Learning in public policy analysis, modes and outcomes

A public policy and administration scholar, Claire’s main fields of interest include the politics of expertise and knowledge uti-lization; epistemic communities and advisory politics; ri

International Series on Public Policy

LEARNING IN PUBLIC POLICY

CLAIRE A DUNLOP CLAUDIO M RADAELLI PHILIPP TREIN

A N A LY S I S, M O D E S A N D O U TC O M E S

E D I T E D BY

Series Editors Guy Peters Department of Political Science University of Pittsburgh

Pittsburgh, PA, USA

The International Series on Public Policy—the official series of

International Public Policy Association, which organizes the International Conference on Public Policy—identifies major contribu-tions to the field of public policy, dealing with analytical and substantive policy and governance issues across a variety of academic disciplines A comparative and interdisciplinary venture, it examines questions of policy process and analysis, policymaking and implementation, policy instru-ments, policy change and reforms, politics and policy, encompassing

a range of approaches, theoretical, methodological, and/or empirical Relevant across the various fields of political science, sociology, anthro-pology, geography, history, and economics, this cutting edge series wel-comes contributions from academics from across disciplines and career stages, and constitutes a unique resource for public policy scholars and those teaching public policy worldwide

More information about this series at

http://www.palgrave.com/gp/series/15096

Claire A Dunlop · Claudio M Radaelli

Philipp Trein Editors Learning in Public

PolicyAnalysis, Modes and Outcomes

International Series on Public Policy

ISBN 978-3-319-76209-8 ISBN 978-3-319-76210-4 (eBook)

https://doi.org/10.1007/978-3-319-76210-4

Library of Congress Control Number: 2018933055

© The Editor(s) (if applicable) and The Author(s) 2018

This work is subject to copyright All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights

of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction

on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed.

The use of general descriptive names, registered names, trademarks, service marks, etc in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Palgrave Macmillan imprint is published by the registered company Springer

International Publishing AG part of Springer Nature

The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

For Kaye, Luisa and Fausto, Thenia

vol-or so years with wvol-ork on policy transfer, lesson drawing and diffusion.The explanation for this sustained interest in developing knowledge about policy learning is unlikely to lie exclusively in the promise that it holds for explaining policy The impact of conscious and analytical policy learning on the shape of policy often turns out on closer examination to

be less strong and direct than it appears to be at first sight Over fifteen years ago, I directed the Economic and Social Research Council initia-tive ‘Future Governance’ that took a cross-national comparative look at how policy learning develops One of the consistent findings across many

of the projects in the research programme was that what either looked like, or was claimed to be, a policy ‘borrowed’ from another was, in fact, more significantly shaped by a range of other political and organizational constraints such that it was hard to identify what, precisely, was bor-rowed over and above terminology or the germ of an idea

The rewards of understanding policy learning are above all tical Basing decision-making on a clear appreciation of what has worked and what has not—elsewhere, at different times or in different

prac-viii FOREwORD

contexts—holds enormous promise for innovation, improving the quality

of government policies and avoiding dangerous and expensive mistakes.The editors point out how far the field of policy learning has devel-oped conceptually, theoretically and empirically we know a lot more about how policy learning works, the constraints and limitations, the conditions under which it is done and the political and organizational support needed to sustain it The nature of the field is not such that we can expect to produce a list of ways in which learning can be encour-aged and promoted in public organizations: a list of dos and don’ts It retains a range of features that make it an especially difficult area for such generalization, among which one might count; learning is an individual activity, it is hard to observe it in action as it essentially refers to a frame

of mind, it has to be developed and sustained in a collective or izational environment, and the conditions that allow more or less sys-tematically drawn lessons to shape policy are as wide and variable as the conditions shaping policy itself It has the added problem arising from any branch of study that seeks to make normative recommendations: of deriving an ‘ought’ from an ‘is’

organ-what is possible, and organ-what the field has developed above all, is an understanding the place of learning in the wider context of policymak-ing; the range of constraints, heuristics, structures, norms and proce-dures that encourage and discourage the application of different forms

of policy learning As the editors point out, we still have a way to go

in developing this understanding They themselves have significantly advanced this research agenda through their earlier elaboration of dif-ferent types of learning (epistemic, reflexive learning, bargaining and hierarchical learning) The exciting part of this collection is that it opens the field of policy learning still further conceptually and empirically It develops the relationship between policy learning and other key litera-tures in understanding policymaking including agendas and policy entre-preneurship, political economy and collaborative governance through a series of fresh and imaginative papers including a range of methodologies including more traditional documentary and qualitative analysis as well

as a range of quantitative methods By carefully taking stock of the field, demonstrating new ways of looking at learning in practice, and finding new places and contexts to look at it, this collection significantly extends our understanding of learning in the wider policy process

Acknowledgements

First and foremost, we extend sincere thanks to our 15 authors who have provided such thought-provoking chapters This volume is written by a real mix of scholars—those who are the start of their learning journey and others who have been learning, unlearning and teaching for a bit longer! The cross-fertilization of ideas between different generations of researchers is crucial for all types of research, and this is certainly the case

in policy learning

Our volume emerged from a series of conference panels at the International Political Science Association’s Biennial world Congress held in Poznań in July 2016 we thank everyone who contributed to those debates, animated our thinking and helped us elevate our ambi-tions for the papers The chapters were honed further at a closed work-shop in January 2017 at the British Academy, London we thank Prof

