Rural economic development, wicked problems and complex adaptive systems
(Don’t let that title keep you from reading)
Co-Director, Indiana Business Research Center, Indiana University Kelley School of Business
At a recent meeting, attendees discussed what I would call the grim future prospects of many rural counties in Indiana and what Kerry Thompson, the executive director of IU’s Center for Rural Engagement, called the need for a “pathway to hope.” While speaking specifically about the opioid crisis in rural Indiana in terms of the need for a pathway to hope, she described the multi-dimensional and multi-layered interdependent characteristics, forces and needs to build that pathway to hope: economic growth, available jobs, good educational resources, access to healthy food and water, access to health services, transportation, etc. One cannot speak of all the dimensions in one breath.
What she described is a wicked problem. Not wicked in an evil sense, but wicked in a big, hairy, complicated sense. A problem that doesn’t have a start, end or well-defined middle.
Thinking about any and all the wicked challenges associated with rural economic development is the motivation for this article. Whether the opioid crisis, or fostering job growth, or stemming job hemorrhaging, or slowing out-migration, or making rural places better connected globally via broadband, or preparing rural students for the workforce of the future, these challenges are all interconnected—and wicked.
The usual progression in describing the definition and nature of wicked problems is to start with complex adaptive systems as a research domain, analytical framework and theory. But we are going to start with wicked problems and save the arguably more off-putting discussion on a theory of regional economic development toward the end.
Wicked problems defined
Tom Ritchey (2013)—and most of the following is pulled from his paper—well-advisedly begins describing wicked problems by first examining the traits of a “tame” problem:
- It has a well-defined and stable problem statement.
- It has a definite stopping point and solution. We know when we are done.
- The solution can be objectively evaluated as being right or wrong.
- The problem belongs to a class or category of similar problems with similar solutions.
- The solutions can be tried, tested and either kept or abandoned.
Wicked problems are just the opposite. They are ill-defined, ambiguous, and often aligned with moral, political and professional self-interests. Stakeholders subjectively define the problem from their point of view and self-interest. As a result, it is difficult (if not impossible) to find consensus about what the root problem is, let alone how to solve it. To make matters even worse, “wicked problems won’t keep still: They are sets of complex, interacting issues evolving in a dynamic social context” (p. 2). Wicked problems keep changing—and often they change as a result of an intervention to treat an earlier problem.
Jeff Conklin (2001) offers a cautionary tale, namely that organizations typically either try to study the problem or tame the problem. “While studying a novel and complex problem is natural and important, it is an approach that will run out of gas quickly if the problem is wicked. Pure study amounts to procrastination, because little can be learned about a wicked problem by objective data gathering and analysis. Wicked problems demand an opportunity-driven approach; they require making decisions, doing experiments, launching pilot programs, testing prototypes, and so on” (p. 10).
It is as if Conklin had just read Who Moved My Cheese. The first step to finding new cheese is to take a step to find new cheese, not think about and study how to find new cheese.
Conklin also cautions against trying to tame a wicked problem because “while appealing in the short run, [it] fails in the long run. The wicked problem simply reasserts itself, perhaps in a different guise, as if nothing had been done. Or, worse, sometimes the tame solution exacerbates the problem” (p. 12).
Ritchey (2013) articulates a concern that “wicked problem” (WP) has become a buzzword and, as a result, has lost much of its definitional punch.
With wicked problems, it’s difficult to know where to start.
So, what is the problem that the term “wicked problem” addresses? The common sense approach to WPs is fairly straight forward: WPs are about people—the most “complex adaptive systems” that we know of. They are subjective problems. Everything that has to do with people and society is ultimately subjective. Above all, WPs are about people as stakeholders: competing and cooperating, vying for position, willing to reflect, and to change their positions on the basis of this self-reflection. This is why such problems do not have stable problem formulations; do not have predefined solution concepts; and why their course of development cannot be predicted. (p.3)
There are 10 generally accepted criteria for wicked problems according to Rittel and Webber (1973):
1. There is no definite formulation.
“The information needed to understand the problem depends upon one’s idea for solving it. This is to say: in order to describe a wicked problem in sufficient detail, one has to develop an exhaustive inventory for all the conceivable solutions ahead of time.” This sounds much like an earlier IBR article about survivorship bias. Namely, how one defines a problem motivates the data collection. This criterion seems to suggest brainstorming on the possible solutions first and then capture data to refute or support that solution.
