Policy Problems and Policy DesignNEW HORIZONS IN PUBLIC POLICY Series itor, Wayne Parsons, Professor of Publ Ply, Wales Governance Cnr, Carli Univers, UK ‘This series aims to explore the mejor issues facing academics and petitioners working in the fd of public policy atthe dawn ofa new millennium. seks to reflect on where public policy hasbeen, in both theovetial and pratiel terms, and to prompt debate on where its going. The series emphasizzs the need (0 understand public policy inthe context of international developments and global change. New Horizons in Public Policy publishes the latest research on the study ofthe policymaking process ad public management, and presents orginal and” trial thinking on th poiy iss and protams facing miodera aad post- modem societies. “Titles in the series inlade: ‘The Dynami of Pui Pliey Theoey and Evidence ‘Adrian Kay ies and Ttepity of Governance Perspestives Across Frontier: Bled by Leo WC. Hubert, Jeroen Mogsschalck and Carole Ls Hrkiewice ‘ble Managemeat inthe Postmodern Fra Challenges and Prospects ‘Bie by Jl Fenwick and Janes Medill ‘The Tools of Policy Formation Actors, Capecties, ens and ets Bite by Andrew. Jordan and John R, Tarnpenny Alyse and Public Pty Swoesses, Failures and Dizetions for Reform Swart Shapiro Publi Policy Transfer ‘Micro-Dynamies and Macro-Etects Eihied by Magdalina Hadjlshy, Leslt & Pal and Christopher Welker Policy Experiments, Pllures and Innovations Deyond Accession in Central and Easter Europe Eile by Ages Hatory, Andeew Cartwright and Diane Stone How Faro Nudge? Assessing Beharioual Pblic Policy Peter Jol Pigy Problems and Policy Design 2B. Gay Peters Policy Problems and Policy Design B. Guy Peters Maurice Falk Professor, University of Pittsburgh, USA NEW HORIZONS IN PUBLIC POLICY FEE) Bawand Hlgar (Cheltenham, UK + Northamptor MA, USA,OB, Guy Peters 2018 [Allright reserved, No pact of shis publication may be reproduced, stored ina resevalsyter or transmitted inary form or by any means, eetronic, smethanial or photocopying, recordig, or otherwise without the prior ponmision of the publisher. bled by ‘Edward Plgn Publishing Limited ‘The Lypiats Is Lansdown Road Chita, Gls GLS0294, UK ‘Edvard Elgar Publishing, Ino. Wiliam Prat House 9 Dovey Court ‘Norampion Messacsatts O1OEO USA ‘catalogue ecord for this book is waable rom the Bis Library Library of Congress Control Number 018935745 “This book is salable electronically in the Elgaronine| Socal and Political Science subject eolestion DOL 143379781 786131356, rie om slmentlehovine free (ECE) reeled paper containing 30% Post-Consumer Waste SHIN 9TH 1 78649 1349 (cased) ISBN 978 1 78643 135.6 (eBook) “typo by Servis Filmsoting Lid, Stockport, Chesie Pied and bound inte USA Contents Eee tsa see escent reeset EEE List of figures, tables and boxes vii Preface ix 1 The logic of policy design L 2 The problem of policy problems 35 3 Wicked, complex, or just difficult problems 60 4 Linking policy problems and policy instruments 90 5 The“new” policy design ut {6 Conctusion: toward better policy design 130 References 144 Index 165| | i | Figures, tables and boxes Seeded ESP eee FIGURES, 1.1 Design and non-design 3.1 Google Ngram showing froqueacy of the term “wicked problem” of “wicked problems” in published literature, 1967-2017 TABLES 2.1 Rochefort and Cobb's characteristics of policy problems 22 Types of policy complexity 2.3 ‘Types of problem structures 24 Stone's typology af policy problems based on causa stories 25. Structured and unstructured problems 3.1 Wicked problem characteristies 3.2. Important dimensions of wicked and complex problems 33. Evaluation of climate change as a wicked probiem () 511 “Old” and “new” policy design BOXES 1.1 Bobrow’s principles of policy design 21. Problem characteristics related to agenda-setting 3.1. Characteristics of wieked and super-wicked problems 32. Characteristics of complex problem solving 313 Dichotomies of dificult and easier problems 441. Examples of policy instruments 4.2. Eight characteristics of policy instruments 423 Linking instruments and problems 44. Characteristics of instruments for policy design 5.1 Hard and solt systems analy 20 8 56 65 8 0 108 108 109 13Preface eo ‘have been interested inthe concept of policy design for decades, beginning ‘with ray work with Stephen Linder at Tulane in the 1980s, That said like Jnterest in this concept in poliey studies, my interest has waxed and waned. "This volume represents a return to tis theme, and an attempy to fink the work in policy design that we did previousy with the more contemporary Iiterature on public policy. Further, there has been a more general retarn to jntorest in design, and che development of innovative approaches that also should be linked to the older design literature, “This volume is an attempt to dest with questions of policy design in a ystematicmianner; but inevitably some things that my colleagues think are important will be left out, and some things they may think ave trivial are included. The literature on public poliey is now so extensive and so rich that some seleetion is necessary. [have endeavored to make the arguments ‘about design as clear as possible, and that sometimes involves leaving tt interesting material that could contribute to the discussion of policy design. Tt is ato important to point out that policy design is more than just ‘old wine in new bottles, although there are certainly some themes that ‘Would be common to almost any analysis cf pubic policy. The impos tauve Of emphasizing design is that rather than formulating policies through path dependence or minimal consideration design emphasizes the need for careful cogitation about the elements of policy, and somo attempt to ‘make workable linkages among those elements, This is not an easy task to pevforn well; but it can, we assume, produee better polis after that investment of effort. "There are many people who deserve thenks for their direct and indirect contributions to this book. As noted above, Steve Linder was important jn developing my earlier conceptions of policy design. My colleagues land frionds Jon Pierre and Bill Dunn have contributed in more ways than they may realize to this enterprise, And my colleagues in the International Public Policy Association — perhaps especally Mike Howiett, M. Ramesh, and Philippe Zittoun — have also help me develop some of these thoughts Thanks also go to Sheryn, who endures the absences at confer fences and workshops that have contributed to the development of this tsx Poley problems and potieydeston ‘project. And, last but not least, Alex Petifer and his colleagues at Edward Elgar Publishing have been patient, profesional and supportive. But, of ‘course, errors and misinterpretations are all mine, B. Guy Peters Pitlsbargh r 1. The logic of policy design ‘Scholars andthe individuals involved in making public policy use a variety of words to describe how they actually arrive at the content of those policies. Perkaps the most commonly used word is “formulation” (see Jordan and Turmpenny, 2015), but words such as creation, innovation, and velopment are also used to describe the process of finding some Form of intorvention to confront a policy problem. The hope is always that the policy thats formulated or created will beable to “solve” the problem, and ‘that government (and citizens) ean go on to eope with the next problem that arises When Herbert Simon (1995, 11) wrote that “everyone designs who devises courses of action aimed at changing existing situations into proferred ones, the definition was somewhat generic but was definitely speaking to policy desig. ‘Although thinking about policy design hes become more common in policy studies, it should be considered as 2 significant alternative to more casual ways of thinking