Ed Page (B.A fellow and author of our Foreword) for his generosity in securing access to this most inspirational of venues That workshop was part-funded by the University of Exeter’s College of Social Science and International Studies via the Centre for European Governance’s research fund, and we gratefully acknowledge that support we also thank our editors at Palgrave Macmillan—Jemima warren and Oliver Foster—for making the publishing process straightforward (and quick!) Claire and Claudio wish to acknowledge the European Research Council’s grant Analysis of Learning in Regulatory Governance grant no 230267 This ERC advanced project allowed us to expand our understanding of policy learning theoretically, as well as empirically In addition, we acknowledge

x ACKNOwLEDGEMENTS

the generous support from the INSPIRES project, which was funded by the 7th Framework Program of the European Union and provided gen-erous funding for Philipp’s research

More personally, Claire thanks Claudio and Philipp for their tional contributions throughout the project, from inception to publica-tion She looks forward to continuing working together Moreover, she gives her love to Kaye who has taught her more about learning than any book! Claudio thanks Claire and Philipp, two fabulous fellow travellers who are never tired of explorations in the field of policy learning… and beyond He dedicates this book to his parents, who are still part of him but no longer with us Philipp thanks both Claire and Claudio for the inspiring and productive collaboration and looks forward to common academic journeys in the future Moreover, he sends his love and grati-tude to Thenia who guided him to the subject of policy learning

inspira-Exeter, UK

Exeter, UK

Lausanne, Switzerland

Claire A Dunlop Claudio M RadaelliPhilipp Trein

contents

1 Introduction: The Family Tree of Policy Learning 1Claire A Dunlop, Claudio M Radaelli and Philipp Trein

2 Lessons Learned and Not Learned: Bibliometric

Nihit Goyal and Michael Howlett

3 Learning in the European Commission’s Renewable

Energy Policy-Making and Climate Governance 51Katharina Rietig

4 Mechanisms of Policy Learning in the European

Semester: Pension Reforms in Belgium 75Christos Louvaris Fasois

5 Individual Learning Behaviour in Collaborative

Vidar Stevens

6 Learning from Practical Experience: Implementation

Epistemic Communities in the European Union 123Daniel Polman

xii CONTENTS

7 The Rise and Demise of Epistemic Policy Learning:

The Case of EU Biotechnology Regulation 145Falk Daviter

8 Public Versus Non-profit Housing in Canadian

Provinces: Learning, History and Cost-Benefit Analysis 167

Maroine Bendaoud

9 Blocked Learning in Greece: The Case of

Thenia Vagionaki

10 Structure, Agency and Policy Learning: Australia’s

Multinational Corporations Dilemma 215Tim Legrand

11 Median Problem Pressure and Policy Learning:

An Exploratory Analysis of European Countries 243Philipp Trein

12 The Hard Case for Learning: Explaining the Diversity

of Swiss Tobacco Advertisement Bans 267Johanna Kuenzler

13 The Policy-Making of Investment Treaties in Brazil:

Policy Learning in the Context of Late Adoption 295Martino Maggetti and Henrique Choer Moraes

14 Interdependent Policy Learning: Contextual Diffusion

of Active Labour Market Policies 317Jan Helmdag and Kati Kuitto

editors And contributors

About the Editors

Claire A Dunlop is Professor of politics and public policy at the

University of Exeter A public policy and administration scholar, Claire’s main fields of interest include the politics of expertise and knowledge uti-lization; epistemic communities and advisory politics; risk governance; policy learning and analysis; impact assessment; and policy narratives She explores these conceptual interests at the UK and EU levels principally, and most frequently in relation to agricultural, environmental and LGBT issues Claire has published more than 40 peer-reviewed journal articles

and book chapters—most recently in Policy and Politics, Policy Sciences,

International Public Management Journal, Regulation and Governance

and Journal of European Public Policy She is editor of Public Policy and

Administration.

Claudio M Radaelli is Professor of political science and Jean Monnet

Chair at the University of Exeter, where he directs the Centre for European Governance He has addressed policy learning in several research articles on Europeanization, impact assessment, impact assess-ment, the politics of best practice, policy transfer and the crisis of the Eurozone He was awarded two Advanced Grants by the European Research Council, one on learning (ALREG-Analysis of Learning in Regulatory Governance)

xiv EDITORS AND CONTRIBUTORS

Philipp Trein is a Senior Researcher in political science at the Institute

of Political, Historical and International Studies, University of Lausanne

He is spending the academic year 2017–2018 as a visiting scholar at the University of California, Berkeley His research interests cover public pol-icy, policy learning, federalism and multilevel governance, institutional analysis and economic voting Among others, his research appeared or is

forthcoming in the following outlets: Annals of the American Academy

of Political and Social Science, European Journal of Political Research, German Politics, Journal of Comparative Policy Analysis, Journal of Public Policy, Public Administration and Publius His book project Healthy or Sick? Coevolution of Health Care and Public Health in a Comparative Perspective is under contract with Cambridge University

Press

Contributors

Maroine Bendaoud is a Social Sciences and Humanities Research

Council Postdoctoral Fellow in political science at McMaster University

He completed his Ph.D at University of Montreal and research stays in several universities including Yale University and Sciences Po, Paris His work on housing policy has been supported by scientific funding agen-cies at the federal and provincial levels in Canada The chapter contained

in this book is a revised version of a paper presented at the 2016 IPSA conference, which won the Francesco Kjellberg Award for Outstanding Paper Presented by New Scholar