2. There are no stopping rules.
With wicked problems, you never come to a final, complete or fully correct solution—since you have no objective criteria to make such a judgement. The problem is continually evolving and mutating. You stop when you run out of resources or when a result is deemed “good enough.”
3. Solutions are not binary (true or false), but better or worse.
The criteria for judging the validity of a “solution” to a wicked problem are strongly stakeholder dependent. Different stakeholders see different solutions as simply better or worse.
4. There is no ultimate test of a solution.
Because wicked problems are constantly changing and evolving, the outcome of any solution is time dependent. What is a success tomorrow may be a catastrophe next year.
5. Every solution is a one-off.
There is no opportunity to learn by trial and error. Because the system is irreversibly changed with each solution, every attempt counts significantly.
6. There is no exhaustive and describable set of potential solutions.
There are no criteria to show that all the possible solutions have been identified.
7. Every wicked problem is unique.
The art of dealing with wicked problems is not knowing too early which type of solution to apply. Stay in the mess as long as you can. There are no classes or categories of wicked problems and, therefore, there are no common principles to apply to find a solution that can be developed to fit all members of that class. This point is particularly salient for those working in regional development. After an exhausting (but admittedly not exhaustive) analysis of some 1,200 counties in the U.S., the various statistical analyses did not yield any sort of consistent narrative across any of the multitude of dimensions examined. In other words, there were no neat categories to place regions based on size, industry composition, proximity to metro areas, educational attainment, average age of the population, and, most significantly, based on performance measures like employment growth, productivity or income equality. Each of those counties were analytically unique and, based on the evidence, their wicked problems were also unique to their particular circumstance.
8. Every wicked problem can be considered to be a symptom of another wicked problem.
There is mutual and circular causality. Finding the right level of abstraction and analytical framework is itself a complex judgment.
9. Causes can be explained in many ways.
There is no definitive and “correct” explanation, or combination of explanations.
10. The planner and problem solver has no right to be wrong.
In the hard sciences, the researcher may make a hypothesis that can be refuted. Such is the nature of science and gaining knowledge. A bad hypothesis is just wrong. The problem solver can’t be wrong, however. The aim is not to find the truth, but to improve some characteristic of the real world. Problem solvers are liable for the consequences of their solutions and actions.
Ritchey (2013) proposes a group-facilitated, computer-aided General Morphological Analysis (GMA) and aligns this approach with the above criteria. (In the interest of space, we recommend that the interested reader read his full 2013 paper and even dig deeper.) That said, there are two particularly important criteria that apply to the question about rural economic development: generating potential problem solution concepts (#1) and remaining in the mess of an unsolved problem as long as possible (#7). The two criteria are related. Criterion #1 suggests a long and possibly contentious process—see also criterion 3—because the goal is to generate as many potential solutions from stakeholders and to encourage stakeholders to better understand the problems, solutions and the consequences of the solutions. By better understanding the solutions, the stakeholders and facilitators will have a better grasp of the actual problem. This means keeping the options open as long as possible—criterion #7—to discover as many relationships in the problem landscape as possible.
Given the “wicked” nature of rural economic development and the general approach of GMA, one can see the possibilities of applying this framework to the study of rural development.
The framing of problems that were not amenable to ready solutions was an offshoot of the study of complex adaptive systems (CAS). We now turn to briefly describe complex adaptive systems and suggest how it may apply to regional economic development.
