Wicked Problems

In the 1973 article, “Dilemmas in a General Theory of Planning,” Horst Rittel, who was a professor of the science of design at the University of California, Berkeley and Melvin Webber, who was a professor of city planning at the same institution, recognized many problems include a social dimension. While many problems are complex and irreducible, those with a social dimension were fundamentally different from those without, and Rittel and Webber introduced terms to differentiate the types of problems.

In many areas of human endeavor, problems are definable, understandable and consensual: Each problem can be attributed to one (or a small number of) clearly identified and defined causes that can be understood through systematic study, and there is consensus in the population that the problem should be solved as well as agreement when it has been solved satisfactorily. Rittel and Webber used the term tame to capture the definable, understandable, and consensual nature of these problems. Tame did not mean the problems were simple to solve or that the problems were not important and ethically challenging. Tame did mean that the problems were generally studied and solved through science and engineering that was guided by defining goals and thinking about systems. While recognizing that no problem can be completely tame, Rittel and Webber suggested that the living conditions for large populations of humans had been improved by solving tame problems related to (for example) health care and medical procedures, transportation and sanitation systems in cities, and agriculture.

Rittel and Webber concluded, however, that most social problems were not tame. These problems cannot be easily understood, are not completely definable, and there is no agreement about the existence and resolution of the problem. For ill-defined social problems (such as education, law enforcement, and local governance) understanding the problem and arriving at consensus on the nature of the problem and the evaluation of the solution is difficult to the point that any judegment must be tentative and qualified. Rittel and Webber called these problems wicked. They argued the methods for solving tame problems will not produce adequate solutions for wicked problems, and even defining a wicked problem as if it were tame is an impediment to planning solutions that can be deemed sufficient.

Fundamentally, wicked problems and tame problems are different; they require different planning approaches, planning actions, and approaches to evaluating success. In general, planning so as to improve efficiency and achieve desired outcomes appears to be effective in solving tame problems, but not wicked problems.

Characteristics of Wicked Problems

In their original article, Rittel and Webber defined 10 characteristics of wicked problems. Other scholars have reduced the number of characteristics by combining some that are similar. Regardless of the number of characteristics used to define wicked problems, scholars who focus on these problems concur the wickedness of a problem arises from the social nature of the problem and the diversity of valid and mutually contradictory perspectives on the problem that can be supported by individuals who are affected by the problem and its solution. Each individual’s perspective affects how he or she frames a wicked problem and evaluates its solution. Individuals’ perspectives vary and even an individual’s perspective changes over time, which makes it difficult to objectively deconstruct the problems into completely and clearly definable causes and effects as well as to objectively evaluate outcomes. Rittel and Webber’s list of properties of wicked problems include the following:

Ill-defined

There is no definitive formulation for wicked problems. In the natural sciences, problems are clearly and completely defined, and all who are familiar with the field will concur on the nature of the questions and the nature of the methods to be employed to solve the problem. (This holds as long as the research works within the dominant paradigm of the field.) If a research problem is too complex or too multifaceted, the scientist defines the problems with great precision and takes steps to control all relevant factors; this is commonly called reducing the problem. Such control over the boundaries of and within a problem is not possible when solving wicked problems; the boundaries between causes and effects blur and unidentified causes may be influencing the resolution of the problem for the entire population or for a part of the population. In many instances, the result is the wicked problem and its solution become inseparables. The decision to implement a different solution would have been a decision to solve a different problem or to address a different factor affecting the wicked problem.

No Stopping Rule

Wicked problems have no stopping rules to identify with certainty when it has been solved. Although certain conditions may have been met by the solution, and the solution may have been deemed sufficient, there remains uncertainty. By gathering more information or looking at the problem or the selected solution from a different perspective, planners may arrive at new understanding of the problem or the solution that may lead the planners to reassess either or both. Also, the conditions that contribute to wicked problems change over time, so the problem and its solution are dynamic. Just as there is no way to be done building a sand castle, there is no way to be done with solving a wicked problem. Solutions can be deemed sufficient for the current situation, but further refinements or enhancements to the solution to a wicked problem are always possible.

Subjective Evaluation

Rather than being judged as accomplished or not (in a true-false manner), solutions to wicked problems are judged as better or worse by those who experience the solution. When solving tame problems, scholars and planners make systematic and objective observations. Solutions to wicked problems are judged by the people who are affected by the solutions, and their judgments are made according to their (probably changing) circumstances at the moment the judgment is made. Because of this, wicked problems are marked by subjectivity. Adjectives such as good or bad, working or dysfunctional, acceptable or not acceptable, satisfactory or unsatisfactory are common when people assess solutions to wicked problems, and these terms demonstrate the necessarily subjective assessments. These adjectives are not acceptable to those who seek to solve tame problems in an objective manner. These adjectives can cause school leaders to react with consternation when there is discordance between their assessment and others’ assessment of solutions that are implemented in school settings.