Henrique Choer Moraes is a diplomat working in the Ministry of

Foreign Affairs of Brazil Throughout his career, he has been involved

in a number of areas of economic diplomacy, with postings in Brasilia, Brussels (Mission of Brazil to the EU) and Montevideo (Mission of Brazil to the Mercosur) He is also the author of academic work in the field of global economic governance, including articles and chapters cov-ering topics such as transgovernmental networks, global financial govern-ance and intellectual property rights

Falk Daviter is Assistant Professor of public administration and policy

at the University of Potsdam He holds a Ph.D in Political and Social Sciences from the European University Institute, Florence, and a mas-ter of public administration from the University of Konstanz His main

EDITORS AND CONTRIBUTORS xv

research interests are the politics of expertise and the use of knowledge

in the policy process, especially policy analysis of complex and tured policy problems Other areas of specialization include EU politics and administration, regulatory policymaking and policy framing

ill-struc-Christos Louvaris Fasois is Ph.D candidate at the Department of

Political Science of the University of Amsterdam and the Department

of Social and Political Sciences at the University of Milano His research focuses on the European Semester and its effects on social and employ-ment policies at the national level, drawing evidence from the case of Belgium

Nihit Goyal is a Ph.D candidate at the Lee Kuan Yew School of Public

Policy (LKYSPP), National University of Singapore He is interested

in research on public policy theories and policy evaluation, particularly comparative policy analysis Empirically, his work has focused on issues

in energy, environment and climate change, using both qualitative and quantitative techniques Nihit has a bachelor of engineering in com-puter science from the Visvesvaraya Technological University, India, and a dual-degree master of public affairs/master in public policy from Sciences Po, Paris, and LKYSPP, Singapore He has worked as a software developer at Philips Healthcare and a research scientist at the Center for Study of Science, Technology and Policy, a think tank based in India

Jan Helmdag is a Ph.D candidate and Teaching Assistant at the

University of Greifswald His research focuses on comparative social policy analysis, labour market policies and diffusion of social policies

In his doctoral thesis, he investigates the causes of labour market icy reform in the OECD, with an emphasis on parties’ influence on the reciprocity of active and passive labour market policies Jan is also an associate researcher in the research project ‘Career Discontinuities and Pension Security—Parenthood as wage Risk’ at the Finnish Centre for Pensions and a member of the research team of the Comparative welfare Entitlements Dataset CwED2

pol-Michael Howlett is Burnaby Mountain Chair in the Department

of Political Science at Simon Fraser University and Professor in the Lee Kuan Yew School of Public Policy at the National University of Singapore He specializes in public policy analysis, political economy and resource and environmental policy His articles have been published in

xvi EDITORS AND CONTRIBUTORS

numerous professional journals in Canada, the United States, Europe, Latin America, Asia and Australia and New Zealand He is editor of

the Annual Review of Policy Design and Policy Sciences and is the rent chair of Research Committee 30 (Comparative Public Policy) of the

cur-International Political Science Association He also sits on the Executive

Committee of the International Public Policy Association.

Kati Kuitto is Senior Researcher at the Finnish Centre for Pensions in

Helsinki Previously, she was Researcher and Lecturer at the University

of Greifswald, where she received her Ph.D., and Research Assistant at the Berlin Social Science Centre wZB Kati’s research focusses on com-parative welfare state analysis, the interplay of social security and social investment policies, public pension systems and policy diffusion She

is the author of Post-Communist welfare States in European Context: Patterns of welfare Policies in Central and Eastern Europe and articles

among others in European Political Science Review, Journal of European

Social Policy and Journal of Public Health Kati is one of the Principal

Investigators of the Comparative welfare Entitlement Dataset CwED2 project

Johanna Kuenzler is a Teaching Assistant and Ph.D candidate at the

Centre for Public Management of the University of Bern She holds a Master of Arts in Swiss and Comparative Politics Her thesis focuses

on the reform of the Swiss child and adult protection services Other research interests include the multiple streams framework, public policy discourse, policy implementation and evaluation as well as studies in the fields of tobacco prevention policy and agricultural policy Methodically, she relies on qualitative comparative analysis and on various methods of qualitative analysis

Tim Legrand is a lecturer at the National Security College, Crawford

School of Public Policy at the Australian National University and is adjunct Associate Professor at the Institute for Governance and Policy Analysis at the University of Canberra His research is concerned with international policy transfer, global governance and the trans-national dimensions of security within the Anglosphere His research

has been published in leading international journals including Public

Administration, Political Studies, Review of International Studies, Security Dialogue, Policy Studies, British Politics and the Journal of Comparative Policy Analysis.