about policy formulation, As Jan Tinbergen (1958, 3), a Nobal Izureato in economics argued, design (in particular are members of professions with strong coramitments to particular perspectives on society, may also impose ideologies and sclctively use evidence when designing policy. This ‘commitment on the part of professional to particular designs appears in the instruments literature, For example. When asked how they would address range of policy problems many economists would always sclect, an economic instrument, while lawyers would opt for «legal, regulatory instrument (Linder and Peter, 1989) These choices clearly reflected their education and training, but did not demoastrate significant willingness to ‘move beyond thei profesional comfort zone. Palltial De Another important style of design can be labeled a “political design”. This style of design reflects contention between allenative designs and the outcome of a political process that chooses among them, or mor’ likely produces some compromis> among them. The political design often reflects the compromise be:ween two or more conflicting ideological designs as described above. However, rather than having & hegemonic party or actor there are contending actors, ack presenting « desiga, aod some means must be undertaken to find an outcome for that conflict. ‘While compromise is often seen as a polities! virtue, for policy design it may be less so. Fritz Scharpf (1988) has argued that when policy choices are being made when there are veto players the tendency is to crcate the “joint decision trap” and to produce decisions which are the lowest common denominetor. While that ray be the extreme version, any compromise may produce policy designs ‘hat ace less than optimal. The6 Polie problems an policy design implementation literature, for example, points to number of instances in ‘which compromises in formulation in the legislature produced “designs” tbat were virtually impossible to impleme in some circumstances Scharpl's analysis may be too pessimistic. This assuraption about poor-quality designs resulting from compromise among, attemative policy proposals is based on an assumption that the parties fnvolved ace playing this “game” only once, and hence will attempt to gain everything they wantin this one decision. On the other hand, if we femember that poliosmaking is repetitive and almost continvous then there are multiple opportunities to rectify any failures during the initial round of policymaking (Peters, 1997), and deals ean be made that trade prosent losses for future gins. “The idea of politica feasibility is often important in the political design of « policy or program. On the one hand, this is @ real constraint. Ifthe poligy being proposed ~ no matter how technically sophisticated it may bo-is mot acceptable to policymakers and/or to the public then itis not @ pod proposal, On the other han, i is easy forthe estes of a program {o argue that itis not politically feasible, and thercby deter any serious iiscusion of the program, The task of the designer, therefore, beeomes determining the real feasibility of a policy, making the right arguments if itis feasible, and redesigning if itis not really feasible Desig by Transfer Tntheliterature on public policy as well as in the real world of policymake ing, there isa significant level of dlifusion of policy idcas and designs ‘Thoteaze over « hundred countries, and thousands of sub-national gov- cmments, involved in the process of making publie policies; and there is therefore significant body of experience outside of any one goverament, ‘or any one designer, that cin be used to shape policies, The academic lit erature on policy diffusion is extensive (see Hadisky ta, 2017) and there § pesaps en even larger body of experience in government on borrowing from other political systems, Diffusion of policy ideas has bocome more prominent with increased interest in “evidence-based policymaking”. Indeed, the idea of using idence systematically to inform policy has become a fad within govern- Ghonts and academics have followed with a large number of studies of this phenomenon.l? These studies have included a number of highly Sfepiealasessmeants ofthe utility of this concept, based on the positivist, sumptions in much ofthis iterature 26 wel as on the general problem of translating information and practice from one seting to another (see Prwson, 2006), That skepticism has been somewhat more intense about the The logie of policy design 0 excessive touting of the concept than about the intrinsic uility of utilizing ovidenoe from other settings to make better policies, ‘While using the experience of other policymaking systems to formulate 1 policy desiga makes a great deal of sense, itis not as easy as it may appear. Although a policy design may perform extremely well in one setting, t may not readily ransfer (o another. This i especially tre if we consider design in a more open and social perspective than in the more ‘conventions! enginosring or architectural manner (see sbove, p.6). Even if we do adopt the more mechanical conception of design, the successful diffusion of # design is far from guaranteed. As I will emphasize below, effective policy design involves understanding the context within which the policy wil function, as wel as the context from which iis being extracted “The success of any policy, whether the subject of conscious design or not, then depends heavily upon the extent to which it matches the internal politcal and social patterns of the host country, or host city even within the same country. Far too often potioy transfers ignore the social foundations of designs and assume thal a program can be picked up and moved with litle or no thought, The numerous problems thatthe countries of Central and aster Europe encountered in translating the ‘acquis ofthe European Union (EU) into their own settings provides ample tevidence ofthis point (Bantory et al, 2017), but these are far from isolated examples, Design by Innovation “The type of design that most individuals, whether academic or poticymak- crs or ordinary citizens, would consider real desiga is design by innovation. “Although Thave argued already that policymaking is remaking existing policies, there are some genuine innovations in policy. This is in many ‘ays dificult claim to make. Even polices that may appear innovative tnd path-breaking may be simply taking a model or an idea from one domain an spplying it to another. For example the 1990 Americans with Disabilities Act (ADA) was taking the well-established template of civil rights legislation and extending it to another class of people who faced discrimination and socal barriers (Burgdort, 1991) While we may arguc that in practic design by real innovation is unlikely, that has not stopped numerous governments and students of design sore generally from creating innovation lboratorcs that are meant to gencrate new policy designs and new ways of thinking about public policy (Windrum and Koch, 2008), These innovation policies have been directed primarily a economic polices, and increasingly at sustainable econotnie ‘development, but the ida of festering innovation and new policy desiens8 Polieyproblems and poiey design has been tried ueross the expanse of government activities (see Kattel, 2004), “The attempts at innovative desiga in the public sector have two features that are important for thinking about the design process more generally ‘The first is that many of the designs that emerge are more about the Jelivery of public services then they are about the actual content of those services. This is not necessarily a major problem, and the delivery of services implementation —i crucial for che success of the programs. That said, however, policy designs are