Complex adaptive systems
A system can be defined as a collection of inter-related components that work together toward some objective. Systems can be very large—macro—or smaller in sub-systems. There can be systems in between the macro and micro as well, and all these systems at different scales/size also interact. There are four attributes to complex systems. They are (1) self-organizing, (2) are non-linear and (3) dynamic, and (4) exhibit emergence. The first three terms are probably recognizable, but “emergence” may not be. Emergence can be thought of as the “behavior” of a system or collection of systems that cannot be accounted for or explained by the “agent” level or individual actors alone.
Vertical to ground dampers installed on the London Millennium Bridge
Photo by Dave Farrance - Own work, CC BY-SA 3.0,
The swaying of the London Millennium Bridge is an example of system behavior in contrast to agent actions. When the bridge first opened, pedestrians—the agents—walked across the bridge randomly and, as typical bi-peds, attempted to keep their normal balance. But this caused the bridge to sway and as the sway increased, the pedestrian attempts to stabilize themselves in opposition to the sway exacerbated the sway. (The bridge was quickly closed and was retrofitted with a dampening system.)
The agent concept is important. Agents in the regional economy are dissimilar individuals who may or may not organize in some group or structure and who can respond to incentives or environmental forces. Complex systems theory is a framework that explains how rules govern agent interaction within a system. Those interactions result in emergence—system-level behaviors. These system-level behaviors change with time and show non-linear patterns.
Complex adaptive systems change and restructure themselves based on changes in their environment. https://t.co/wQcsQYpgAs— IBRC (@IUibrc) September 20, 2019
An adaptive system and an interdependent system-of-systems change due to environmental forces. This change enables the system to persist. Adaptive systems alter their constituent components or internal properties, or interact with and alter their environment. While an adaptive system is complex, a complex system is not necessarily adaptive. That is, a complex system may not undergo change due to shocks. On the other hand, it may change in a manner that puts into question its ability to persist—maladaptation. Also, whether a system is appraised as having adapted depends on the scale/size of the component or structure being evaluated. For example, a person’s cells and organs may adapt to living at high altitude by changing their structure and metabolism, but the person as a macro system (embedded in an environment) may also adapt by using supplemental oxygen. Continuing with this example: Given the many levels of interacting systems and structures, how does one measure whether, or the degree to which, a system as a whole has adapted? A common method for altitude adaptation would be to test one’s hemoglobin. The simple metric may not tell the whole adaptation story, but it is a solid indicator of adaptation.
In summary, complex systems theory is a scientific framework that explains how system rules are manifest in emergence—structural change. It is this emergence that may help guide theories and practice of regional, and especially rural, economic development.1
Complex adaptive systems and economic development
This article began with describing wicked problems. If wicked problems did not exist in regional economies, either in Indiana or across the nation, there would not be a field or discipline of regional economic development. We began with how one might apply some of the principles of wicked problems to finding solutions. That was followed by an explanation of CAS and a brief overview of the theory of complex systems. That is, complex systems theory is a scientific framework that explains how system rules are manifest in emergence—structural change. It is this emergence that may help guide theories and practice of regional, and especially rural, economic development.
Economics is a theory-heavy discipline. It has a long and exemplary history of theoretical development that borrowed, if not outright stole, the principles of physics in the 19th century. For generations, economists have organized their work around the notion of equilibrium, for example, where the supply curve intersects the demand curve and where all other forces are abstracted away. Economics has developed theories and models that have well-explained aspects of economic outcomes. In more recent decades, there has been movement away from that assumption and/or the teleology of equilibrium, for example, embracing multiple equilibrium states. For over 30 years, many preeminent economists and complexity scientists have gathered to find the intersections between complex adaptive systems and economics (for example, see Blume & Durlauf, 2005).