EDITORS AND CONTRIBUTORS xvii

Martino Maggetti is Associate Professor in political science at the

Institute of Political, Historical and International Studies (IEPHI) of the University of Lausanne His current research interests mainly focus

on regulatory agencies, multilevel policymaking, the domestic impact

of soft rules and transnational private governance He teaches on public policy, mixed methods and the politics of regulation His research arti-

cles have appeared in several top journals, including: The Annals of the

American Academy of Political and Social Science, Business and Society, European Journal of Political Research, European Political Science Review, International Review of Administrative Sciences, Journal of Comparative Policy Analysis, Journal of European Public Policy, Journal of Public Policy, Political Research Quarterly, Political Studies Review, Public Administration, Regulation and Governance, Swiss Political Science Review, and West European Politics His latest book is Comparative Politics: Theoretical and Methodological Challenges (Edward Elgar 2015,

co-edited with Dietmar Braun)

Ed Page is Sidney and Beatrice webb Professor of Public Policy at the

London School of Economics and Political Science’s (LSE) Department

of Government, and Fellow of the British Academy His recent lications have addressed accountability in the bureaucracy and civil ser-vice roles in the production of policy, especially in writing primary and secondary legislation Professor Page also leads the LSE-based research collective with undergraduate students which has published on policy-making issues

pub-Daniel Polman is a Ph.D candidate at the Department of Political

Science and the EUROPAL research group at the Institute for Management Research (IMR) of the Radboud University Nijmegen His Ph.D project focuses on processes of feedback from the implementation

of policies to future policy developments More specifically, his current Ph.D project studies how the domestic implementation of policy pro-grammes of the Common Agricultural Policy affects subsequent policy changes at the European level

Katharina Rietig is Lecturer/Assistant Professor in International

Politics at Newcastle University She holds a Ph.D and M.Sc from the London School of Economics and Political Science Her research inter-ests focus on agency in global governance, especially in the areas of environmental and climate change policy She examines how learning,

xviii EDITORS AND CONTRIBUTORS

leadership and multilevel governance dynamics between the European Union and United Nations influence policy change Her work appeared

in the Journal of European Public Policy, Policy Studies Journal, Policy

Sciences, Global Governance, Environmental Policy and Governance, as

well as International Environmental Agreements.

Vidar Stevens is a Researcher at the Department of Political Sciences

(research group public administration and management) at the University of Antwerp His main research interest is in the management

of collaborations for the development and formulation of innovative icy strategies, particularly in the fields of transport, climate change and spatial planning

pol-Thenia Vagionaki is a Doctoral Researcher at the Institute of Political,

Historical and International Studies, University of Lausanne She holds a B.A in international relations from the University of Sussex and an M.A

in international studies from the University of Reading Her doctoral sertation focuses on the policy learning impact of EU soft modes of gov-ernance in Greece, and other southern European countries, in the field

dis-of poverty and social exclusion Her main research interests include EU soft governance, policy learning typologies, welfare states and social pro-tection and inclusion policies among others Prior to her doctoral stud-ies, she worked in the Greek national administration for several years in the field of social policies She is spending the academic year 2017–2018

at the University of California, Berkeley, for a research stay

AbbreviAtions

ACFI Agreement on Cooperation and Facilitation of Investments

AG Aargau

AHI Affordable Housing Initiative

ALMP Active Labour Market Policies

BEPS Base Erosion and Profit Shifting

BITs Bilateral Investment Treaties

BS Basel-Stadt

CD&V Christian Democratic and Flemish Party

COREPER Committee of Permanent Representatives

CVP Christian Democratic People’s Party

DG AGRI Directorate General for Agriculture

DG ECFIN Directorate General for Economic and Financial Affairs

DG EMPL Directorate General for Employment, Social Affairs and

Inclusion

DG ENERGY Directorate General on Energy

DG MOVE Directorate General on Transport

DG TREN Directorate General for Transport and Energy

xx ABBREVIATIONS

DPTA Diverted Profits Tax Act

ENGO Environmental Non-Governmental Organization

FEANTSA European Federation of National Organizations working with

the Homeless

FOPH Federal Office of Public Health

FSSPP Flemish Sustainable Spatial Planning Plan

GMOs Genetically Modified Organisms

GR Graubünden

ISDS Investor-State Dispute Settlement

LN Learning Network of Paying Agencies and Co-ordinating

Bodies

LU Luzern

MAAL Multinational Anti-Avoidance Law

MSRs Multilateral Surveillance Reviews

NAPincl National Action Plan on social inclusion

ABBREVIATIONS xxi

OECD Organisation of Economic Cooperation and Development

Open VLD Open Flemish Liberals and Democrats

PLMP Passive Labour Market Policies

QCA Qualitative Comparative Analysis

REFIT Better Regulation and Regulatory Fitness and Performance

SFSO Swiss Federal Statistical Office

SPPM Social Protection Performance Monitor

list oF Figures

Fig 2.1 Citation network of publications relevant to policy learning 31 Fig 2.2 Co-author network of authors with three or more

Fig 2.3 Co-occurrence network of titles and abstracts

Fig 2.4 word cloud of terms pertaining to policy areas

in publication titles and abstracts relevant to policy learning 40 Fig 2.5 Co-occurrence network of publication titles

and abstracts of articles pertaining to learning in public policy 41 Fig 5.1 Learning activities of representative i with some

possible representatives j (j 2 and j 4 ), but not with

Fig 8.2 Responses from British Columbia interviewees 177 Fig 9.1 Direction(s) of policy learning via the OMC according

Fig 10.2 Press release, UK and Australia agree to collaborate

Fig 10.3 Institutional morphogenesis in Australian MNC

Fig 10.4 Institutional morphogenesis in policy transfer 237 Fig 11.1 Problem pressure and learning in a comparative perspective 257 Fig 11.2 Relationship between problem pressure and policy learning 260

xxiv LIST OF FIGURES

Fig 12.1 Plot of the sufficient solution for “BANAD” 286 Fig 12.2 Plot of the sufficient solution for “banad” 288