usually conceived as mechanisms that will go beyond delivery systers to include formulation of more innovative approaches to the problem being considered. Designing the implementa- tion structure i important, bt itis not the sole focus of design ‘The second characteristic of these efforts at innovative design is that they tend to be conceived a8 participatory and collaborative exercises (Gee Bason, 2010, 2016), Rather than having innovations be the product ‘of cogitation on the part of a few experts, policy innovation fends to be conceptualized as being more likely and more effective if it involves mul- tiple actors, especially stakeh>Iders in the policy area. While some of this ‘commitment to collaboration can be seen as an ideological commitment t0 public participatiog, iti also usified by the assumption that it can bring ‘a wider range of ideas for soktion ofthe policy problem into the process, “The process of innovation, or design, may require more time ~ but the assumption is that superior designs will be the result ‘Accidental and Experimental Design ‘The notion of an accidental design may appear to be an oxymoron, but governments may simply happen on policy designs, and some of those designs can be effective. This perspective on designs is especially compat- ‘ble with the third wave of participatory thinking about design, meaning that rather than necessarily being imposed from the top down, many designs are more bottom up. Even in more traditional conceptions of design some possibilities for thinking about policy from the bottom up were accepted and oven advosated (Bhnore, 1985). The bottom-up features of design can be driven by publi participation, or they may be driven by lower-level public servants. ‘Accidental designs for policy may appear most commonly in times of wat, or economic or social esis. The crisis management literature points to the need for governments to do something al these times, even if itis ‘wrong (Boin etal, 2017). Fer example, studies of Franklin D. Roosevelt heaving to cope with the problems of the Great Depression and then World ‘War Il point to his unending search for ways to manage problems. Some he The logic of policy design » happened upon ~ for example the Works Progress Administration ~ were very suocessfi, while others were not (Leuehtenberg, 1963), but that was to some extent irrelevant during the cries. What was important was 10 Keep trying to inncwate and find something that worked, ‘While some designs may emerge through simple good fortune, scholars and governments also need to consider the possibilty of more systematic experimental design efforts, At a certain level almost all policymaking is experimental (Campbell, 1969; Sabe] anc Zeitin, 2010), The social and teonomie processes into Which governments are interjecting their policies are complicated, and predicting that a cttain intervention will produce a certain outcome is at best difficult, This is true especially tor complex and ““vicked” problems, but may also be true for many more familiar targets of public sector action. ‘There appear to be two major challenges for designing policy through experimentation. The first is the political shallenge of making @ notion of ‘experimentation accaptable to lawmakers and citizens. While there have been sone major policy experiments, it is often difficult to get a program that is deemed simply an experiment adopted and funded. Second, if a ‘program is openly deemed to be an experiment then will citizens or other targets of the intervention take it seriously and behave as they would if it were a “real” policy? These challenges iply that governments may have to continue to experiment without informing the public, or perhaps even themselves, of the nature of the intervention Can There Be “Non-Design”? ‘While I consider policy design an interesting concept, if itis to be useful jn a scientific context then there mst be Snon-design”; that i, chete aust ‘be policies that emerge from a process that is not design (see Howlett and Mauikherjee, 2014), While, as stated above, all policies appear to involve ‘some level of design, there are atthe same time degrees to which design is involved, Te notion of “fall design” meationed above implies that there isconscious cogitation and the considerat on of all four ofthe elements of| utlined below. This would be analogovs to design by innovation rmeationed above, and the antithesis of design by accident or perhaps esiga by experiment The other types of design mentioned earlier would contain varyi degrees of “designedness”, if there can be such a term (see Figure 1.1) ‘With full design at one end, other types of design might be lower on some of the elements involved in design. For example design by borrowing ‘might be simply thal, and involve zelativey little cogitation about how the policy actually functions, and especially little about how it might work2» elie problems and poley design Full Thought sign eplcement Programmed replacement ‘Soueer Base on Hole and Make @014, Figwe 1 Design cod non-design in the new context, The basic question, therefore isthe extent to which ‘designs emerge from conscious attempis to find a solution, considering ‘a range of options, to processes haved on routine, path dependence, and simple borrowing of interventions. ‘And itis important to remember that for public policy non-decisions are «decisions. This point has been highlighted in the agenda-setting literature, jn which the ability 1o prevent decisions, or even the consideration of decisions is perhaps the most powerful political ploy avilable (Bachrach ‘and Baratz, 1962; Lukes, 2005). Thus, non-design in which the demands for policy change are kept off the agenda and the existing policies (or lack thereof are allowed to persist can also be considered a form of non-desin. "Thus, while designing policy is active, important policies persist hecause of the absence of that activity, A MODEL OF POLICY DESIGN ‘The above discussion has pointed to some basic features of policy design, but itis necessary to pat some flesh on those bones and preseat a basi¢ model of desiga. This model of “full design’ isan atterapt to describe the Fundamental components of a working design for public policy, rather than, the process through which that design is created, Thus, we ae distinguishing design from designing for the purposes of presenting this mode, recognizing that process is far from unimportant when conceptualizing policy design Indeed, as Iwill point out in subsequent chapters, the nature of the process ‘pursued will influence the type and quality of the desigas finally adopted. ‘Tho model presented here builds on the ideas of policy design that Stephon Linder and I published some 40 years ago (Linder and Peters, 1984, 1989). Although now seemingly somewhat out of date, the basic ideas for design presented there are still relevant. There have been a number of critiques of this approach to design, including some T have penned myself (see Peters and Rava, 2017}, but the approach ~ even if seemingly highly technecratic and apolitical remains a very useful place from which to begin a consideration of policy design, The logic of policy design a “The approach to design presented here argues that any working poliey design requires at least four elements: an understanding of eausation; an ‘understanding of instruments available to adress the problem; and aa explicit understanding of the nalues by which the outcomes of the policy should be evaluated; and, Finally, 2 plan for incervention. Further, the design must understand the manner in which these four elements interact and can be used for attempting to correct ongoing social and economic ‘processes, While this notion of design may appear extremely simple, in practice itis not, and