and Northampton, MA, USA: Edward Elgar).
Peters, B.G., E. Schröter and P. von Maravić (2013) The Politics of Representative
Bureaucracy (Cheltenham, UK and Northampton, MA, USA: Edward Elgar).
Petridou, E. (2014) “Theories of the policy process: contemporary scholarship and
future directions”, Policy Studies Journal 42 (S1), S12–S32.
Pierre, J. (2014) “Can urban regimes travel in time and space?”, Urban Affairs
Review 10, 85–102.
Pierre, J., B.G. Peters and J. de Fine Licht (2018) “Is auditing the new evaluation?
Can it be? Should it be?”, International Journal of Public Sector Management
31, 726–39.
Piven, F.F. and R.A. Cloward (1993) Regulating the Poor, 2nd edn (New York:
Vintage).
Poama, A. (2019) “Application or construction?: two types of public policy
ethics”, in A. Lever and A. Poama (eds), Routledge Handbook of Ethics and
Public Policy (London: Routledge).
Pollitt, C. (2013) “The logics of performance management”, Evaluation 19,
346–63.
Pollitt, C. and C. Talbot (2004) Unbundled Government: A Critical Analysis
of Global Trend to Agencies, Quangos and Contractualization (London:
Routledge).
Posner, E.A. and A. Vermeule (2003) “Reparations for slavery and other historical
injustices”, Columbia Law Review 103, 689–748.
Pralle, S.B. (2003) “Venue shopping, political strategy and policy change: the
internationalization of Canadian forest policy”, Journal of Public Policy 23,
23360.
Pressman, J.L. and A. Wildavsky (1974) Implementation (Berkeley: University of
California Press).
Quill, T.E. (2005) “Terri Schiavo – a tragedy compounded”, The New England
Journal of Medicine 352, 1630–3.
Rachels, J. (1986) The End of Life: Euthanasia and Mortality (Oxford: Oxford
University Press).
Rawlinson, K. (2013) “Police crime figures being manipulated, admits Chief
Inspector”, The Guardian, 18 December.
Rawls, J. (1971) A Theory of Justice (Cambridge, MA: Harvard University Press).
Rawls, J. (2001) Justice as Fairness: A Restatement (Cambridge, MA: Harvard
University Press).
Rhodes, M. (2001) “The political economy of social pacts: ‘competitive corporat-
ism’ and European welfare reform”, in P. Pierson (ed.), The New Politics of the
Welfare State (Oxford: Oxford University Press).
Richardson, J.J. (1982) Policy Styles in Western Europe (London: Allen and
Unwin).
Ridde, V. (2009) “Policy implementation in an African state: an extension of
Kingdon’s multiple streams approach”, Public Administration 87, 938–54.
Rittel, H.W.J. and M.M. Webber (1973), “Dilemmas in the general theory of plan-
ning”, Policy Sciences 4, 155–69.
 REFERENCES 207