Fig 13.2 Growth of Brazilian outward direct investment 307 Fig 14.1 Fitted values of changes in ALMP expenditure

per unemployed over range of spatial lag variables 337

list oF tAbles

Table 2.1 Clusters in the co-occurrence network of titles

and abstracts of publications relevant to policy learning 35 Table 2.2 Theoretical concepts in the titles and abstracts

Table 5.2 Results of ERGM analysis on the dimension

Table 8.1 Low-income housing units in Quebec (QC),

Table 9.1 Key domestic features of blocked learning in Greece 198 Table 11.1 Social policy reforms and policy-oriented learning 253 Table 12.1 Analysis of sufficiency for the outcome “BANAG” 282 Table 12.2 Analysis of sufficiency for the outcome “banag” 283 Table 12.3 Analysis of sufficiency for the outcome “BANAD” 286 Table 12.4 Analysis of sufficiency for the outcome “banad” 287 Table 13.1 Brazil as a destination of foreign investment 305 Table 13.2 why not opt for an agreement with ISDS provisions?

Lessons from international and domestic experiences 309 Table 14.1 Possible combinations of institutional similarity

xxvi LIST OF TABLES

Table 14.2 ALMP expenditure in 22 OECD countries

Table 14.3 Functional form and connectivity function

Table 14.4 Diffusion of ALMP expenditure in 22 OECD

Table 14.5 Descriptive statistics for dependent and independent

list oF Appendices

CHAPTER 1

Introduction:

The Family Tree of Policy Learning

Claire A Dunlop, Claudio M Radaelli and Philipp Trein

Demand for learning is high in practically all policy domains, whether

we consider growth, the control of corruption, improvement in schools and health, or the dissemination of benchmarking and good practice

by international organizations At the same time, the supply of research findings shows that learning mechanisms are often stymied, the most obvious triggers like evidence-based policy do not work or work differ-ently than expected Or learning is not desirable, either because it is inef-ficient, for example by persevering in listening to the wrong teachers or

by implementing the wrong lesson, or by applying the right lesson to the wrong institutional context In other circumstances, learning may fail our

© The Author(s) 2018

C A Dunlop et al (eds.), Learning in Public Policy,

International Series on Public Policy,

2 C A DUNLOP ET AL.

criteria of democratic quality—such as, transparency, fairness, equality, accountability

One way to describe this state of play is to say that the constellation

of actors, incentives and norms in a policy process or a political system is not aligned with the objective of learning how to improve on public pol-icy and following the criteria of democratic theory Another is to say that bureaucracies, politicians in office, pressure groups, organized citizens and experts have objectives that are normally different from policy learn-ing, such as consensus, the control of expertise and knowledge, cultivat-ing membership, influence over the definition of a social or economic problem, and the management of implementation processes

This raises a number of questions that today define the field of icy learning First, what exactly do we mean by learning in the context

pol-of comparative public policy analysis and theories pol-of the policy process

of learning, its mechanisms, how it develops in different policy cesses, within and across countries? Third, what are triggers and hin-

what are the consequences of different types of learning for the ciency of public policy as well as for the normative criteria of the dem-ocratic theory we adopt? The first question brings us to definitional issues The second and third question are about causality—in fact, they refer to causes, mechanisms and consequences Even if our ambition is not to develop policy learning as stand-alone theory of the policy pro-cess, but rather to perfect our knowledge of learning within the estab-lished theories of the policy process, we have to assemble the building blocks of causality more systematically, in terms of micro-foundations, learning in organizations, and how group learning becomes policy learn-

fourth question is about the outcome of learning—in the literature, this is often captured by the relationship between learning and policy

outcomes, and some involve normative issues that haven’t always been prominent in the field

In this introductory chapter, we explain how the study of policy ing has evolved to the point where it is today, and show how the con-tributions to the volume provide empirical and conceptual insights that,

learn-to be entirely honest, do not answer the four questions, or at least not completely, but assist us in providing the building blocks for a research

1 INTRODUCTION: THE FAMILY TREE OF POLICY LEARNING 3

agenda that has potential to provide successful answers In doing this, we aware of the existence of some important reviews of the state of play in the field, and we refer the readers to these in order to keep our chapter

pro-vides a systematic bibliometric review of policy learning based on the most recent data by Goyal and Howlett—hence all the bibliometric data

we need to support our discussion are in there Other comprehensive

At the outset, what do we mean by learning? Obviously there isn’t a single definition in the field Indeed, the history of the political science literature on this topic suggests that learning is seen by different strands and authors as the solution to different problems, including:

• the problem of cybernetic equilibrium in a system,

• the problem of managing and reducing radical uncertainty,

• the problem of cross-national diffusion and convergence,

• the problem of knowledge utilization, and, (more recently),

• the problem of learning in different modes or types of policy processes

Thus, definitions do not come out of thin air Rather, they are linked

to approaches that capture one problem-solution association instead of another

For us, it is sufficient to begin our brief overview of the historical development of the field by keeping in mind a basic definition of learn-ing as updating of knowledge and beliefs about public policy (Dunlop

inter-action among policy actors, or personal-organizational experience, or the provision of new or different evidence It can of course also result from variable combinations of the three

the roots

Let us now briefly see how political science research on policy learning has emerged and developed we do so because an appraisal of what has already been done is the strongest foundation to design the coordinates

of an agenda like the one supporting our volume, and more generally the research agenda in the field for the near future The metaphor of

