Introduction
Policy design has been occurring as long as governments have been making policies. Design often has been more implicit than explicit, but the academic literature on policy design (Bobrow and Dryzek, 1987; Peters, 2018) has emphasized the need for more conscious thought about designing. The more explicit design process should increase the number of options considered by designers, and can make the process more open to stakeholders, as well as ordinary citizens. A more conscious approach to designing may also approximate the idea of “evidence-based policymaking” that uses available information to make better policies (Cairney, 2016).[i] The academic literature on policy design does abstract from the real world, but the purpose of that research is to improve policy as well as score academic points.
This paper examines the elements of good policy design and analyses possible processes and institutions used for designing.
What is a Good Policy Design?
Public policy, like architecture or engineering, is a design science. Its ultimate purpose is to produce public policies that improve the lives of citizens, and to do so at a reasonable cost. Policy design is arguably more difficult than other forms of design. This is largely because it is attempting to alter the behavior of people who do not always behave in the way that designers want or expect. Likewise, policies involve systems of action involving multiple actors, rather than an object, Finally, policy design involves more unknowns and uncontrollables than do other forms of design. A well-designed policy may be upended by changes in the economy, or changes in social conventions, or any number of other unanticipated events,
The most obvious question to ask about policy design is what constitutes a good design? Individuals and organizations may work for days or months to produce a design, but what makes it a good one? There are a number of important attributes of a good design, but the following four are especially relevant.
- It has to work. This is perhaps obvious, but any number of policies have been implemented by government that were unlikely to work. That said, few policies are totally successful, so the criterion of “working” needs to be considered carefully.
- It should be elegant as well as functional. Well-designed policies not only perform the tasks but they do so with a minimum of resources and a minimum of burdens on recipients (Herd and Moynihan, 2018) or taxpayers. In some instances, however, the use of sheer power may be needed to solve a problem, and elegance will not be relevant.
- Flexibility. A good policy will work with changing circumstances, although if changes are severe or central to the design of the policy it may still work.
- Interdependent. A good policy design recognizes that a policy is just one among dozens or even hundreds affecting the same target populations, and therefore it must take those other policies into account (see the Model of Coordination Below).
What are the Elements of a Good Policy Design?
The above general points about policy designs lead to the question of what has to be included in a good design. I will not be using the term “model” in this discussion in a restrictive sense, but only to point out that the designer should have considered these points carefully when making the design. Further, the designer should ensure that the various elements of the design are compatible with one another if the design is to have much chance of success.
1. A Model of Causation
Public policies are created to solve problems, and those problems are caused by some social process. The problem for designers is that there may be many ideas about that causation. Poverty, for example, is argued by people on the political left to be caused by the failures of the capitalist economy, while for those on the political right it is caused by individual failures such idleness. Different ideas about causation will quite naturally require very different types of solutions
Policy problems are often discussed in terms of the policy domain in which they occur. Something may be defined as a health problem or an agriculture problem, but the policy domain does not always provide much guidance in terms of causation or characteristics of problems. There may be as much variance within a policy domain as there is across domains. In education, for example, the policy issues in a kindergarten are vastly different from those in a research university. Therefore, it is important to think about the nature of policy problems more analytically rather than thinking only about what government department is involved (Peters, 2018)
2. A Model of Instrumentation
After a decision about causation has been made, the instruments that will be used to “solve” the problem must be selected. Governments have a very large tool chest at their disposal, so the choice may be difficult. Scholars (Hood, 1984; Bemelmans-Videc et al., 1989) have classified these tools in various ways, but the fundamental differences are among tools that use information, those that use law, and those that use money to create changes in the society. Also, it is important to note that although tools may be equivalent ways to achieve goal, they will have very different political economies.
The choice of instruments is important, but so too is the way in which people in the public sector use those instruments. In one empirical study (Linder and Peters, 1989) when civil servants were asked what instruments they choose to solve policy problems the answers were somewhat surprising. One large group said they always would choose one type of instrument–lawyers chose legal instruments, economists chose market-based instruments, etc. Another large group of respondents said that the nature of the instrument did not matter; they as public administrators could make any instrument work. Only a small group said they would make efforts to match the instruments to the nature of the particular problem.
3. A Model of Intervention
Public policies, once adopted, have to be implemented. Governments and their allies in the private sector have to develop the mechanisms to make a policy work. Too often designers have developed wonderful schemes to achieve a policy objective without thinking of how it could actually be made to work (see Pressman and Wildavsky, 1974), but a successful policy design must pay attention to that component of governing.