Roberts, W.T. (2004) “Nongovernmental organizations and the state in the

developing world”, Unpublished Dissertation, Department of Political Science,
University of Arizona.
Robinson, S.E. and K.J. Meier (2006) “Path dependence and organizational
behavior – bureaucracy and social promotion”, American Review of Public
Administration 36 (3), 241–60.
Rose, R. (1993) Lesson-Drawing in Public Policy: A Guide to Learning Across Time
and Space (Chatham, NJ: Chatham House).
Rose-Ackerman, S. (1996) Controlling Environmental Policy: The Limits of Public
Law in Germany and the United States (New Haven, CT: Yale University Press).
Rossi, P.H., M.W. Lipsey and H.E. Freeman (2004) Evaluation: A Systematic
Approach, 7th edn (Thousand Oaks, CA: Sage).
Sabatier, P.A. and H. Jenkins-Smith (1999) “The Advocacy-Coalition Framework:
an assessment”, in P.A. Sabatier (ed.), Theories of the Policy Process (Bounder,
CO; Westview Press).
Saetren, H. (2014) “Implementing the third generation research in policy imple-
mentation research: an empirical assessment”, Public Policy and Administration
29, 84–105.
Salamon, L.M. (1979) “The time dimension in policy evaluation: the case of New
Deal land reform”, Public Policy 27, 129–83.
Salamon, L.M. (2001a) Handbook of Policy Instruments (New York: Oxford
University Press).
Salamon, L.M. (2001b) “Introduction”, in L.M. Salamon (ed.), Handbook of Policy
Instruments (New York: Oxford University Press).
Sandmo, A. (2000). The Public Economics of the Environment (Oxford: Oxford
University Press).
Scharpf, F.W. (1986) “Policy failure and institutional reform: why should form
follow function”, International Social Science Journal 38, 179–89.
Scharpf, F.W. (1988) “The joint decision trap: lessons from German Federalism
and European integration”, Public Administration 66, 239–78.
Scharpf, F.W. (1997) Games Real Actors Play: Actor-Centered Institutionalism in
Policy Research (Boulder, CO: Westview Press).
Scharpf, F.W. (2009) “Legitimacy in the multilevel European polity”, European
Political Science Review 1, 173–94.
Schattschneider, E.E. (1962) The Semi-Sovereign People (New York: Holt, Rinehart
and Winston).
Scheuerman, W.E. (2006) “Survey articles: emergency powers and the rule of law
since 9/11”, Journal of Political Philosophy 14, 61–84.
Schillemans, T. and J. Pierre (2016) “Entangling and disentangling governance
and the media”, Policy & Politics 44, 1–8.
Schlager, E. and M. Cox (2018) “The IAD Framework and the SES Framework: an
introduction and assessment of the Ostrom Workshop Frameworks”, in C.M.
Weible and P.A. Sabatier (eds), Theories of the Policy Process, 4th edn (Boulder,
CO: Westview Press).
Schneider, A.L. (2013) “Policy design and transfer”, in E. Araral, S. Fritzen, M.
Howlett. M. Ramesh and X. Wu (eds), Routledge Handbook of Public Policy
(London: Routledge).
208 ADVANCED INTRODUCTION TO PUBLIC POLICY

Schneider, A. and H. Ingram (1988) “Systematically pinching ideas: a comparative

approach to policy design”, Journal of Public Policy 8, 61–80.
Schneider, A. and H. Ingram (1993) “Social construction of target populations:
implications for politics and policy”, American Political Science Review 87,
334–47.
Schön, D.A. and M. Rein (1994) Frame Reflection: Solving Intractable Policy
Disputes (New York: Basic Books).
Schulman, P.R. (1980) Large-Scale Policymaking (New York: Elsevier-North
Holland).
Schwartz, J.E. (1987) America’s Hidden Success: A Reassessment of Twenty Years of
Public Policy (New York: W.W. Norton).
Scriven, M. (1991) “Pros and cons about goal free evaluation”, American Journal
of Evaluation 12, 155–62.
Self, P. (1975) Econocrats and the Policy Process: Nonsense on Stilts (London:
Macmillan).
Sen, A. (2009) The Idea of Justice (Cambridge, MA: Harvard University Press).
Shadish, W.R., T D. Cook and D.T. Campbell (2001) Experimental and
Quasi-Experimental Designs for Generalized Causal Inference (Boston:
Houghton Mifflin).
Shanahan, E.A., M.D. Jones and M.K. McBeth (2018) “How to conduct a narrative
policy framework study”, The Social Science Journal 55, 332–45.
Sieber, S. (1980) Fatal Remedies: Dilemmas of Social Intervention (New York:
Plenum).
Simon, H.A. (1947) Administrative Behavior (New York: Free Press).
Simpson, D. and J. Walker (1987) “Extending cost-benefit analysis for energy
investment choices”, Energy Policy 15, 217–27.
Skocpol, T. (1992) Protecting Soldiers and Mothers: The Political Origins of Social
Policy in the United States (Cambridge, MA: Belknap Press).
Slemrod, J. (2007) “Cheating ourselves: the economics of tax evasion”, Journal of
Economic Perspectives 21, 25–48.
Slovic, P. (2000) The Perception of Risk (London: Earthscan).
Snyder, B.F. (2015) “Tax and trade: a hybrid climate policy instrument to control
carbon prices and emissions”, Climate Policy 15, 743–50.
Somashekar, S. and J. Millman (2014) “Obamacare enrollment reaches 7.5
million”, The Washington Post, 10 April.
Sørenson, E. and J. Torfing (2007) Theories of Democratic Network Governance
(Basingstoke: Palgrave Macmillan).
Stoker, R.P. (1991) Reluctant Partners: Implementing Federal Policy (Pittsburgh:
University of Pittsburgh Press).
Stone, D.A. (1997) Policy Paradox: The Art of Political Decision-Making (New
York: W.W. Norton).
Stone, D.A. (2002) Policy Paradox: The Art of Political Decision-Making, rev. edn
(New York: W.W. Norton).
Sum, N-L. and B. Jessop (2013) “Competitiveness, the knowledge-based economy,
and higher education”, Journal of the Knowledge Economy 4, 24–44.
Sunstein, C.R. (2014) Valuing Life: Humanizing the Regulatory State (Chicago:
University of Chicago Press).
 REFERENCES 209