The foundations of policy learning are philosophically grounded in pragmatism and its concern for what works Pragmatic thinking marked

a fundamental historical turn away from ideological approaches to public policy If all that matters is what works, we have to be open to whatever mechanism may empirically occur in public policy, and learn how to gen-erate usable lessons from experience and evidence

Dewey’s pragmatism, however, went beyond that, because it included the seeds of a profound reflection on the normative issues we mentioned above One of his core ideas was that education, policy and the publics define a single social problem of learning In fact, Dewey’s (1927) classic

The public and its problems, reprinted in 2012, was all about re-connecting

a public distracted and un-interested in public policy problems with the essence of democracy—a normative direction that is also appropriate in the current mood of anti-politics Dewey’s utopia was to turn the Great

he even thought of mobilizing the arts to draw the attention of the

140) we can call this ‘nudging the attention’ of the citizens, to use

macro-level of analysis pragmatism implies an evolutionary learning

His lesson was not lost on Harold Lasswell: most of his intellectual effort was directed towards closing the gap between academic knowledge and society: ‘[O]ne thing Lasswell learned from the pragmatists, and Dewey in particular, was that inquiry requires community’ (Torgerson

Thus, Lasswell called for a ‘… collective cultivation of professional

of the field that in the contemporary literature, very few scholars have

1 INTRODUCTION: THE FAMILY TREE OF POLICY LEARNING 5

Mechanisms / No rmative Theor y

Theory

Po licy Proce ss

Learning in Existing Theories

of the Policy Process Learning and Democratic

Learning and Education of Publics

DEWEY

Learning as Inquiry

of the Community

LASSWELL

Learning as netic Problem

Cyber-DEUTSCH

Heuristics

SIMON

Learning as uct of Bargaining

By-Prod-LINDBLOM

Partisan Mutual Adjustment as Intelligence of Democracy

Lens on the Theore

6 C A DUNLOP ET AL.

developed a Deweyan perspective on learning in public policy (Ansell

In the Deweyan perspective, learning is the solution to the lem of aligning what works with public policy and an informed, vigilant democratic public Instead for Deutsch learning is a cybernetic prob-

prob-lem, classically exposed in his 1963 The nerves of government (celebrated

other seminal works like Political community and the North Atlantic area

insti-tutions These capabilities are indispensable to manage ‘the burden’ of the ‘traffic load of messages and signals upon the attention-giving and decision-making capabilities of the persons or organizations in con-

that learns and a system that does not is to secure these capabilities For Deutsch, organizations are webs of communication Their core function

is to generate and transmit information, to react to signals and events,

to deploy self-controlling mechanisms and to manage feedback The essence of learning is a kind of special dynamic capacity of systems to recombine resources when something changes and to manage feedback coherently Feedback is not simply finding something in the system that provides a response to an information input in the external environment

In fact, the information input ‘includes the results of its own action in the new information by which it modifies its subsequent behaviour’

of equilibrium in the system It is the capacity to pursue changing goals Thus, the kind of learning that Deutsch has in mind is similar to the zig-zagging of the rabbit in a field, when the rabbit re-adjusts its direction continuously, as new changes and opportunities arise Applied to pub-lic policy, this is a notion of learning-as-improvement (today we would say ‘instrumental learning’) that is socially progressive In fact, it allows a society to draw on learning capacity to pursue new and changing goals

An implication of this cybernetic view is that learning is not limited to making policies work better or stay efficient when the economy or demo-graphics are altered It also includes adapting and transforming policies

to follow the search for new equilibria of a zig-zagging, open society.Lindblom’s most profound intuitions did not arise out of a concern with learning but rather with the analysis of decisions, and its empiri-

pressure groups, experts and civil society organizations make policies

1 INTRODUCTION: THE FAMILY TREE OF POLICY LEARNING 7

because they have objectives of power, influence, prestige or epistemic authority in society They are not pupils in a class where the main goal

is to learn They are partisans They mutually adjust in a process of gaining, not a process of truth-seeking And yet, bargaining is also a

bargaining and mutual adjustment, an actor learns about the strategies, intentions, volitions and preferences of other actors This actor will accu-

all actors are involved in discovering the possibilities for cooperation, assuming that bargaining ends up with an agreed choice They are therefore involved in a process of exploration of what can be achieved together, what problems can be solved, and how Their goal may be power, votes or more members, but collectively they engage in a process

of knowledge generation, exchange, and utilization

Now, combine this with Herbert Simon’s proposition that actors have limited rationality and pursue ‘satisficing’ solutions instead of impossi-

opened the door to the world of heuristics, biases, framing and ing that provides a realistic account of how partisan mutual adjustment ends up in more or less functional and desirable forms of learning The Lindblom-Simon roots have paved the way for today’s interest in cog-nitive psychology and experimental social sciences A fundamental cor-ollary of Lindblom’s approach to the analysis of decisions and Simon’s bounded rationality is that partisan mutual adjustment is not just an explanation of how collective decisions are taken It is also a norma-tive model of a pluralist, open democracy The public good is not pre- determined by intellectual cogitation but results from conflict, different opinions, pluralist views… in one word, it emerges from partisan mutual adjustment The latter is the equivalent (in political systems) of the mar-ket in the economy

nudg-Heclo is famous for having identified learning as solution to the lem of puzzling, or, less metaphorically, to social problem-solving under conditions of uncertainty However, for Heclo the social dimension of learning has its own micro-foundations Heclo argued that learning

individu-als interact: learning is therefore acquired and then diffused through patterns of collective action Heclo’s approach to learning combines individual cognition as well as the features of the social and institutional environment