requires very significant levels of both substantive policy knowledge and political understanding if i isto be successful. Causation An understanding ofthe cause ofa problem isthe fast stp in Formulating 4 design, although there may be an even earlier stage of recognizing that there i problem, oat least a problem that can be addressed eletiely through public policies, Peior to the acceptance of Keynesian economics in the 1930s the economy was problem, but it was bulived tht it was only slightly more controllable than the weather. And in somo instances issues such as spousal abuse or child abuse may be ignored entirely simply because they are somehow assumed to be “normal” (Nelson, 1984). ‘Once policymakers recognize the existence of a problem, they mast have some idea about the causes of the problem if they want to formulate a response inthe form of a policy design, The trouble is that many problems Fave multiple and complex causes, and thoreore setting on a eause in order to construct a design is a dificult analytic and politcal problem Take a major policy iswe such as poverty and socal exclusion. There may be as many causes forthe problem 2s there are snalysts, and those prosumed causes are often ideologically and politically charged, Analysts on the politcal right will argue that individual inadequacies are the root cause of the problem, while on the political lft the finger will point to structaral problems in the economy and the feilure of governments to ‘break the intergenetational transmission of policy with effective policies > ‘Who is ight? Given thatthe causes of policy problems are often unknown, ora east. disputed, policymakers are often left just treating the symptoms rather than the root causes, Just as a physician might trete fever of unknown cause with aspirin and uid, so too might the policymaker simply attempt to ameliorate the symptoms. Tu the case of poverty, for example, giving people money without attempiing to make them more capable of earning an income and becoming more inchude in society may bean effective and Inumane treatment without addessing the underying causation2 Poly problems and policy design Instramentation The selection of instrumests is the second component of poliey design. Governments havea large (901 chest at their disposal for addressing policy problems, These tools have been categorized in a number of different ‘ways (Hood, 1983; Salamon, 2002); but while those categories are useful analytically, they may not do much to solve the dffiulties faced by policy ‘designers. The dilemma for the designer is that there ae so many options, ‘each with its own political economy, that the choice of instrument may appear overwhelming, and designers will revert to the familiar or the ‘conventional Although the classficaion of policy instruments may not provide ‘lear enswers forthe wouk-be policy designer, attempting to understand the underlying mechanisms st work in those instruments may provide at east some clues about the manner in which they will produce action, For example, Hood's classification of instruments is based on resources for government ~ authority, nodality, treasure, and organizations, These resources may be seen us analogous to the core mecknisms associated with analyte sociology (Helstrim and Swedberg, 1998; Real-Dato, 2009). Understanding these underlying mechanisms contained within policy instruments in turn gives some idea about under what circumstances 19 select particular instruments for a policy design. ‘Much of the thinking about policy instruments and design has been an attempt to find an algorithn that can map a set of policy instruments onto fan array of problems and “ind optimal solutions for the problem, While ‘the optimism that characterized uch uf the cavlice thinking about policy design has waned, there is tila lingering hope that designing can provide clear guidance forthe individuals and organizations responsible for making, policy There arc more tools or policy formulstion than in the past Jordan ‘and Turnpenay, 2015), and experience and scholarly research are providing ‘more guidance, At the same time, however, the number and complexity of | policy problems appear to have increased (sce Chapter 3) along with the range of possible tools, mating design even more challenging. Evaluation ‘The third clement of this model of design is evaluation. This does not necessarily mean the appliation of evaluation research alter the program is implemented, although taat could certainly be involved. Rates, this is ‘a more geueral demand thst policy designers be explicit about the values \which they seek to achieve through the policy being designed. In other words, what outcome is hoped for and what outcome would be minimally ‘The log of poloy desien 23 acceptable? This clement of the design is obviously related (othe goals of the progeam as well asthe political and social values of the designer Just as T identified the possibility of multiple causes for an observed policy problem, so too are there multiple ss of values that can be applied (ee Hausman et al, 2015), Some may be simple empirical tests ~ there is 4 program goal and it is achieved or not. But rarely are the outcomes of policymaking as simple, and the gosls wit often be only partly achieved, ‘And even ifthe goals are reached there may beside effects and unintended consequenees that also need to be assessed and weighed against the more positive outcomes on the main objectives (see Vedung, 2006, 2013). Intervention ‘The fourth clement of this design approsch is a model of intervention, If government, possibly also receiving assistance from its partners, from society and the market, isto intorvene in socoty it will nea to do more than just select instruments. Making policy function effectively means the development of a program infrastructure that utilizes the available instru ‘ments and ensures that they are implemented. Thus, good public policy roquires good public administration. Policies do not work on their own; they must be made to work, and thereforesome attention must be paid to signing structures for implementation along with designing policies per se (Peters, 2014), ‘To some extent the problems of public policy can be seen mirrored in public administration. In particular, whik we focus on complex policy probleum ia this bok, theie are equally complex implementation steuc- tures designed to deliver the policies, Some years ago Porter and Hiern (1981) discussed that emerging complexity, and the involvement of social ‘actors in implementation has only become more extensive (se also Peters, 2015), This complexity in the intervention model, while leveraging private resouress, also creates needs for policymakers to consider monitoring and ‘compliance as major pars oftheir intervection strategy (see Donalue and ‘Zeckhauser, 2011). Policy Paradigms The above aspects of policy design can be seen as combining into policy paradigns, meaning the “framework of ideas and standards that specify... the gous of policy and the kindof instruments that can be used toattain them” (all, 1993, 279)" These raradigms provide general guid ance tothe would-be designer ofa policy, telling him or her jut what they should be attempting to do and how to goabout pursuing the appropriatem Policy problems an poly desion policy goals Paradigms aze therefore useful shortcuts for the would-be designer, providing guidance on how to make a decision, There are a number of existing paradigms in the social sciences, Some paradigms can be derived from major ideological perspectives such as Marxisin, utilitarianism, of neo-liberalism. Other paradigms may be derived from academic disciplines such as economics or philosophy. Stil others may be specific 10 4 policy area such as energy or defense. No