Sydow, J., G. Schreyögg and J. Koch (2009) “Organizational path dependence:

opening the black box”, Academy of Management Review 34, 689–709.
Taylor, S. (1984) Making Bureaucracies Think: The Environmental Impact
Statement Strategy of Administrative Reform (Palo Alto, CA: Stanford
University Press).
Termeer, C., A. Dewulf and M. Van Lieshout (2010) “Disentangling scale
approaches in governance research: comparing monocentric, multilevel and
adaptive governance”, Ecology and Society 15, 29–46.
Thaler, R.W. and C.R. Sunstein (2003) “Liberal paternalism”, American Economic
Review 93, 175–9.
Thaler, R.W. and C.R. Sunstein (2008) Nudge: Improving Decisions About Health,
Wealth and Happiness (New Haven, CT: Yale University Press).
Thomann, E., N. van Engen and L. Tummers (2018) “The necessity of discretion:
a behavioral evaluation of bottom-up implementation theory”, Journal of
Public Administration Research and Theory 28, 583–691.
Thompson, D.F. (1999) “Democratic secrecy”, Political Science Quarterly 114,
181–93.
Tilly, N. (2000) “Realistic evaluation: an overview”, Speech to Danish Evaluation
Society, Copenhagen, September.
Torfing, J., B.G. Peters, J. Pierre and E. Sörensen (2012), Interactive Governance:
Advancing the Paradigm (Oxford: Oxford University Press).
Truog, R.D. (2008) “Consent for organ donations – balancing conflicting ethical
obligations”, New England Journal of Medicine 358, 1209–11.
Truog, R.D., C. Mitchell and G.Q. Daley (2020) “The toughest triage – allocating
ventilators in a pandemic”, New England Journal of Medicine 382, 1975–7.
Tsebelis, G. (2000) Veto Players: How Political Institutions Work (Princeton, NJ:
Princeton University Press).
UNICEF (2005) The “Rights” Start to Life: A Statistical Analysis of Birth Registration
(New York: UNICEF).
Van Dooren, W., G. Bouckaert and J.A. Halligan (2010) Performance Management
in the Public Sector (London: Routledge).
Van Hulst, M. and D. Yanow (2016) “From policy ‘Frames’ to ‘Framing’:
theorizing a more dynamic, political approach”, American Review of Public
Administration 46, 92–112.
Vanderbroght, Y. and S. Yamasaki (2004) “Des cas logiques ... contradictoires”,
Revue internationale de politique compareé 11, 51–66.
Vedung, E. (2007) “Policy evaluation”, in B.G. Peters and J. Pierre (eds), Handbook
of Public Policy (London: Sage).
Vedung, E. (2010) “Four waves of evaluation diffusion”, Evaluation 16, 263–77.
Vedung, E. (2013) “Six models of evaluation”, in E. Araral, S. Fritzen, M. Howlett.
M. Ramesh and X. Wu (eds), Routledge Handbook of Public Policy (London:
Routledge).
Visser, J. (2007) “Deutero-learning in organizations: a review and a reformula-
tion”, Academy of Management Review 32, 659–67.
Wald, M.L. (2012) Court Backs E.P.A. over emissions limits intended to reduce
global warming”, New York Times, 26 June.
Walker, J. (1977) “Setting the agenda in the US Senate; a theory of problem selec-
tion”, British Journal of Political Science 7, 423–56.
210 ADVANCED INTRODUCTION TO PUBLIC POLICY