Most implementation in contemporary governments is done by “implementation structures” rather than by single organizations (Hjern and Porter, 1980). That is, public sector organizations tend to cooperate with private sector and third sector organizations to deliver programs. This enables government to leverage the resources, including legitimacy, that the non-government actors may have, but it also exacerbates the complexity of implementation. Therefore, careful design is needed to ensure that the right sort of cooperative structures are being created.
4. A Model of Evaluation
After a policy is adopted and implemented it must be evaluated. Is it working? Do the negative unintended consequences outweigh the positive intended effects of the policy.[ii] As was true for causal reasons for a policy problem, there can also be a number of dimensions of evaluation. Historically there were major evaluation programs within government agencies, universities and think tanks performing detailed analysis of the success of programs. These have largely been closed, and much of the analysis of success and failure is done through performance management that focuses more on the short-term effects measured through relatively simple indicators (Bouckaert and Halligan, 2007).
As well as the usual questions about efficiency and effectiveness, evaluation also implies examining the values that are being pursued in the policies. Are the results produced by policies fair and equitable? Do the policies allow citizens to make as many choices as possible about what will happen to them? Do the policies take future generations into account adequately? These normative questions and many more need to be asked when considering whether a public policy should be adopted or continued.
5. A Model of Coordination
The final requirement for a workable policy design is a model of coordination. As already argued, few if any policies exist without conflicts with some other programs. Even if there are no conflicts, the policy may perform better if it can work effectively with other programs. This point also emphasizes that there are few entirely novel policies, and therefore making a new policy involves finding out how to work with existing policies, For example, much of the immensely long Affordable Care Act (Obamacare) passed in the United States in 2010 involved making this new program work legally with existing programs.
Coordination among policies can occur prior to implementation, or as a part of the implementation process. Everything else being equal, coordination is easier when undertaken as a part of implementation. When coordination is attempted while the policy is being adopted (policy integration, see Tosun and Lang, 2017) then the bureaucratic politics and competition for resources endemic in the political process may prevent building effective coordination designs. However, when a policy is being implemented and there are real clients with real needs to be served, agencies may be more willing to find solutions.
Bringing all these elements of design together in a coherent policy is a challenging task. All of this thinking about policy and its components must be conducted in a politically charged environment, with multiple actors pursuing their individual and organizational goals. As it is being implemented, the policy depends on the behaviors of numerous actors, including perhaps most importantly the intended targets of the policies. Understanding the dimensions of good policy design, and appreciating the difficulties involved, can however facilitate producing the types of effective programs desired.
Institutions and Processes for Designing
With the requirements for good designs in mind, we can proceed to consider the possible processes and institutions used for designing. Each of the three options I will identify has its own virtues and vices, and therefore the policy designer must weigh the probabilities of success for his or her policy problem with each of the available options. These options also involve political values, so that a designer may choose a difficult process simply because s/he thinks it is the right thing to do. There may be other ways to design policy, but these three cover the major structural patterns of performing this task for the public sector.
Technocracy. Policy design has been criticized since its inception for being too technocratic. The assumption is that design involves the use of a small group of experts who will use their expert judgment to design a policy, seemingly ignoring the wishes of the people, and of democratically chosen representatives. There are instances of this type of policy designing, by planning agencies, central banks, independent commissions, and, at times, the courts. In most democracies, this is a small, but still significant, percentage of total policymaking.
This autonomous form of policy designing has the advantage of being removed from politics, giving the decision-makers the capacity to make decisions based on technical and professional criteria. Technocratic policymakers should be able to make decisions quickly, and thus respond to crises more effectively than other forms of policymaking. On the other hand, designing policy in this way can limit the number of options considered, as well as the type of criteria that are used to evaluate the options. “Expert failure” is a real possibility with this type of policy designing (Koppl, 2018).
Participatory Policy Design. Participatory policy design is the antithesis of technocratic policy design. As the name implies, this form of designing seeks to maximize opportunities for citizens to make inputs into the policy process and to shape the emerging designs. Some forms of this type of designing, such as public hearings, have been used for many years, but the first major innovation came with participatory budgeting in cities, initially in Brazil, that allowed citizens to have substantial influence over local budgets. This innovation then spread around the world (Bartocci et al., 2023) and has been expanded to include participatory processes for all types of local policies (Boosabong and Chamchong, 2019), Other methods for participation, such as citizen juries, deliberative policymaking and policy labs (Wellstead et al., 2021) have followed.