No Ultimate Evaluation

There are neither immediate nor ultimate tests for a solution to a wicked problem. Solutions to wicked problems have different effects on different people, and their evaluation of the solution varies depending on their experience before, during and after interacting with the solution. Further, the effects of a solution can be quite different from those intended by the solver; the same non-neutrality of technologies described in the chapter “Views of Information and Technology” can be applied to solutions to wicked problems. As a result, there is no single measure of the success of a solution to a wicked problem, and a single measure can return different results when applied to different population or when applied to the same population at different times. For educators who design a school to (for example) prepare students for college, the program may be evaluated differently by those who enroll in different colleges or those who study different majors in college. Further, the program may be evaluated positively by students when they perform well on college entrance exams, but negatively after they enroll in college.

Every Solution Matters

Tame problems are generally solved through processes that are tolerant of error: Natural scientists propose hypotheses and through the collective process of scientific inquiry inaccurate hypotheses are separated from accurate ones. Engineers typically build and test models and prototypes before any solution is fully deployed. Through this process, errors and flaws in the design are identified and mitigated prior to exposing the design to use. (Examples of engineering failures are well-known, but given the number of engineered products that are used on a daily basis, the number of failures with catastrophic consequences is remarkably low.) Planners solving wicked problems do not have the option to test their solutions in the manner scientists and engineers do.

Wicked problems are social, so when planners design and implement a solution, it affects the experience of individuals and groups exposed to it in permanent ways. For example, educators who implement an instructional plan in mathematics will affect how students understand mathematics. The instruction may improve their ability to solve problems and learn more advanced mathematics later, or the instruction may inhibit their abilities.

Each Wicked Problem is Unique

When solving tame problems, scientists and engineers go to great lengths to identify the conditions under which they made their observations. When evaluating their work, they account for all factors that may not have been reliably controlled and they assess the transferability of the conclusions they draw to other situations. Scientists also interpret the results of their experiments and observations in light of what was known and unknown about the factors that appear to have influenced their observations. When recommending further research, scientists recommend changes that may improve their methods and the observations they made. All of this is done so that other researchers can recreate conditions and confirm observations. When dealing with wicked problems, the factors affecting a situation are too complex to be known reliably and to be controlled. As a result, attempts to transfer solutions often do not produce similar results as important conditions of the original setting are not replicated in the new settings.

For educators, this character of wicked problems suggests that the methods designed for one classroom and evaluated as effective in that classroom may not produce the same results in another classroom. Factors ranging from the previous experiences of the students to the skills of the teacher to the sociocultural context of the communities in which the school is situated all influence how a curriculum and instruction problem and solution is understood and instantiated in a community. Those same factors influence how the solutions are evaluated by the populations served by the school.

Each Wicked Solution is Unique

Because each problem is unique, so is each solution unique. While it may not be possible to transfer a wicked solution developed in one setting to another, it is possible to mix and remix solutions designed in one setting when designing solutions for another. This characteristic of wicked problems necessitates planners demonstrate innovative and creative thinking when designing solutions; simply following the recipe developed by another planner is a dubious strategy for planners of solutions to wicked problems.

Further, there are no rules for guiding or limiting planners as they design solutions to a wicked problem; any solution is permissible. Rittel and Webber use chess as an example of a tame problem that has well-known rules limiting and defining solutions to the problem of taking the opponent’s king. Such solution-limiting rules do not exist for wicked problems.

Interconnected Problems

Wicked problems exist at multiple levels, and each wicked problem is the cause of another (or several other) wicked problems and each wicked problem is the effect of other wicked problems. For those who seek to understand a wicked problem and create and implement a solution to it, this complicates the task by making it impossible to identify and control all relevant factors. Even if key causes can be identified, those causes may arise from a source over which the problem solver has no control. Further, a sufficient solution that can be implemented may cause unacceptable problems to arise for others, so the solution must be abandoned. As a result, even the best made plans may be prevented from working by a cause that cannot be controlled or by an effect judged to be unacceptable by others who are politically more powerful.

Multiple Valid Perceptions

Wicked problems (and their solutions) can be accurately described from many perspectives. During one particularly contentious school budget debate, I observed the schools within a district being described as “a baby-sitting service,” “a burden on the taxpayer,” and conversely “the best investment we can make in our future,” and “a priceless source of pride for our community” by different members of an audience comprised of citizens from the same community. Even excluding these opinions which were expressed in a politically and emotionally-charged situation, these comments are evidence that schools are technologies designed to meet a variety of human goals and that different people’s perceptions of the same solution can vary widely, and each interpretation is accurate. Objective measures of wicked problems and solutions do not exist; the subjective experience and the interpretation of problems and solutions in light of that experience is the only method of evaluating wicked problems and solutions that is reasonable.

No Right to be Wrong

Wicked problem solvers have no right to be wrong. Given the social nature of wicked problems and the permanent effect that solutions can have on participants, it is essential that planners of wicked problems design the best possible solution. This final character of wicked problems can be problematic for many. It does capture the responsibility of each problem solver to take action on the best possible information and to fully implement the best solution possible.

References

Rittel, Horst, and Melvin Webber. 1973. “Dilemmas in a General Theory of Planning.” Policy Sciences 4(2): 155-169. doi:10.1007/BF01405730.