Wallis, J. and B. Dollery (1999) Market Failure, Government Failure, Leadership

and Public Policy (New York: St. Martin’s Press).
Walzer, M. (1973) “Political action: the problem of dirty hands”, Philosophy and
Public Affairs 2, 160–80.
Watson, C. and L. Mullen (2020) “Risk-informed decision making: more impor-
tant than ever before”, Health Security 18, 153–4.
Weaver, R.K. and B.A. Rockman (1993) Do Institutions Matter? Government
Capabilities in the United States and Abroad (Washington, DC: The Brookings
Institution).
Weber, E.U. and P.C. Stern (2011) “Public understanding of climate change in the
United States”, American Psychologist 66, 315–28.
Weber, M., P.P.J. Driessen and H.A.C. Runhaar (2014) “Evaluating environ-
mental policy instruments mixes; a methodology illustrated by noise policy
in the Netherlands”, Journal of Environmental Planning and Management 57,
1381–97.
Weible, C. (2014) “Introducing the scope and focus of policy process research
and theory”, in P.A. Sabatier and C. Weible (eds), Theories of the Policy Process
(Boulder, CO: Westview Press).
Weible, C.M., P.A. Sabatier, H.C. Jenkins-Smith, D. Nohrstedt, A.D. Henry and P.
deLeon (2011) “A quarter-century of the Advocacy-Coalition Framework: an
introduction to the special issue”, Policy Studies Journal 39, 349–60.
Weimer, D.L. (1993) “The current state of design craft: borrowing, tinkering and
problem solving, Public Administration Review 53, 110–20.
Whittington, D. and D. MacRae Jr (1996) “The issue of standing in cost-benefit
analysis”, Journal of Policy Analysis and Management 5, 665–82.
Wildavsky, A. (1980) “Policy as its own cause”, in A. Wildavsky, The Art and Craft
of Policy Analysis (London: Macmillan), reprinted in 2018.
Wilson, J.Q. (1980) The Politics of Regulation (New York: Basic Books).
Winter, S. (2011) “The implementation perspective”, in B.G. Peters and J. Pierre
(eds), Handbook of Public Administration, 2nd edn (London: Sage).
Wolf, C. (1987) “Market and non-market failures: comparison and assessment”,
Journal of Public Policy 7, 43–70.
Workman, S., J. Shafran and T. Bark (2017) “Problem definition and information
provision by federal bureaucrats”, Cognitive Systems Research 43, 140–52.
Workman, S., B.D. Jones and A.E. Jochim (2009) “Information processing and
policy dynamics”, Policy Studies Journal 37, 75–92.
Xue, J. (2012) Growth with Inequality: An International Comparison on Income
Inequality (Singapore: World Scientific Publishers).
Zito, A. (2001) “Epistemic communities, collective entrepreneurship and
European integration”, Journal of European Public Policy 8, 585–603.
Index

