Journal Information

Article Information

Systemic planning: Dealing with complexity by a wider approach to planning


On the basis of the author’s latest book, Systemic Planning, this paper addresses systems thinking and complexity in the context of planning. Specifically, renewal of planning thinking on this background is set out as so-called systemic planning (SP). The principal concern of SP is to provide principles and methodology that can be helpful for planning under circumstances characterized by complexity and uncertainty. It is argued that compared to conventional planning—referred to as systematic planning—there is a need for a wider, more systemic approach to planning that is better suited to current real-world planning problems, often characterized by complex issues.


On the basis of the author’s latest book, Systemic Planning (Leleur, 2005), this paper presents an overview of the ideas behind systemic planning (SP). The principal concern of SP is to set out—by making use of systems science—some principles and methodology that can be helpful for planning under circumstances characterized by complexity and uncertainty. The paper is arranged as follows: After an introductory overview, the second section gives the main ideas behind the SP approach, with an outline of different categories of change processes and of three principal types of complexities that can make a conventional planning approach inappropriate. Then systems science is used to formulate the basic SP principles. In this respect, the paradigms of simplicity and complexity thinking as set out by the French philosopher of science Edgar Morin are called on, and a generic, methodological framework for SP is formulated.

The following section presents SP as an appropriate multimethodological approach characterized, among other things, by combining “soft” and “hard” operations research (OR) methods to set up an exploration and learning cycle that can guide planners and decision makers in a real case. A comprehensive set of relevant tools is presented, and it is discussed how this SP toolbox can be made use of together with the general principles of SP. In the subsequent section an example is dealt with. This concerns the complex public decision whether or not to build the Øresund Fixed Link at a cost of €3.2 billion for road and rail traffic between Copenhagen and Malmoe. This ex-post study—carried out four years after the opening of the link in 2000—has been found relevant to demonstrate a planning case where the planning environment can be described as complex and uncertain. The case is treated with the emphasis on presenting the applicability of SP in general; that is, as being of interest also for other societal sectors than transport and infrastructure. The final section gives some findings and a perspective on further work.

Simplicity, complexity, and systemic decision making

Based on work carried out by the British organizations and complexity researcher Ralph Stacey, change processes in general can be categorized as closed change, contained change, and open-ended change (Stacey, 1993; Leleur, 2005: 16—17). These processes, described below, will influence the kind of planning approach that should be made use of when dealing with planning tasks relating to specific change processes.

  • Closed change: The key features of closed change are unambiguous problems, opportunities, and issues; clear connections between cause and effect; and the possibility of accurately forecasting the consequences of change. Faced with such change, people tend to behave in easily understandable ways. The decision maker can use rational planning techniques and the processes of control are formal, analytical and quantitative. There is a clear purpose with clear preferences, and alternative ways of achieving the purpose are known.

  • Contained change: The key features of contained change derive from those change situations where it is possible to make probabilistic forecasts based on actions taken now and their most likely consequences. This is made possible because the consequences appear to some degree as repetitions of what has happened in the past or relate to large numbers of essentially the same event. As a planner looks into the future, accurately predictable closed change declines in relative importance, while less reliably predictable contained change increases in relative importance.

  • Open-ended change: Planning and control in open-ended situations in practice means something completely different from what it means in closed and contained situations. In such situations, the future consequences are to some extent unknown and forecasting is made difficult due to the sometimes ambiguous purposes and equivocal preferences of the planning agents involved. The whole situation may be ill structured and accompanied by inadequate information that is more or less subjective and conditioned by personal ambitions, beliefs, and values. There can be considerable problems interpreting data and applying statistical techniques in uniquely uncertain conditions, for which reason forecasting and simulation become problematic.

Planning methodology and problem solving can in many cases be reasonably well specified and developed so they can facilitate planning-based decision making relating to closed and contained change, whereas open-ended change remains a challenge for several reasons. One basic consideration in this respect is that the uncertainties involved in complex decision making are principally of a generic type that cannot be satisfactorily dealt with by detailing and refining conventional planning methods that work well in situations with closed and contained change. Specifically for complex planning problems the planning environment when approached by conventional planning thinking is seen to “complexify” along three dimensions, which is why the following complexity categories are relevant to deal with:

  • Detail complexity;

  • Dynamic complexity;

  • Preference complexity.

Basically we will associate detail complexity with “means” and dynamic complexity with “path,” whereas preference complexity relates primarily to “ends.”

The application of systems science for improving our problem-solving capabilities holds two promises (Leleur, 2005: 22):

  • By seeing our problem or study object as a system, we may make use of the systems concepts to create a better representation of it and here capture (and model) various interrelations among elements and so on in a more qualified way.

  • By seeing our problem as a system, we may be able to focus less on step-by-step approaches and capture more holistic impressions that can qualify our study.