matter what the source of the paradigm, they can shape policies - and that can also constitute a barcier to policy change. Policy design therefore ‘must understand the existence of these paradigms and develop means for contesting dominnt patterns of thought that may stifle needed change, ‘A policy paradigm would provide some answers for the various ele- iments of design mentioned above. A policy paradigm first will provide fa means of understanding the problem ~ the same economic issue will be understood differently by a Marxist or « neoctiberal. That definition ‘of the problem will thea evoke instruments that can be used in response, ‘and again paradigms may rely upon different types of instruments to do the work of changing the perceived problem. And pechaps most clearly the paradigms will evaluate outcomes very differently based on the values ‘embedded in the paradigms. These policy ideas may have less to say about intervention aud institutional design, although that part of the model may be derived from the desir to utilize diferent sorts of instruments, Paradigms can also pose problems for the designer, Although there is ‘very convenient template for how to make the policy, there is also the danger that a designer would continue to repeat old mistakes and not lool, for more ereative solutions to the policy problem, Much as paradigms in scienoe enforee "normal science” (Kuhn, 1962),so too do policy paradigms constrain the poteatial creativity of actors inthe policy process. As Iwi bbc arguing in a later chapter, policy design is now moving away froma the essentially deductive approach dependent upon policy paradigms toward ‘a more induetive perspective. By opening up the design process to a wider range of perspectives the task of designing bevomes mote dificult, but the rewards of doing so may be substantial CRITIQUES OF THE POLICY DESIGN LITERATURE ‘The discussion up to this point has been generally positive about the concept of policy design and the capacity to adress, if not solve, problems in the economy and society. The optimism about design has some justifica- tion, at east as nn academic enterprise, but we should also express some skepticism about the design literature as it has developed. This skepticism {The ogi of poiey design 25 will be expressed primarily in the direction of the academic debates on policy design, but many of the same points have substantial relevance for the practice of policy desiga. And, as important as the academic debates ray befor those of us involved in them, the effects of policy choices in the real world are of much greater significance. ‘Narrow Focus ‘The first point that shouldbe made about this design literature is that most policy design thinking has been focused on extremely narrow topics. That is, it has tended to focus on an individual policy problem, often defined in a very constrained manner, with the intention of producing a rather narrow program, Thus, rather than thinking about alternatives for urban transport, a good deal of the design literature focusos on things such as means 10 expedite automobile traffic movement.!® While those problems are not trivial, especially for individuals eaught in traffic jars, thinking about policy alternatives could benefit substantially from & broader con- ceptualization ofthe problems. And failure to think more comprehensively about policy may only create now problems in other areas — mote highways rmeans lower air quality and pethaps more people moving to the suburbs and gutting the inner ets. “This natrow focus for most policy designs requires greater attention to policy integration and coordination (Peters, 2015; Jordan and Lenschow, 2010). While coordination is often portrayed as an administrative and institutional design issue, some of the problems inherent inthe specaliza- tion of govecament organizations can be rectified in advanes by careful policy design. Such a design would necessarily involve instruments that ean link polices and organizations and produce raore coherent policies across vatious policy areas (Jordan et al, 2005). Devising those policies is difficult, but wtimately neoessary if policy problems are indeed to be solved. Policy design and poliey integration are therefore two eemeats of policy analysis that should be brought more closely together. For policy design to be effective at more than the simple program level ths logic of policy integration needs to be brought into play. This awareness of the need {or integration will mean that the elements of the policy being designed ‘nust bo abl to work wll with other relevant policies in the domain. AS ‘one simple example: when attempting to make health care more ecient by moving people out of hospitals more quickly, it is important to be aware of the limitations of the social cae system that will have to necept patients who will not be capable of caring for themselves (Bamberger, 2016)%6 Pot problems and policy design Ronetive Style of Desig me [A second issue, related to che first, isthe tendency in much of the design literature to think about the problems primarily ina reactive and retrospec- tive manner. That is the sssamption is that problems emerge from s and economic processes, ad that after the problems become sufficiently vexing to citizens and/or actors in the public sector something will be done about them, The assumption appears tobe that designing toa great extent involves governments siting around waiting for something, adverse to ‘happen, and js related to the emphasis on agenda-setting in the alysis of| policy This characterization of the design literature is somewhat too harsh perhaps; but governments are generally responding to prior evens, and often long strings of market and policy failures, rather than anticipating policy needs “Time and the anticipation of events are always major problems in policy design and analysis Politically, there is not very much intrest in the future ~ those Future citizens are not voting now. And many of the analytic techniques used for policy analysis, such as CBA, tend to discount the futur. This heavy discounting of the future is characteristic of the “super- wicked" problems that governments are beginning to confront, albeit ‘with little success (see Chapter 3). The challenge then isto find means of {aking time and the long tm into account in societies that are generally interested in almost immediate gratification, Although most policymak- ‘ers work with optimistic essumptions, the effectiveness of polices often decays overtime ~ and that must be considered in design (Callender and ‘Marin, 2016). ‘Narrow Temporal Considerations Flowing from tht rtrossectve nature of much of policy design, a good deal ofthe literature on policy design is dfined rather narrowly in tem poral terms (se iacobs, 201). That is, there are pokey problems now that ‘appear difinlt; but design should also be considering the problems that vill arse and which may be anticipated with some degree of certainty” “There ae means of integrating foresight, and especially static foresight, into policymaking. Many of these devices have been developed for the tnsinss community, but they aso ean be applied to the public sector (Kuose, 2012). In this manne of thinking about policy design the ieipera tives of design begin to sug with those of planning. Given that planning has been generally evalued in contemporary goveraments, is important to identity ways through which some sense of preparing programs forthe furure can be reintroduced int the public seta. ‘The log of policy desien n ‘As Iwill point out in following chapters, the most important problems {acing contemporary governments involv: long-time dimensions, and in some cases represent the consequences of a failure to design programs with effective