abortion 172 child abuse 82

abuse 82 child benefit 174
accountability 125, 136, 148, 149 choices 44–5, 116, 118–19, 123, 178,
adaptive perspective 8–9 182
Adler, Matthew 161 freedom of 168–70
administrative man 48 tragic 30
advocacy coalition framework (ACF) civil servants 130
63–5, 68, 70, 188 classification 108–16
Affordable Care Act (US) 97, 122, 169 clearance points 89–90, 91, 93
agency 69, 186 client model 138
agendas 71–82 climate policy 26
agenda-setting 72–82, 190 coercion 61, 170
alarmed discovery 80 collegial model 139
Allison, Graham 45, 51 command and control 107, 108, 170
Army Corps of Engineers 28–9 comparative analysis 187–90
auditing 144–5 competitiveness policy 21
Australia 83 complexity 26–8, 34
autonomy 170, 171, 172 compliance 95, 107
confluence of events 56–8
backward mapping 91–2, 93, 104 Constitutivists 117–18
Béland, Daniel 146–7 constructivist position 3–4, 113
benefits 62, 173 consumer’s surplus 153
Birkland, T.A. 81 Contingentists 117
Blyth, Mark 66 contraception 169–70
Bobrow, Davis 4 cooperation 184
Botterill, L. 83 coordination 36–7, 184
bottom-up perspective 102, 103 cost-benefit analysis 46, 151–63, 165,
boundary spanning problems 20–21 180
bounded rationality 9, 23, 47–53, 68 costs 16–17, 50, 51, 62, 93, 142, 171
Bowen, Elinor 90, 93 COVID-19 23, 27, 74, 82
budgets 46, 50, 73 credible commitments 186
bureaucracy 87, 89, 101–2 crime 26, 100, 159
bureaucrats, street-level 101–2 statistics 144
criminals, sentencing convicted 146
“Carrots, Sticks, and Sermons” 111 crises 68, 75, 81
certainty 28–9 Cuban Missile Crisis 45, 67

211
212 ADVANCED INTRODUCTION TO PUBLIC POLICY

Culpepper, Pepper 66 enforcement 127

cultural influences 189 entrepreneurs 57, 58, 69–70, 90
environmental impact analysis 162–3
death penalty 171 Environmental Impact Assessment
decay 140 (EIA)/EU 162
decision-makers 116–18 Environmental Impact Statements
decision-making 12, 40, 42–7, 48–53, (US) 163
55–8, 66 Environmental Protection Agency
and ACF 64 (EPA)/US 96
and punctuated equilibrium 79 epidemic diseases 23
seven stages 59 epistemic community 120–21
decision science 45–6 equality 173, 174
decision trees 154 of opportunity 167, 168
defense 22 of outcomes 168
delegation 95–6 equilibrium 78–9
demerit goods 169 equity 121, 129–30
“deserving” poor 173–5 Esping-Andersen, G. 189
detectors 110 ethics 128–30, 168, 169, 170, 175, 176,
deutero-learning 148 178
difference principle 167 European Union 162–3, 188
diffusion 121, 148 evaluation 135–42, 145–50, 151–63
digital divide 128 events, confluence of 56–8
diplomats 176 evidence-based policymaking 142,
“dirty hands” 175–6 148, 149
discount rate 157–8 evolution 8
discretion 95–6, 101–2, 105–6, 129–30 exogenous shock 79
Downs, Anthony 78, 80, 182 experimentation 27–8
drift 8, 96, 99, 103, 104 expert organizations 186
Dror, Yehezkel 45 externalities 17
drought 83
drug policy 83, 101 failure 88, 102, 103, 105, 180, 182
Dryzek, John 4 and governing in the shadows
185–7
earned-income tax credit 174 social 18–19
economic criteria 125–6, 130 fairness 166–8
economic elite 76 faith-based organizations 185–6
economic inequality 18 farmers 122
economic man 48 fatal remedies 141
economic policy 26–7 feasibility 104
effectors 110 feedback 146–8
effects, range of 153–4 Fenna, A. 83
efficiency 156–7 Finland 27–8, 146
elderly abuse 82 floods 28–9
elites 76, 78 focusing events 80–82
Elmore, R.F. 111–12 framing 5–6, 82–4
emergency 74–5 freedom of choice 168–70
empowerment 139 Freeman, Gary 188
energy policy 188 full employment 147
 INDEX 213

