Abbe Boulah’s Weblog

RE-EXAMINING WICKED PROBLEMS: UNDERSTANDING AND IMPLICATIONS

Thorbjørn Mann 2021

This is the second post on the question of claims by proposed problem-solving ‘approaches’ to successfully ‘solve’ Wicked Problems.

Looking for reassuring answers to the question whether some approach, method or ‘perspective’ can be expected to live up to claims that using the respective approach will reliably result in ‘solving’ Wicked Problems, it may be useful to turn the question around and look at the concept of ‘wicked problems’ itself, and its understanding. Are its ‘properties’ and implications really justifying the frequent automatic rejection of such claims, or claims of a technique guaranteeing solutions? The following first attempt, for discussion, takes a stab at this question, examining each of the WP properties:

* “No definitive problem formulation”:

This feature reflects the fact that different people involved in a project will have very different opinions about the problem, and that the acceptance of one view of ‘what the problem really is about’ is a choice or decision. It is a stern challenge to the habitual recommendation to begin a problem-solving process with a ‘clear statement of the problem’. The implication: to avoid controversies and disruption from occurring later in the process it is necessary to not only begin such a process with a widely open invitation to affected and interested parties to contribute many different perspectives of the problem, but to keep the process open to emerging insights on this issue. 

* “Every WP can be explained in many different ways”: 

The same recommendation holds for this WP property: 

* “Every WP can be seen as a symptom of another problem or set of problems.”:

One obvious implication of this feature is that any proposed ‘solution’ idea, however promising, can be dismissed as ‘only treating the symptom’. The question should therefore be raised early in the process to be discussed, and any necessary decisions resulting from it agreed upon – such as having to shift the entire effort to a different institutional level or entity – before devoting much time and energy to develop a ‘symptom-treating’ solution. 

* “Every wicked problem is essentially unique”:

The implication of this feature is that ‘tried-and true’ methods and lessons from previous cases may not be applicable to a new WP.  However, could it be that the stark formulation of the property unnecessarily hides the fact that the significance of similarities and differences between the new problem and similar cases are a matter of degrees? There may be part of the problem that are sufficiently ‘similar’ to warrant the application of known tools. The process should address this question by looking at details and make decisions about using known methods where applicable and devote efforts to develop new tools as needed. 

* “WP ‘solutions’: not ‘True or False’ but ‘Good  or Bad’:

This reminder was especially necessary at the time the WP issue was raised and published: there was a veritable movement of stressing ‘fact-based’ decision-making, that is, using ‘objective facts’ about a proposed solution’s measurable performance as the decision criterion. This trend seems to re-emerge periodically, (under slightly different banners such as ‘science’ or ‘expert advice’), perhaps because of inappropriate populist switching to decision-making based on ill-informed intuitive ‘goodness’ judgment or insistence on ideological principles decrying the facts presented by discipline experts as ‘elitist oppression’. So the reminder should perhaps be revised to reflect that the real issues are 

    –  the selection of the performance measures for which the ‘facts’ are then established – of course factual information must be provided and assessed for any problem, wicked or tame;

    – these ‘facts’ will always be qualified by probability; especially the  predicted ‘facts’  offuture solution performance (which of course aren’t even facts yet!); and

    – the necessity of communicating about how fact-measurements and predictions relate to the ‘good’ or ‘bad’ judgments (the process Rittel called by the somewhat problematic name of ‘objectification’); and

    – the most important question of  whose judgments  should determine the common decisions about accepting or rejecting proposed ‘solutions’. 

* “No immediate nor ultimate tests”:

This property refers to the difference between scientific hypothesis-testing and the discussion of proposed plans to remedy social problems, as well as to the feature that plans and policies will have chains of consequences that make ‘immediate tests’ meaningless even if we had such tests and ‘ultimate tests’ un-specifiable because the time span of those consequences is indefinite. However, any reaction of doing without any ‘testing-like’ efforts is seriously mistaken. Two considerations:

    – Simulation models used properly (that is, to explore the possible consequences of different actions and strategies taken today) can be considered a kind of test or better ‘evaluation’ – the best tools we have for predictions, none of which are establishing true facts since they all deal with future developments: probabilities.  And

    – Argumentative discourse: The sharing and assessment of the proverbial ‘pro’ and ‘con’ arguments about proposed plans.In which simulation model results may play a significant role, but the essential difference is that planning arguments contain the ‘ought’ premises that are not properly assessed as ‘true’ or ‘false’ and thus the same ought claim may be ‘plausibly’ seen as ‘ought’ or ‘desirable’ by some affected parties but as ‘not desirable by others. The degree to which a plan is perceived as achieving the ‘ought’ state (of a problem perception) is the basis for ‘good’ or ‘bad’ judgments of the plan. 