8 C A DUNLOP ET AL.

And yet, there isn’t anything pre-determined in this process Here, Heclo is mindful of Deutsch’s lesson He presents learning as a process taking place in a maze But, this is a special maze The walls are re- patterned all the time Individuals work in different teams or groups Each group has an idea of how to get out of the maze and gets in the way of other groups Some teams even reason that getting out of the maze may not be the best solution! Note that Heclo’s learning mecha-nisms are not random, but they are significantly shaped by social inter-action, organizations (the teams), and institutions (the structure of the

we cannot conclude the section on the roots of policy learning without mentioning the body of work developed by Carol H weiss

and hundreds of articles, she made a remarkable contribution to the fields of usable knowledge and policy evaluation But, deep down in her work runs a powerful stream of ideas about learning as solution to the problem of knowledge utilization Her findings about instrumen-tal, conceptual, enlightened and, later on, forced usages of evidence by organizations and in policy processes defined the pathway for the litera-

Thanks to her, some hard lessons dawned on the minds of political scientists: research is often not utilized by policy-makers; knowledge can be manipulated for political reasons, for example to support pre- fabricated biases; and, learning can end up in policy endarkenment, thus destroying the benign assumption that all learning is policy improve-ment and enlightenment we can call this section of the roots the knowl-edge utilization perspective In more recent times, authors like Boswell

con-tributions to learning to this organizational perspective on knowledge utilization and weiss in particular, showing the vitality of these roots

the tree gets stronger

The learning family tree grew stronger in the 1990s Bennett & Howlett,

policy learning This was the decade of learning types

Essentially, these authors break down the concept of learning by considering different types, and identifying the prima facie evidence

1 INTRODUCTION: THE FAMILY TREE OF POLICY LEARNING 9

that lead us to recognize a type (of learning) as ‘political’, ‘social’,

evi-dence that would lead us to conclude that learning is instrumental (that means supportive of policy improvement), political (that is, learn-ing about policy in a way that assists political strategies like winning elections or gaining more popularity, with policy improvement being

at best an indirect outcome), and ‘social’ (this is the broadest type of learning, where a whole society moves from one set of ideas or pol-

published one of the most-cited political science articles of all times in which he described paradigmatic change Although Hall was eminently interested in building a neo-institutional theory of change, his notion

of policy paradigm and the accompanying reference to social learning

Hall, a whole literature on ideational politics, discourse, epistemic munities, technocracy, and policy paradigms developed, crossing roads

decade and the 2000s This was not exactly a new trajectory Since the 1970s, scholars of American federalism observed the interaction between innovation in a state and adoption of the same innovation by other US

1990s debate generalized the topic of laboratory federalism in the USA in terms of more explicit theories of cross-national learning

This turn in the research agenda takes us to learning as lens to identify solutions to the research questions and practical problems arising out of cross-national diffusion The idea is attractive: instead of learning from the experience at home, which invariably includes making mistakes, a government can learn from vicarious experience, that is the experience

of other countries Analytically, the most powerful and more general way to capture this is spatial inter-dependence (Plümper and Neumayer

Convergence may or may not be the end-point of the learning process Diffusion can be limited by spatial or institutional variables, like legal tradition or family of welfare Learning from early adopters follows

an S-shaped curve, but diffusion is also triggered by conditionality,

on the 1990s typological approach, distinguishing instrumental learning from political learning, and adding considerable empirical precision to his findings about the dynamic of diffusion The difficulty with the quan-titative study of diffusion (similar to other macro-level analysis of latent concepts) is ensuring the cross-national patterns identified empirically are truly representative of learning processes rather than other processes It

is plausible to attribute patterns of adoption to learning, but it’s equally plausible to think of other variables Qualitative work comes to the res-cue by specifying the causality of learning in diffusion process (weyland

to spatial inter-dependence as mutually exclusive, we prefer to think of them as complementary

But, how do countries ‘learn’ exactly, given that they do not have cognitive capabilities of their own? And, how should they learn? The

makers should follow in their search for adaptation of lessons emerging abroad to the home destination Theoretically, Barzelay has the merit

of having introduced an explicit conceptualization of the mechanisms

of learning Indeed, his approach to extrapolation is anchored to lytical sociology, where mechanisms play an important explanatory role

Another impulse to the research agenda came from policy- makers rather than academics Over the last twenty-five years governments and international organizations have embraced the evidence-based policy agenda—at least in their manifestos and official discourse (for exam-

evi-dence-inspired policy, its biases, presuppositions and consequences has naturally emerged Often it has crossed roads with research on