game theory 46 institutional analysis and development

gaming 143–4 framework (IAD) 53–5
garbage can 9, 55–7, 77, 118 institutionalism, historical 147
gender mainstreaming 121 institutions 118–19, 189–91
generations, protecting future 159 Instrumentalists 117, 120
genetically modified organisms interests 121–2
(GMOs) 29 inter-generational equity 159–60
geographical boundaries 20 internalities 19
goal attainment model 137 Interstate Highway System (US) 141
goal-free evaluation 137–8 intervention 87–8, 91–2, 103
Goodhart’s Law 143 issue-attention cycle 71, 78–80
governance 47, 48, 50, 64
failures 19 Jenkins-Smith, H.C. 63
Government Accountability Office
(US) 145 Kaldor-Hicks criterion 152, 158
government benefits 173–5 Kingdon, John 57, 80–81
Kirschen, E.S. 109, 115
habit 118–19 Kiser, Larry 53
health care 97, 147–8, 183, 184, 188 Knoepfel, Peter 111
health insurance 122, 168–9
Heclo, Hugh 65 labeling 174
hedonic pricing 155 large-scale policy problems 23–4
highways 141 Lascoumbes, P. 113
Hill, M. 60 Lasswell, Harold 1, 59
historical institutionalism 147 league tables 144
Hjern, Bennie 93 learning, and evaluation 148–9
“Hobby Lobby” case 169–70 Le Gales, P. 113
Hood, Christopher 110 legislation, as barrier to
Hoppe, Robert 65 implementation 97–8
housing 126 life
Howlett, Michael 112 preserving 170–72
Hupe, P. 60 valuing 155–6
Hurricane Katrina 28–9 Lipsky, Michael 101
logic of appropriateness 44, 165
ideas 120–21 logic of consequentiality 44, 165
implementation 87, 88–91, 92–106, logic of incrementalism 49
190–91 logic of satisficing 48
income maintenance 146 Lowi, Theodore 60–61
incrementalism 23, 49–50 lying 176
inequality 18, 158–9
informal agenda 72 Macdonald, Roderick 131
information mainstreaming 121
control of 184 Majone, Giandomenico 3
perfect 17 Managerialists 117
information processing 79 manpower training programs 141
innovation 187, 189 Maor, Moshe 80, 182
institutional agendas 72 March, J.G. 44, 49
214 ADVANCED INTRODUCTION TO PUBLIC POLICY

market actors 10–11, 52, 94, 104, 105, Organisation for Economic
185 Co-operation and Development
market failure 16–18, 185 (OECD) 162
Marshall, T.H. 174 organizational actors 50
McConnell, Allan 181 “Organizational Process” model 51
McDonnell, L.M. 111–12 organizational routines 51, 98
means-tested benefits 173, 174 organizations, internal dynamics of 50
measurement 138, 140, 143–4 Ostrom, Elinor 53
media 80
merit goods 169 pacting 66–7
monetization 31 Pareto optimality 152
monopolies 17 participation 65–7
multi-level governance 98 path dependency 58–9, 78, 147
multiple actors 23–4, 99 pensions 129
Multiple Governance Framework 60 performance indicators 144
multiple streams 55–8, 62, 68, 69, 77, performance management 100, 142–5
188 performance measurement 135,
“Muscles and Prayers” 111 142–3, 144
performance standards 100
National Audit Office (UK) 145 Perrow, Charles 29
National Environmental Protection persuading 66–7
Act (NEPA)/US 162 Planning, Programming and
national leaders 175, 176 Budgeting System (PPBS) 46
NATO (Nodality [Information], police 101–2
Authority, Treasure and policy bubbles 182
Organization) 110, 112, 115 policy diffusion 121, 148
natural monopolies 17 policy drift 8, 96, 99, 103, 104
Nelson, Richard 25 policy entrepreneurs 57, 58, 69–70, 90
neo-liberalism 11, 157 policy regimes 62–3
network governance 11 political complexity 26–7
networks 93–4 political elite 76, 78
New Jersey Income Maintenance political framing 84
Experiment 146 political models 58–63
New Public Management 11, 99, 142, political power 19, 65, 66, 67, 123
157 political science approach 110–13
nodality 31, 110 political success 182–3
nodding 144 political symbols 76
normative criteria 7, 23, 30, 40, 44, political systems 68–9, 78, 188
149–50, 165–78 politics 56, 98, 119, 124–5, 176
not-for-profit organizations 185–6 pollution 16–17
nudge 20, 114, 129, 170 the poor 122
nutrition programs 122 Porter, David 93
poverty 18–19, 159
Obamacare 122, 147–8, 168, 183 powering 65, 66–7
Olsen, J.P. 44, 49 precautionary principle 29, 160
opportunity, equality of 167, 168 preferences 43–4, 52–3
opportunity costs 153, 158 Pressman, J.L. 89–90, 189
organ donation 171–2 private goods 21–2
 INDEX 215