I have suggested that to Popper’s advice about scientific hypothesis-testing: 

“We are entitled to accept a hypothesis as corroborated (only) to the extent we have done our very best to show that it is false, plausible and it has survived all those tests” 

the closest analogous ‘test’ we have in design and planning is the following:

“We are entitled to accept a plan as plausible and ‘supported’ (only) to the extent we have done our very best to expose it to all the most plausible counter-arguments  (‘cons’) and those have all been shown to be flawed or outweighed by supporting arguments (‘pros’). 

Evaluation procedures and approaches to develop measures of plausibility of individual judgments of planning arguments have been described, as the closest we have to ‘testing’ plan proposals.  

* “No well-described, finite sets of admissible operations”

This feature is set against disciplines like mathematics where the operations of addition, subtraction, multiplication, division etc. are the admissible operations that define ‘tame’ problems. It warns that approaches attempting to specific a finite set of such operations for WP’s are liable to encounter new ways to tackle them – ‘anything goes’ if it works. 

* “No enumerable set of potential solutions to a WP”:

The insight that the ‘solution space’ for WP’s may be infinite and impossible to define implies that claims of finding ‘optimal’ solutions are meaningless: there may be even better solutions somewhere in regions of the solution space that were not explored. This constrains the discussion to the more modest quest for solutions that are ‘good enough’ within the regions an approach is able to examine: a feature that was earlier proclaimed as ‘satisficing’?

* “No inherent ‘stopping rule’ for efforts to deal with WP’s”:

The implication of this feature is simply that stopping rules  for efforts to tackle WP’s are not provided by the problem itself but by constraints outside the problem but acting on the task force working on it – and as such ‘arbitrary’ and debatable: financial and time constraints being the most common such limitations. For the problem itself, ‘we can always try to do even better”. 

* “Every effort to deal with a WP is a ‘one-shot operation’”: and

* “No ‘trial and error’”:

These two properties are intimately related. We cannot rely on the trial and error strategy to learn how to ‘solve’ a WP: Any actions on the problem itself will expend resources, generate new consequences: the ‘next trial’ will now be a very different problem.  

* “The WP-planner has no ‘right to be wrong’:

Like the hypothesis-testing issue above, this aspect refers to a fundamental difference between science and WP-planning. The scientist having performed a test that refutes a hypothesis may be disappointed — but that is making a legitimate and rightful contribution to scientific knowledge. The would-be WP- solver failing to remedy the problem is actually ‘making things worse’. The plausible implication then is the call for holding the planner – or the decision-maker for the implementation of the plan liable’ — ‘accountable’ — for the failure. More often than not, such calls are rather meaningless, if there is no ‘account’ involved (other than perhaps a decision-maker’s position or ‘reputation’: how does it balance suffering of people affected by the problem of the wrong solution?). Should efforts be devoted to finding better ‘accounts’ for this issue? Rittel suggested one implication: the ‘complicity model’ of planning. Taking this aspect seriously, no decision-maker would be able to accept responsibility for major decisions if it required ‘investments’ equal to the risks of failure of plans. It would be necessary to find ‘accomplices’ willing to share that risk. Again, what kind of ‘account’, what ‘currency’ might be used for this?  (I have sketched one possible idea: the use of the ‘reputation’ account of ‘merit points’ earned for the value of contributions to the public discourse to have decision-makers ‘pay’ for important decisions.) 