foresight. For example, public pensions ~ both those pro- vided through social seeuity and those fer public employees ~ have been dasigned with litle attention to the changing demographics of the labor force, and of society more generally (Tiasalo, 2017). Therefore, defined benefit programs are seriously underfunded, and defined contribution programs that have been in essence pay zs you go are beginning to lack adequate coatributions to make them viable (Beermann, 2013) Uncertainty and Risk “The time dimension in policymaking points to another prevalent problem — the inadoquate attention to risk and uncotanty. Al polices involve some vncertanty, but this is usually implicit in policy design rather than explicit (but sce Dror, 1986; Mood, 202). Uncertainty and risk have been integrated into much of the regulatory poliy literature, especaly when attempting to understand what the worst-case scenarios for some controllable, or urcontrllbl, events might be, Oue of the best examples ‘nas been in the regulation of genetically modified organisms (GMOs) in the food chain, with European regulators assuming that if there are indeed problems the risks to human health of aocepting GMO crops would be unacceptable This precautionary principe is in marked contrast tothe general acceptance ofthese technologies by American regulators. ‘Tir tendeney wot to deal explicly with uncertainty in policy design is becoming # more important issue as governments increasingly confront “wicked” or illstructured” problems (Ge Peers, 2017; Coyne, 2005) Although dificult enough, most of the problems that governments have boon facing are relatively “tame” and operate in « more or les linear fashion, Emerging problems, notably climate change, da not operate in such a predictable mannes, but rather hav tipping points and non-linear relationships tht make their dynemics less predictable, Consequently, the designs for poly must also buildin adaptive capacities, and perhaps also tbe embedded in more adaptable institutions (see below). “The Tabula Rast ‘Another of the problems with much of the design literature, and with ‘design as practiced by governments, is that it tends fo assume a tabula rasa. Although we noted the importance of redesign above, the design ‘discussion is often addressed as a problem that has been identities, with8 Policy probleme and poley design the object of the exercise to develop a response (or the response) to this perceived problem. The difficulty I find in this discussion is chat most policymaking is actually replacing existing policies, or at least intervening in an already crowded policy space (Hogwood and Peters, 1983; Carter, 2012; Schaffrin eta, 2014) As discussed above concerning the narrowness of thinking in policy design, the failuee to integrate a new policy with the existing array of policies can only complicate the implementation of a “gow” program and reduce its probability of success. ‘The difficulties of integrating new approaches or programs with existing ‘ones has been seen any number of times in reforms of public manage ‘ment (See Pollitt and Bouckaert, 2017). Whether they were the ideas of ‘New Public Management (NPM) or other types of reform (Rouban, 2008), adininistrative reform involved altempling to interject new styles ‘of condueting publie business into the existing institutional structures ‘and administrative processes. The failure to understand the layering and. displacement (Mahoney and Thelen, 2010) in administrative policy change often leads to failure of what might otherwise be workable programs, Existing underneath each new layer of policy reform are numerous layers from previous reform efforts, and unless they can be tied together and integrated success becomes significantly less likely. Context AAs well as not having a strong temporal dimension, policy design is often done without regard to context, This has been especially true for poliey transfer und borrowing (but see Hadjisky et al, 2017), but can also be ‘true for policies being designed within a particular context. If designers are excessively tochnocratic and/or ideological they may assume that their {evorite policies will work, regardless of the institutional or social context. ‘Those assumptions are often the recipes for policy failure; but, that suid, understanding the relevant dimensions of context for policy including the impacts of existing programs) is dificult. These failures to understand ‘context are perhaps especially relevant for development programs imposed by more effluent countries on countries in Aftiea and Asia But it may not be enough simply to say that we need to understand con- text, Which aspects ofthe context are relevant for design, and which can be ignored? And how ean conceptualized context be linked to the outcomes of the policy process and in the ultimate impacts of a poliey once it is implemented? Falleti ara! Lynch (2008) point to the means for integrating, ‘coatext into comparative politcal analysis, and the same should be done more dtccly in the study of public policy and policy design. In particular, ‘more attention must bo paid to integrating insights from comparative The logic of policy design » political scienoe and sociology into the study of public poliey (see Peters, ortheoming) ‘There are a number of coutextual features that we would expect co affect the performance of policy designs (Rose, 2005). At the extreme are the vast economic differences that exist across countries, or that exist across different regions ofa single country. In addition, politcal ideologies fand traditions will affect the choices of policies, and pechaps especially the selection of policy instruments. And many political systems also are deeply divided and sill depend upon customary forms of authority and ‘policymaking rather than the more rational-legal forms upon which most public policy is base, Instrumentalist Much of the design iterature is also focused heavily on instruments ee for ‘example, Howlett otal, 2015). While Ihave identified instruments as one clement of design, and will spend sometime discussing poiy instruments in subsequent chapters, design isa more comprehensive proces Designing involves significantly more than just selecting instruments. Especially when we begin to think about alternatives to conventional forms of design (Chapler 5), the usual panoply of instruments cannot address adequately the range of wicked, complex, and critical policy problems. Further, not all of policy design is simply about selecting the best instrument to match « particular task, with that task typically being narrowiy defined it also involves thinking about context and about the values assoeated with the design. ‘The danges of focusing so heavily on instruments is that it bocomes, as Simon once described for rouch policymaking analysis, a “cookbook” The policymaker has a problem, and then that policymaker attaches 4 solution to it and tho problem is solved ~ QED. As T have beyun to argue above, policy design should perhaps not be considered in such 4 meehanical manner, and a more compete analysis of problems and solutions and values may be required. And those elements must be fitted into institutional structures that may not be as favorable to the new policy idea as the designers night lke. Design then becomes working within the existing paradigms for policy while simultaneously looking for the means to overthrow those paradigms Although (am eritciing much of the literature on policy design fr its emphasis on instrument, somewhat paradoxically design will avays involve instruments “Instrument” isa generic term for means of inter ening in a problem. What Tam criticizing is any simplistic linkage of problem and a solution, especislly esting a single instrument to adsest20 Polley