private sector 147 Sieber, Sam 141

problem definition 35 Simon, Herbert 48
proceduralists 119 Simpson, D. 161
process success 181 sleeper effects 140
professionalism 184 smoking 184
program success 181–2 SNAP 7, 122
public budget 73 social actors 185–6
public bureaucracy 87, 89 social assistance programs 173–4
public goods 16, 21–2 social benefits 173, 182
public interest 166 social discount rate 157–8
punctuated equilibrium theory 78–9 social failures 18–19
puzzling 65, 66–7 Social Impact Analysis (SIA) 163
social pacts 66
rationality 9, 10, 40–42, 43–53 social problems 35
Rawls, John 167 Social Security 147
realistic evaluation 142 soft law 112
recidivism 145–6 solubility 24–6
recurrent agendas 73 solutions 56
reframing 83–4 spousal abuse 82
regressive interventions 141 stages model 58–60, 68, 69
regulatory impact analysis 162 stakeholder model 138–9
regulatory issues 19–20 status quo 64
Rein, Martin 83, 84 stigma 173–5
relevance 137, 141 Stone, Deborah 4
resources, depletion 159 Strategic Environmental Assessment
returns to scale 17 (SEA)/EU 162
risk 9, 10, 28, 29–30, 161 street-level bureaucrats 101–2
risk-benefit analysis 160–61 subsystems 62–3
Rittel, H.W.J. 33–4 success 102–3, 180–83
road building 25–6 suicide 171
Rockman, B.A. 188 “super wicked” problems 34–5
routines 49, 93 Swiss Army Knife 131–2
synoptic decision-making 49
Sabatier, P.A. 63 synoptic rationality 40–41
sabotage 96 systemic agenda 72
safety devices 171
Salamon, Lester 112–13, 115 targeting 126–7, 129
satisficing 48 tax 109, 125, 127, 169, 174
scale 22–4 evasion 166
Scharpf, Fritz 185, 188–9, 190 expenditures 110–11, 126, 173,
Schattschneider, E.E. 78 174
Schön, Donald 83, 84 technical complexity 26, 27
scientific approach 3 technical evaluations 147
seatbelts 171 termination 59
secrecy 176–7 third-party actors 99
shadows 185–7 time 127–8
shock 79 tobacco farming 184
side-effects 137, 141 tools 113, 122, 123, 131
216 ADVANCED INTRODUCTION TO PUBLIC POLICY