But, significantly for our volume, it has drawn attention to ing mechanisms from a critical angle Indeed, the applied results (as opposed to academic findings) of the evidence-based policy vision are

learn-at best mixed—possibly because of the weak and simplistic theoretical

1 INTRODUCTION: THE FAMILY TREE OF POLICY LEARNING 11

foundations of this agenda In its most simplistic approach, dence-based policy does not take into account the mechanisms, especially triggers and hindrances, of knowledge utilization It assumes a deficit model (that is, policy-makers have a deficit of good, sound evidence) that is unrealistic Policy-makers are bombarded with information, and in any case they look at information with their own priorities, preferences, biases and political interests At the same time, this impulse towards more evidence-based choice has inspired the adoption and diffusion of specific policy instruments, such as evaluation programmes, randomized controlled trials, and regulatory impact assessment Thus, the last twen-ty-five years have also produced findings on whether evidence-based pol-icy instruments really support learning, and if do what type of learning:

symbolically, to show compliance with the beliefs and policy doctrine of

the brAnches todAy

In more recent years, the learning tree has produced new branches and research leaves and fruits Indeed, there have been no fewer than six journal special issues on policy learning since 2009—two on learning

volume exploring the nexus of policy learning and policy failure (Dunlop

more a continuity and deepening of themes that are foundational to this research programme

we make the following claims First, the new branches are more itly theoretical Second, they are less concerned with the type of learning

explic-per se (instrumental, political, social …) and more focussed on the

charac-teristics of the policy process that determine varieties or modes of ing Third, the field is healthy, but needs genuine cross-fertilization from other disciplines Hence, we point towards examples of interdisciplinary

learn-12 C A DUNLOP ET AL.

research on learning we cannot tell whether there is a rise, since atic data are not available to demonstrate the percentage of interdiscipli-nary studies across time But, we would certainly advocate that there is an objective need to go in this direction Fourth, we are more aware of the mechanisms, especially in terms of hindrances and triggers, and of what learning does or does not do to our normative criteria

system-More Theoretical

The classic way for a theoretical turn is to build explanations of ing within theories of the policy process that are explicitly formulated to

learn-explain policy change Theories of the policy process, edited by weible and

turned into more theoretical directions inside major public policy works Empirical applications of the advocacy coalition framework have demonstrated however that the framework itself provides limited expla-nations for learning processes and how the beliefs of different coalitions

narrative policy framework now talk about narrative learning (Shanahan

conceptual insights to our understanding of beliefs systems, knowledge utilization, the presentation of knowledge in policy narratives, and mech-

Moreover, recent projects have taken learning as the object of

is a good example of how we can go theoretically from individual to collective learning taking into account both cognitive and behavioural

put learning in James Coleman’s famous bath-tub They start from the classic proposition, dominant in the 1990s, that learning produces pol-icy change There may be correlation between learning and change at the macro level But, to make this an explanation, Dunlop and Radaelli

individual-to-individual relationships, and finally aggregate from micro

to macro In yet another iteration of how to go in an explicitly

per-spective of principal-agent modelling

1 INTRODUCTION: THE FAMILY TREE OF POLICY LEARNING 13

It’s the Policy Process….!

Our second claim is that the field has become more confident in necting learning to the characteristics of the policy process In the 1990s, political scientists were breaking down learning by type The types made empirical sense They were, and are, plausible But, they are inductive and, so, not anchored to theory

con-Today, and here we take the liberty of referring to the work of two

looking at the features of the policy process in which constellations of actors operate These features are derived from explanatory typologies—these typologies are built theoretically, it’s the combination of two independent variables that generates the cell in the dependent variable They are also built inter-disciplinarily, using political science as well as

learn-ing modes differ greatly dependlearn-ing on whether the policy process is

has piloted an empirical instrument to control for the presence of these different types of learning

The advantages of this approach are threefold To begin with, it opens the door to an explicit consideration of what can go wrong with learn-ing If an expert takes an epistemic attitude within a bargaining process, they will most likely become irrelevant or professional knowledge will

this approach allows us to bring together the classic intuitions of the roots, especially Lindblom’s partisan mutual adjustment, together with more recent developments on the policy process, specifically the analy-

approach is amenable to policy recommendations: essentially it tells actors that are frustrated with their attempts to produce learning to focus more

on the structural characteristics of the policy process and less on the

Cross-Fertilization

Beyond political science, there has been an exciting number of scholarly waves concerned with learning Here, there is definitively continuity with

re-discovered Herbert Simon in their ‘Systematically Pinching Ideas’

14 C A DUNLOP ET AL.

article on heuristics Today, the logics of biases and heuristics is present

in the analysis of how coalitions relate to each other (Leach and Sabatier

intuition of ‘Pinching Ideas’ has supported a whole sub-field on policy design, which is explicit about the aim of drawing on how explanations

Or, take the organizational and managerial lens Grounded in

solution to the problem of triggering mechanisms of innovation in the public sector Again, we have continuity: in the early 1980s Etheredge

rise in governmental sophistication, intelligence and analytical capacity

links learning types to organizational capacity (see also Zito and Schout

a critical voice on how public organizations self-describe themselves as learning organizations to camouflage the politics of controversial choices

or to silence criticisms of what the organization does A similar take on the manipulation of language as learning appears in Thomas Alam’s anal-

on learning as solution to the problem of manipulating organizational reputation and muting controversial political choice

Other important disciplines explicitly embraced by political tists in the field include cognitive psychology, evolutionary and experi-mental economics, and adult education Sociology and network theory and models have provided fundamental tools to understand learning In conclusion, the scene is set for interdisciplinary work Experiments have become a classic way to ascertain whether learning follows the pathways hypothesized by political scientists, or other, intriguing but less explored

draw on cognitive psychology and experimental economics to argue that there are empirical instances where change causes learning, instead

of learning causing change They show that contingency and surprise change behaviour without updating of beliefs First actors change, then, when the right feedback conditions appear, they sit down and make sense of what they have done, hence they learn afterwards