Again, the strengths and weaknesses of this approach to design are perhaps rather obvious. In strictly design terms, it has the advantage of allowing more options to be considered before making a decision, as well as the opportunity to discuss those options fully. But that full consideration of options is purchased at the price of slower responses, and the need to govern through the lowest common denominator when there are conflicts among multiple groups involved in the process. Further, from a democratic perspective, the openness of the participatory method of design helps to legitimate policy decisions, and that legitimacy may help to facilitate their implementation.
Traditional Policy Design. Most policies have been, and continue to be, designed by conventional processes existing in legislatures, bureaucracies, and political executives. These institutions vary across countries in their capacity to make and implement effective policy design, but they are central components of government structures, in both democratic and authoritarian regimes. They also have the virtue of being institutionalized and having some political legitimacy, whereas the other methods may appear more ad hoc.
These policymaking institutions combine some of the strengths and weaknesses of the other two forms of designing policy. The conventional institutions, through the bureaucracy and advisors for legislatures and executives, have some of the technical capacity of the technocratic mode of design. At the same time, through representative institutions, as well as public hearings and other opportunities for participation, they can fulfill at least some of the democratic goals for the participatory forms of designing. And these institutions, if imperfect, are familiar to citizens while the others–especially the technocratic ones may appear inaccessible.
Conclusion
This short paper has sought to provide some background on the concept of policy design by first discussing what is needed for a viable policy design. The five elements outlined, if developed well, allow policymakers to move from some vague idea of the existence of a policy problem to a final design containing all the elements needed to make the policy work. This does not guarantee that the policy will be a success. Making policies is not easy and there are any number of factors that will upset the best intentions of the designers. And even if the policy achieves some or all of its goals, there may still be negative unintended consequences that outweigh the good that has been accomplished.
This paper has also discussed the alternative institutions and processes that are available to would-be designers. A designer may want to select the process of design as well as the elements of design. For example, if making policy in a relatively unknown domain, the designer may choose a participatory method that opens the process to a wide range of ideas and new proposals for policy. On the other hand, a policy with a strong knowledge base may be dealt with best through a more technocratic process. Choosing designs also involves selecting the instruments for designing.
References
Bartocci, L., Grossi, G., Mauro, S. G., & Ebdon, C. (2023). The journey of participatory budgeting: a systematic literature review and future research directions. International Review of Administrative Sciences, 89(3), 757-774.
Bemelmans-Videc, M. L., R, C. Rist and E. Vedung (1998). Carrots, sticks, and sermons: Policy instruments and their evaluation. (New Brunswick, NJ: Transaction).
Bobrow, D. B. And J. S. Dryzek (1987) Policy Analysis by Design (Pittsburgh: University of Pittsburgh Press).
Boossabong, P., & P. Chamchong (2019). The practice of deliberative policy analysis in the context of political and cultural challenges: lessons from Thailand. Policy Studies, 40, 476–491.
Bouckaert, G. And J. A. Halligan (2008) Managing Performance: International Comparisons. (London: Routledge).
Cairney, P. (2016). The Politics of Evidence-Based Policymaking. (London: Palgrave).
Hjern, B. And D. O. Porter (1980) Implementation Structures: A New Unit of Administrative Analysis. Organization Studies 2, 211–227.
Hood, C. (1984). The tools of government. (Chatham, NJ: Chatham House).
Koppl R. 2018. Expert Failure. (New York: Cambridge University Press).
Linder, S. H., & Peters, B. G. (1989). Instruments of government: Perceptions and contexts. Journal of Public Policy, 9, 35–58.
Peters, B. G. (2018) Policy Problems and Policy Design (Cheltenham: Edward Elgar).
Peters, B. G. And M. L, Nagel (2025) From Benign to Malign: Unintended consequences and the growth of zombie policies. Policy and Society, 44, 136-49.
Pressman, J. L. And A. Wildavsky (1974_ Implementation (Berkeley: University of California Press).
Tosun, J., & Lang, A. (2017). Policy integration: mapping the different concepts. Policy Studies, 38(6), 553–570.
Wellstead, A.M., A. Gofen & A. Carter (2021) Policy innovation lab scholarship: past, present, and the future – Introduction to the special issue on policy innovation labs, Policy Design and Practice, 4, 193-211.
[i]A good deal of implicit policy design becomes “policy based evidence making” in which political leaders select, and or create, evidence to support their policy preferences rather than having a more open process.
[ii]There are almost always unintended consequences of policies. See Peters and Nagel (2025).