The Tools of Government (Hood) 110 universal benefits 174

top-down perspective 102–3 unstructured problems 33–5
tragedy of the commons 17–18
tragic choices 30 vaccinations 169
transparency 176, 177 Vedung, Evert 136–7
transportation 24, 25–6 veil of ignorance 167
Treasure 110 venue-shopping 73
trust 93, 95 vertical coordination 36, 184
truth-telling 176–7 veto players 41, 42, 181, 188
turf 184 visibility 125
visible lives 172
uncertainty 9, 10 in vitro fertilization 172
“undeserving” poor 173
unemployment 74 Walker, J. 161
unemployment insurance 74–5 Weaver, R.K. 188
unintended consequences 141, 163, Webber, M.M. 33–4
182 welfare benefits 161
United States welfare economics 152
Affordable Care Act 97, 122, 169 welfare states 189
EPA 96 well-being 140, 159
and framing 83 wicked problems 20, 28, 33–5, 68
government 188 Wildavsky, Aaron 26, 89–90, 189
“Hobby Lobby” case 169–70 willingness to pay 155, 169
Interstate Highway System 141 Wilson, James Q. 62
NEPA 162 window of opportunity 55, 57, 68
New Jersey Income Maintenance World Health Organization 23
Experiment 146
Titles in the Elgar Advanced Introductions series include:
International Political Economy Comparative Constitutional Law
Benjamin J. Cohen Mark Tushnet
The Austrian School of Economics International Human Rights Law
Randall G. Holcombe Dinah L. Shelton
Cultural Economics Entrepreneurship
Ruth Towse Robert D. Hisrich
Law and Development International Tax Law
Michael J. Trebilcock and Mariana Reuven S. Avi-Yonah
Mota Prado
Public Policy
International Humanitarian Law B. Guy Peters
Robert Kolb
The Law of International
Post Keynesian Economics Organizations
J.E. King Jan Klabbers
International Intellectual Property International Environmental Law
Susy Frankel and Daniel J. Gervais Ellen Hey
Public Management and International Sales Law
Administration Clayton P. Gillette
Christopher Pollitt
Corporate Venturing
Organised Crime Robert D. Hisrich
Leslie Holmes
Public Choice
Nationalism Randall G. Holcombe
Liah Greenfeld
Private Law
Social Policy Jan M. Smits
Daniel Béland and Rianne Mahon
Consumer Behavior Analysis
Globalisation Gordon Foxall
Jonathan Michie
Behavioral Economics
Entrepreneurial Finance John F. Tomer
Hans Landström
Cost-Benefit Analysis
International Conflict and Security Robert J. Brent
Law
Nigel D. White
Environmental Impact Assessment International Trade Law
Angus Morrison-Saunders Michael J. Trebilcock and Joel
Trachtman
Comparative Constitutional Law
Second Edition European Union Law
Mark Tushnet Jacques Ziller
National Innovation Systems Planning Theory
Cristina Chaminade, Bengt-Åke Robert A. Beauregard
Lundvall and Shagufta Haneef
Tourism Destination Management
Ecological Economics Chris Ryan
Matthias Ruth
International Investment Law
Private International Law and August Reinisch
Procedure
Peter Hay Sustainable Tourism
David Weaver
Freedom of Expression
Mark Tushnet Austrian School of Economics
Second Edition
Law and Globalisation Randall G. Holcombe
Jaakko Husa
U.S. Criminal Procedure
Regional Innovation Systems Christopher Slobogin
Bjørn T. Asheim, Arne Isaksen and
Michaela Trippl Platform Economics
Robin Mansell and W. Edward
International Political Economy Steinmueller
Second Edition
Benjamin J. Cohen Public Finance
Vito Tanzi
International Tax Law
Second Edition Feminist Economics
Reuven S. Avi-Yonah Joyce P. Jacobsen

Social Innovation Human Dignity and Law

Frank Moulaert and Diana James R. May and Erin Daly
MacCallum Space Law
The Creative City Frans G. von der Dunk
Charles Landry National Accounting
John M. Hartwick
Legal Research Methods Global Production Networks
Ernst Hirsch Ballin Neil M. Coe
Privacy Law Mental Health Law
Megan Richardson Michael L. Perlin
International Human Rights Law Law and Literature
Second Edition Peter Goodrich
Dinah L. Shelton
Creative Industries
Law and Artificial Intelligence John Hartley
Woodrow Barfield and Ugo Pagallo
Global Administration Law
Politics of International Human Sabino Cassese
Rights
David P. Forsythe Housing Studies
William A.V. Clark
Community-based Conservation
Fikret Berkes Global Sports Law
Stephen F. Ross
Public Policy
B. Guy Peters