* “Distributed information”:

This issue refers to several aspects of large public WP’s: 

    – The need to assemble ‘factual’ information about how the problem – and any proposed solutions — affect many different parties ‘out there’ – that are not yet documented and certified in knowledge bases and experts’ knowledge. This may require research and information-gathering for which the need will only become apparent as the discourse proceeds, so that initial estimates of needed resources will be unreliable; and initial surveys to gather such information will be insufficient: the questions to be answered will only emerge later on: the information-gathering effort must accompany the process throughout;

    – The question of ‘getting ‘ the information may require offering some incentives for people to contribute it – early enough to be useful (rather than complains after the fact) – and mechanisms for assessing its truthfulness or validity;

    – If the property of ‘not true or false but good or bad’ is valid, and thus should determine the decision, these judgments will have to be the judgments of the affected parties. This will require a clear distinction between ‘factual’ information (that must be ‘verified’) and goodness/badness judgments that must be accepted as individual’s assessments and aggregated into overall statistics of sentiments of approval or disapproval. This task is not adequately addressed by many ‘approaches’; the effort to achieve consent or even consensus in small task groups seems to sidestep rather than systematically and transparently confront it. (See also the issue of ‘making decisions ‘on behalf of others’, below.)

* “Nonlinear and counter-intuitive system behavior”:

It is the merit of ‘systems modeling’ to bring this issue to the attention of planners and decision-makers. The simulation models aim at overcoming the resulting prediction difficulties of this ‘complexity’ of the systems involved in WP’s. The connection between the prediction results and the ‘goodness’ judgments (of the many affected parties) has not been sufficiently well explored much less convincingly resolved.

* “The ‘doorknob’ syndrome:” 

The warnings against getting lost in the upward or downward ‘cause’ or ‘symptom’ issues of WP’s are understandable but carry the risk of under-estimating the reality and significance of such relationships.  The rules that can guide decision about how much attention to devote to them, like the ‘stopping rules’ discussed above, are often extraneous to the problem – which can lead to flawed decisions. 

* “Making decisions ‘’on behalf of others”: 

Governance and planning decisions on public issues have traditionally been taken by leaders, officials, or representatives of the community, with the justification that these decision-makers are sufficiently familiar with their constituencies to make decisions ‘on their behalf’.  This can mean one of two things: Either they know (or claim to know) ‘what’s best’ for the community — even if there are people in the community who disagree — or they know the basis of judgment (the way the community members relate their goodness judgments to the facts of the matter) well enough to make judgments ‘as the people themselves would’. Both assumptions have been questioned, and current efforts to validate either assumption are cumbersome and unconvincing, adding to the wickedness of the problem at hand. 

* “The ‘making a difference’ syndrome”:

Many people are perfectly content with the provisions of planning decisions being made by leaders, officials or hired consultants: delegation of work allows us to focus on ‘our’ work and priorities.  But to the extent people are – in the name of ‘citizen participation’ – becoming more extensively involved in public problem- solving issues, this makes that involvement a part of their lives, in which they may want to ‘make a difference’ – a somehow outstanding contribution. Consciously or subconsciously, this may mean ‘doing things differently’ from the way things have been done, or from what some recommended ‘approach’ or method is proposing. The planning process itself becomes a part of the plan, and they want to make it ‘theirs’. Regardless of how appropriate or allowable this may be in the view of other participants or approach promoters, this will introduce unforeseen complications into the process. If it is seen as part of these individuals’ ‘right to pursuit of happiness’ – that governments are supposed to ensure: should all public planning efforts include provisions for such efforts – and what would they look like? 

There may be some commonalities of implications in these properties that are not apparent in the individual items, and that deserve closer examination.  One such common assumption is the reference to the ‘WP solver’. Is this an unspoken and unquestioned assumption of a single designated person or team to do the problem-solving ‘on behalf’ of the community affected? The reality of public projects is that there are always multiple institutions with various decision-making responsibilities – the task then also involves the organization of constructive coordination between all these entities. 

A larger common aspect is that meaningful response to WP properties requires some common communication and coordination platform. For all the progress of information technology over the last decades, an appropriate and effective platform for this purpose remains to be developed.

The platform, finally, will also be the venue for reaching decisions. None of the WP properties mention this explicitly, but their implication is that the traditional decision-making modes (such as voting) do not meet the expectations of suitable responses to the issues – e.g. being based on transparently explained individual ‘goodness’ judgments. Especially for problems transcending existing governance boundaries with different decision-making entities and rules, this will become an urgent consideration. 

Are these sketchy observations indicating an urgent need for wider discussion? 

— o —