problems and poliey design ‘2 complex problem. The danger in instrumentals is that a too quick Closure on selecting an instrament will lead the designer to ignore other important elements of the design process, and perhaps the full range of available instruments ‘Summary In short, the existing policy design literature has numerous strengths, and the general idea of thinking about policymaking in design terms is important but there ate also weaknesses that limit its capacity to deal ‘sith importent public problems. For the more straightforward problems Facing governments the narow and linear conceptions of design may be appropriate; but for the homing “Continuous Critical Problems” or CoPs (Grbektian, 1970) tht narrow and largely technocratic approach appears inadequate and, in William Dunn's (1988) cherecterization of some policymaking, may be solving the wrong problom (see also Mitroff dnd Featheringham, 1974). Pethaps most importantly, given the array of problems now facing the pubic sector, the policy problems may be defined in the linear and constrained manwer while they are actually complex and cross-cutting LINKING DESIGN WITH INSTITUTIONS. Our discussion of policy design has been, atleast to this point, somewhat abstract and not as closely Lukel to the political seaites of governance ‘as might be desirable, In panticula, I have been discussing policy design ‘without placing the design process into any institutional setting. That sort of analysis can only go so far given that policymaking is inherently institutional, and the nature of public sector institutions, and theit alles in the private sector, will influence how designs are actually made and implemented. If we want to sudy policy design, we need to think about the designers, both individual an1 institutional For the conventional model of poliey design sketched out above the institutional linkages appear vitualy identical to those that are used to dleseribe the standard policy process model. For example, the definitions of| causation and the identification of policy problems are very similar to the idea of agenca-setting, and can be described using the similar institutional ‘actors involved (sec Zahatiadis, 2016), These institutions are very good at dealing with policy in w limited time frame, and generally with relatively sinall departures from the status quo. Although most models of the policy process tend to fink policy : The logic of poley deen 3 formulation (and therefore policy design) with the legislature, it is crucial to understand that design is a game that any number of people can play. Tn particular, bureaucracies arc osatral players in design. The upper echelons of the bureaucracy axe usually assumed to have a significant role in making policy recommendations, but the influence of bureaucratic actors on policy may be much more pervasive For example, Edward Page (2007, 2010) has demonstrated that a good deal of the inigal detailed design of policy is actually done at middle levels of the publie bureaucracy. ‘Alco, although public bureaucracies can have sirong commitments (o the ‘existing policy paradigms, they can also be the source of ideas for policy ‘change (Berman, 2013). Is also important to link the institutions associated with evaluation and assessment of policy performance with design. These institutions typically influence policy at theend of the cycle, but their work constitutes the foundation forthe next sound of design (Donaldson and Lipsey, 2006), Unfortunately for the quality of designs {aat are emenging from govern- ments, however, these evslustion institutions have been devalued in favor of shorter-term “pesformance managermen’” that looks atthe achievement ‘of quantifiable targets more than where those targets came from and their utility for society ‘The ubove esitique of conventional forns of design can be defended easily on intellectual grounds, but they may be dificult to actually imple iment through the conventional institutioas of the public sector, Most governments encounter difficulties in making policies with a long-time horizon (see Schneider, 1991; Jacobs, 2011). This short-time horizon is present both booauco of the uncertainty of the policies and the uncsriainty Of polities (Warwick, 2000). The idea of meta-design as developed hore requires such an extended time frame for thinking about policy, and may also require some continued commitinenis by government in order 10 maintain the policy But gaining that type of commitment is exceptionally dificult in a democratic regime when each new goverament will want (© address its own priorities. While all governments do encounter some difficulties with longer time frames, Jeremy Richardson's notion of rolicy style (1982) argues that some goveraments are better at dealing wit time than others, One of the defining features of a policy style is its anticipatory versus reactive nature. Designing in areas of substantial and unpredictable change will generally be characterized by & more reactive style although most governments might tend toward that end of the dimension anyway. In the best of al ‘worlds a government would seck to anticipate such changes because of their potential for significant negative consequences, but relatively few governments have demonstrated that foresight2 Policy problems end policy design SUMMARY AND CONCLUSION Politics and government are ultimately about making policy. Harold Lasswell famously defined politics as "Who Gets What”, and public policy is the means theough which governments move resources within sociely to benefit some (and hopefully to some degree all) citizens and to impose eosts on others Policies are made through a complicated and often ‘contentious politcal proces, but behir all the political maneuvering there ‘may be some idea about what should be adopted ~ the policy design, 1 have been arguing that these designs are likely to be more effective if they involve eonscious thought about their content, but that may not always be ‘possible given political constraints. But although designing policy is a political game that any number of ‘partleipaats caa play, itis also an exercise in judgment and the application of information. ‘The above discussions of the characteristics of designs sand the types of designs point to important differences in the manner in which a government can make and implement policy designs. While ‘design implies a more or less rational and coherent approach to making policy, in reality the process may be based more on path dependence or on accidental discoveries of options. These forms of design that do not imply full rationality on the part of the designers need not be inferior to more purposive forms of intervention. Further, when governments must confront more difficult and less well- structured policy problems they must also respond with policy designs that are themselves les well-structured and more adaptable. Governments, and adres obey designs “The obvoes coanerxamlesaretheeaton ofthe Grorge W.Duthadnsation to Harviane Katina nd the econ of th Tramp adminstion to Hrreane Mara (One dear example she exten oe atau nthe formato of the Alorable Cte Act and the stengths and wees ofthe sesling rogram. ‘One could even rake the vomovhat pervs aegumeet ttt the selogeal style of Sain is more rational has tat which usu dssed a tonal desea process. ‘The tecopet sl is deductive, and based on ust of fa rapes from waioh setion is deivced, Te conventional design proces i largely indat, and alls {he desi onthe basi of evidence witout neserly havin 9 theoretalfeane ‘Ti sguoen near, canbe conta wth thon of peste sch as Dey (sce Haesvorh, 9H sidenote policymaking has bean expecinllypopalar in hltcae poly. See Dobro etl 2000. This cognition and éeiton of poly problem is refered to as “amin”, See Chapter? {Moet ressdly Dan Catson, Donald Tramp’ Secwary of Heung and Ushaa Developnent has mohasized thet pore isa "sate of ind” teen ons be over come by changing the mines ofthe poor. Tasca understandiag of porary bas phenomena, ‘TF we begin with dichotomies of policy problems, perhaps the most basic uss been between domestic policy and foreig policy. These two basic areas of policy have been dealt with throogh very different intellectual traditions, find generally have not communicated with one another. That said, how- ver this may bea false dichotomy, with thesame analytic issues arising in oth domains (Ingram and Fioderiein, 198; Costa da Silva, 2015). While foreign policy is a “game” played against other nations, domestic policy raking is « game that may iavolvo a wide range of potential adversaries criminals, microbes, weather, poverty, and human behavior, Those games nay be different given that many of the adversaries in domestic policy are mot sentient, bat some of the same logic of analysis ean be used “As well a3 having the potenti for common analytics, foreign poticy ray be diiven by domestic policy concerns (Putnam, 1988), and in tuen domestic policy may be heavily influenced by foreign policy concerns (Goureviteh, 1978). These linkages are dso referred to by the rather Inafortunate term of “intermestie". That i, in reality it may be impossible to separate foreign and domestic policy other than in the most literal fense. At the extreme, foreign policy may rfleet internal needs, such as the histori need for ofl from the Middle East by the United States. ‘One of the most basic of dichotomies that can be applied to policy problems captures the difference betwee» substantive and procedural Feoues facing the public sector. Generally, when we think of policy problems we think of the substantive categories alteady discuss above Por clizens die task of government, and ite allies in the private sector, js to provide goods and services to the public. That ssid, however, some of the most vexing problems facing the public seetor concern the mechanisms available for them to utilize when delivering services and to manage government itself, These issues are largely confined to public fadninstration, but have substantial relerance for policy as well? But procedures do influence substantive policy choices, For example, in ome cases the procedures developed to protect the rights of citizens can. nak policymaking on behalf of those seme citizens more difficult (See ‘Levi-Faus, 2005). Ta seminal work on problem definition in poliey analysis David Dery (1984) posed an intresting trihiotomy of yoliy problems. in Decy’s class. fication problems could be “situations”, meaning jst that something in the tocial and economic environment is not a* it should be. More specifically, problems may be “disrepancis” between what sand what is suppose bas this implies thre is some relatively clear conception ofthe desired state of affairs, And, finally, problems can be “opportunities”, meaning that” Poliey problems ond policy design policymakers can us a perceived problem as # stimulus for moving ahead With a policy plan they may ave been waiting to implement? “The three categories proposed by Dery may say as much about the agendasetting process 28 they do about the content of the problems. Jn particulay, the notion of opportuitescosresponds neely to the ideas of {Kingdon (2003) about windows of opportunity opening and making it possible for policy entrepreneurs 1o move forward with thei ideas. The idea of discrepancies, ikewis, may be almost a generic description of policy problems Simon kad said that policy represents the attempt toeot- reot a discrepancy between an existing condition and a desired condition, Rochefort and Cobb ‘The first comprehensive taxonomy of policy problems was advanced by David Rochefort and Roger Cobb (1993) in their analysis of policy problems and agenda-setting. They first divided issues about policy prob- Jems into two broad groups ~ causation and the nature of the problems themselves. The causation dimension then contained several attcibutes of causation that could be useful in understanding the manner in which agendas are constructed (Table 2.1), Theit dimensions of problems to some extent also foreshadowed more contemporary research on the role of complexity in policy problems and the need to understand the multiple causes of those problems (see below). ‘Rochefort and Cobb's discussion of the nature of policy problems also ‘ends to emphasize the likely political consequences of « problem rather than the more substantive aspects of the problem, and therefore was per- hhaps closer to a diseussion of agenda-setting than policy problems per se For example, the nature of the population affected by a problem appears in several ofthe dimensions they identified, with some problems affecting some portions of the population being more capable of generating positive politcal attention than are others Likewise, the extent to which a problemn js mote social than personal can be seen a8 a means of asking whether this problem is truly a public issue or whether itis one that is best addressed ‘through individual efforts, Issues that are essentially personal will not be ‘g00d foundations for political mobilization ~ unless, of course, there are ‘enough individuals who have the same personal issues ‘Several of the dimensions discussed by Rochefort and Cobis do addtess the substantive nature of the problems. For example, the crisis nature of problem, and its severity, may be important for placing it on an agenda; bat it also can be used to understand the amount of resources required ‘within the public sotor to adaress the problem and the speed with which ‘a response may be required, Likewise, the instrumental-expressive dimen: The problem of policy prolems 4 ‘Table 2.1 Rochefort and Cobb's characteristes of poliey problems Dimension f problems Options Personalfimpersoned Ttended/aecidental ‘Blame allocated/blame avoided Problem causation ‘Simplelcomples Severity Depress of severity Tcidence Growing, stable or dectining Patterned by clas ‘Age, race ete? Novelty ‘Unprecedentedfaiiar Proximity Personal/ocal Criss Cesiinon-riss ‘Target population Worthyfunvorthy ‘Desorvinglundesrving Funllaclerange ‘Sympatheticthestening Means-ends orientation of problom defer _ Instumentalexpressive [Nature ofthe solution “Avalabe/aon-existent ‘Acceptabllabjctionable ‘Affordablelunaffordable Source: Roshelot sad Cob (159), p61 sion of a poliey problem reflets the types of responses that may be most appropriate for government to make, Some problems may require only symbolic interventions, or indeed those intangible interventions may be the only instruments available. Most policy problems, however, will quire riore substantive responses ~ that is, funding, service delivery, or regula tion, But the effective policy designer must consider these dimensions, and the possibility of making inexpensive symbolic responses in liew of the more expensive alternatives that are the usual response, ‘The final dimension of problems in this taxonomy links the problems with solutions. This linkage is not the type of direct mapping of sets of solutions onto sets of policy problems that has been the holy grail of conventional models of policy design. The linkage does, however, ask a set of important questions about the possibilty of a problem being addressed successfully through the public sector, This dimension of problems coming. from Rochefort and Cobbs asks, most simply, whether there are acceptable, affordable solutions available fo policymakers or wil greater creativity be required; oF, at an extreme, is there nothing government can do, even if it peresves that there isa genuine problem facing society