The foresight field has long suggested that modern models of planning and ‘classical approaches’ to policy have serious short-comings. In particular it has critiqued the focus on prediction (‘predict-then-act’) and on control. In my view some of these critiques go too far, such as some colleagues who have argued that we should “rid the world of the word prediction”. Prediction is an essential part of the world, such as in advancing scientific understanding by testing hypotheses.
The original paper on “wicked” problems (versus “tame” problems) by Rittel and Webber provides an early 1970s view of these issues. In many respects it is a highly persuasive critique of attempts to make “the planning idea operational” (those are their words).
I plan to trace these debates more fully through to the present day; however, a few propositions and themes in the Rittel and Webber stand out as important ones to consider and to research further. This post outlines these briefly. At the end I note some big picture reflections.
Key proposition 1: “there is no opportunity to learn by trial-and-error”
Rittel and Webber argue that wicked problems are a “one-shot operation”. This is because “every implemented solution is consequential” and “leaves ‘traces’ that cannot be undone”. This leads them to assert that “actions are effectively irreversible”, preventing traditional trial-and-error learning. That is, you don’t play the same game twice – as you would, for example, in the trial-and-error learning that occurs in games of chess, when solving mathematical puzzles, or in engineering.
This made me think about Australian immigration policy regarding asylum seekers. The Coalition claim they have the solution (to the ‘problem’, as they define it) based on the Howard Government’s experiences. This ignores the fact that the ‘problem’ has evolved in significant ways.
Key proposition 2: “every wicked problem is essentially unique”
Whilst Rittel and Webber acknowledge that for any two problems common properties can be found, they argue that “there always might be an additional property that is of overriding importance”. They add: “there are no classes of wicked problems in the sense that principles of solutions can be developed to fit all members of the class” (p.164), suggesting that “in the complex world of social policy planning, every situation is likely to be a one-of-a-kind” (p.165).
“Despite seeming similarities among wicked problems, one can never be certain that the particulars of the problem do not override its commonalities with other problems already dealt with” (p.165)
In contrast, for ‘tame’ problems, “there are explicit characteristics … that define similarities among them, in such a fashion that the same set of techniques is likely to be effective on all of them”.
If the diagnosis outlined by Rittel and Webber is correct then this has enormous implications. It again reduces the potential to learn from experience (this time from approaches used to address other wicked problems), and suggests practitioners/consultants cannot develop a set of core tools that they apply to all the various wicked problems they encounter. Wow. Talk about throwing a ‘spanner in the works’ of the change consultancy field and wider policy and planning fields.
Key proposition 3: “the full consequences cannot be appraised until the waves of repercussions have completely run out”; we have no way of tracing all these waves “ahead of time” (i.e. in an anticipatory way) or even “within a limited time span”.
This is discussed as part of a dimension termed “there is no immediate and no ultimate test of a solution to a wicked problem” (p.163). Implemented solutions will generate consequences over “an extended – virtually an unbounded – period of time” (p.163). In contrast, they suggest that for ‘tame’ problems one can determine more simply “how good a solution-attempt has been”.
This proposition naturally made me think about evaluative research. A key issue in foresight evaluation is when to do an evaluation (i.e. the timing). That is, when is it fair to assess the consequences/effects? When is the ‘end’ of the intervention/process? There is no immediate and/or simple test. This issue is potentially amplified when foresight processes are used to help tackle a “wicked” problem!
Additionally, more advanced methodologies may give us some more traction on this dimension. (Although, it must be said, in spite of the best analysis, policy and plans often have more unintended consequences than intended consequences.) More recent writers have suggested complexity theory and thinking, and approaches like “applied forward reasoning” may improve such advance assessments – such as through scenario building and analysis processes.
Key proposition 4: “There is no definition formulation of a wicked problem
Here, Rittel and Webber are arguing that for a ‘wicked problem’ it is not possible to state “an exhaustive formulation” of the problem “containing all the information the problem-solver needs for understanding and solving the problem”. This leads to a further claim that the idea that a planning project can be organised into distinct phases (e.g. understand the problem, gather information, analyse and synthesise information, formulate solution) does not work for wicked problems.
This is potentially a big stumbling block when foresight practices are used to address wicked problems. Often problem definition is the first stage of a foresight process. For example, this is the starting point in many backcasting methodologies. Scenario planning often begins with an orientation phase in which a central focal issue (or problem) is defined. If we can’t define the focal issues/problem then how do we proceed to run the whole process? Also Voros’s generic foresight process adopts a sequential phases model, similar to those critiqued by Rittel and Webber.
Finally, and overall, Rittel and Webber conclude that:
“the problems that planners must deal with are wicked and incorrigible ones, for they defy efforts to delineate their boundaries and to identify their causes, and thus to expose their problematic nature. The planner who works with open systems is caught up in the ambiguity of their causal webs. Moreover, his would-be solutions are confounded by a still further set of dilemmas posed by the growing pluralism” (p.167)
Reflections: can there be an applied science for tackling “wicked” problems?
The above propositions are only a selection of those advanced in the paper. Nonetheless, for me it has raised a key question: can there be an applied science for tackling wicked problems? If you cannot learn from experience, if all wicked problems are unique, and the consequences of interventions cannot be known in advance, and no wicked problem can be definitively formulated, then how can an applied science be developed for addressing these problems? If it was a boxing match, it would be a super-effective left, right, left, right combination – possibly a KO. Indeed, Rittel and Webber assert that “social professionals were misled … into assuming they could be applied scientists” (p.160)!
And yet, this seems to be the direction of recent fields such as sustainability science, action research, and related approaches to transdisciplinary research. Some consultancies also offer various forms of applied research that claim to offer effective methods for addressing a wide range of “wicked” problems. The emergence of adaptive policy approaches is also significant (see this book), which draw on better understandings of complex adaptive systems and forms of adaptive management.
Investigating this further appears to be an important line of inquiry. It potentially has important practical implications for the role and value of evaluative research and the potential to develop “reflective practice” when dealing with wicked problems. Rittel and Webber’s analysis appears to suggest that there a serious limits to reflective practice when tackling wicked problems. My principal supervisor, Chris Riedy, has suggested that Rittel and Webber may have overstated the problem – and that, therefore, perhaps learning is possible – and their analysis might further justify a ‘realist evaluation’ framework.
It also appears to have important implications for modes of intervention, such as whether a public policy approach is the most efficacious approach for wicked and “super wicked” problems. Bottom-up, and “organic” (rather than planned) change, may be more important than is currently realised.