As both complexity science and economics develop, whether in partnership or separately, it is likely that some tenets of these disciplines will become redundant, while other principles will be adopted in order to build out more a robust and comprehensive understanding of the systems of interest. In the case here, the systems of economic relationships and phenomena occur in geographic space—regional economics. Progress will likely come in fits and spurts, as there may be conflicting agenda among stakeholders in the disciplines. Data science is sometimes criticized by economists as a “theory-free zone”—all data and no theoretical guideposts. Theoretical convergence will not be easy and may be a little unnerving.
It so happens that theories and data concepts tend to grow together. Testing a theory requires data. Data have typically needed theory to categorize and define what data to collect. In just about every science textbook, there is an introduction chapter showing the self-reinforcing circle of theory and empirical testing. Space and time would not allow a full-throated attempt to build a CAS-friendly regional economic development theory, so one will propose one important attribute of CAS theory and how we may proceed to conceptualize and measure it for regional economic development.
Emergence is an attribute of CAS and views agent, organizational and social behavior as a continuous social re-construction. Put another way, adaptation or transformation over time is a hallmark of emergent systems, akin to evolutionary systems. Rather than fixed structures, there are changing structures. As Heraclitus is often quoted: “You cannot step twice into the same river.” What we need then are markers or signals of change.
For the rural engagement application on the Indiana regional, county or town scale, stakeholders and facilitators would define markers or signals of positive adaptation/transformation associated with the particular wicked problem, as well as the potential solutions.
For regional economic development, a sign that there has been emergence, or adaptive change, is the re-balancing of industry employment in a county or region. Those changes in industrial structure are evidence of adaptation. One would expect that, in a complex adaptive region, some industries wax while others wane. Rebalancing, in contrast to a large share of industries shedding jobs without other industries absorbing them, is a diagnostic metric for adaptation. While this concept and data is not the product of a GMA conducted county by county, it should serve to provide a first-order approximation of adaptation in a region. Exactly what that measure is will be the topic of another article in the not-too-distant future.
This article is ambitious. It had several objectives: (1) explain complex adaptive systems and wicked problems; (2) suggest that the general structural change analysis—GMA—may be a better way, if not the only way, to solve the wicked problems of rural places; (3) explain the tensions and potential of melding economic and CAS theory; and (4) propose a first cut at a concept (and eventually a metric) to identify transformation in a regional economy.
One hopes that the suggestion of a simple metric to gauge regional structural change will provide something of an impetus to do thought experiments on solutions to wicked problems related to rural and regional economic development, as well as considerations about what geographic boundaries are the most relevant and how to measure adaptation and change. Writing theory pieces on which framework, evolutionary or ecological, is better to understand resilience and adaptation only provides half an answer. Theory, when tied to operationalized concepts and data, will advance the study of regional economic development more quickly than thought pieces alone. Regional economic development is a discipline driven more by practical action than abstract theorizing.
One also hopes that, as a result of understanding CAS, people will understand why no one has a ready solution for the wicked problems of rural places.
This article was prepared by Timothy Slaper at the Indiana Business Research Center using Federal funds awarded to the Trustees of Indiana University and as a sub-component under award number ED17HDQ3120040 from the U.S. Economic Development Administration, U.S. Department of Commerce. The statements, findings, conclusions, and recommendations are those of the author(s) and do not necessarily reflect the views of the Economic Development Administration or the U.S. Department of Commerce.
- Please note that complexity science is not a single body of theory. It is a collection of many fields from natural sciences to social sciences to medical sciences and network science. It is more an analytical frame to place over a body of inquiry—an academic discipline, for example—to help find answers for as yet still open questions.
- Blume, L. E., & Durlauf, S. N. (Eds.). (2005). The economy as an evolving complex system, III: Current perspectives and future directions. Oxford University Press.
- Conklin, J. (2001). “Wicked Problems and Social Complexity.” CogNexus Institute.
- Ritchey, T. (2013). Wicked problems: Modelling social messes with morphological analysis. Acta Morphologica Generalis, 2(1), 1-8.
- Rittel, H., & Webber, M. (1973). “Dilemmas in a general theory of planning.” Policy Sciences, 4, 55-169.