The first statement above concerns what is sometimes referred to as systems analysis. In an almost generic process we commence by defining our problem and determining the objectives. After this we turn to envisage or model the consequences of various relevant alternatives. Then we appraise the alternatives to make it possible to select the best one. The final step concerns the implementation of this alternative and it may be decided to continue the process by monitoring it (Leleur, 2000: 18). Our ideal in this undertaking is to be rational in our decision making so that the analytical processing of complete information will lead to an optimal result, be it a decision, design, plan, and so on. We will see this as a systematic approach.

The second statement relating to the use of systems science expresses that wholeness matters and it can therefore be seen as a corrective to the first one. With the systematic


The two paradigms of simplicity and complexity

Simplicity ParadigmComplexity Paradigm
The generalThe particular
Linear causalityMulticausality
The automatonTime
approach we proceed in our planning and problem solving by using a step-by-step approach; with the systemic approach we are concerned with holistic views.

No doubt the systematic approach is tied to the rational-analytical thinking well known to people educated, for example, as engineers and economists, whereas the notion of being systemic is more difficult to understand and come to grips with. In this situation the paradigms about simplicity and complexity by Morin become highly relevant, where a paradigm denotes a basic research orientation and pattern. Fundamentally, Morin sees classic scientific explanation as based on a simplicity paradigm. Although he recognizes the strength of the simplicity paradigm in many respects, he also identifies certain limitations to its explanatory models. As physics and cosmological thinking have always been major suppliers of ideas to other branches of science, it stands out that subnuclear physics is the main example that Morin uses to argue the insufficiency of the simplicity paradigm, as this branch of physics cannot satisfactorily explain new so-called


The SP structure as four interrelated modes of exploration and learning. Different methods are indicated to illustrate some possible method choice.

SP Method StructureSystemicSystematic
ScanningExample: Critical systems heuristicsExample: Scenario analysis
AssessmentExample: Futures workshopExample: Multicriteria analysis and simulation
exotic particles, for example. Against this background, Morin proposes a complexity paradigm to reorient and widen our research activities. Specifically, he suggests principles for the complexity paradigm that are complementary to those contained in the simplicity paradigm. Table 1 indicates Morin’s two paradigms by some paired keywords (Morin, 1974, 1986; Leleur, 2005: 24).

As concerns the development of systemic planning, we will see systemic thinking as rooted in the complexity paradigm and systematic thinking as rooted in simplicity thinking. The idea is not to replace systematic thinking with systemic thinking, but to make wider planning possible by applying both. As the conventional planning approach is tied to systematic thinking, we adopt the term systemic for such wider analysis in which we choose to include both systematic and systemic findings and not just the latter. In this way, systemic planning (SP) seeks to generalize the conventional well-known planning approach that in this context can be seen as relating primarily to systematic planning and problem solving.

Another basic complementary relationship behind SP concerns scanning vs. assessment. Dealing with planning and problem solving, exploration and learning will depend on alternating between these two modes; that is, they cannot both be problematized at the same time but will reciprocally influence each other. By cross-referencing the two pairs of complementary relationships we obtain the basic methodological structure behind SP shown in Table 2 (Leleur, 2005: 127):


The SP toolbox made up of various hard and soft OR methods

Current Methods Available for SP—Bold Type Indicates the Methods Made Use of in the Øversund Fixed Link Case
Analytical hierarchy process (AHP)Computer-aided design (CAD)Conflict analysisCost—benefit analysis (CBA) and cost-effectiveness analysis (CEA)Critical systems heuristics (CSH)Critical path method (CPM)Cross-impact analysisDecision analysis (DA) applying SMART and SMARTERDelphi conferencing techniquesEnvironmental impact assessment (EIA)Expert systemsForecastingFutures workshop (FW)Fuzzy set theoryGame theoryGraph theoryInput—output analysisInteractive planning (IP)Intuitive exploration/brainstorming/metaphor and analogy buildingLinear programming techniquesMulticriteria analysis (MCA)Multiple perspectives (MP)Network theoryOptimization theory and heuristicsProgram evaluation and review techniques (PERT)Scenario analysisSensitivity analysisSimulationSoft systems methodology (SSM)Statistics, probability and queuing theoryStrengths, weaknesses, opportunities, and threats analysis (SWOT)Systems dynamicsTotal systems intervention (TSI)

The tools of systemic planning

The SP approach is developed by making use of the generic structure shown in Table 2. Generally this is carried out by applying appropriate OR methods (see Table 3) in a self-organizing process that embeds conventional optimization in a wider process of exploration and learning (Leleur, 2005: 35). The ongoing search—learn—debate process moves on by contrasting and interpreting the different findings and insights. The process aims at converging into a satisfactory end result for the decision makers.

Generally, “hard” OR methods can be seen to provide first-order findings based on calculative rationality, whereas second-order findings (or even higher) are associated with “soft” OR methods—based on communicative rationality—that relate to the so-called subworld created around a complex problem by the various stakeholders and participants in the process (Dreyfus & Dreyfus, 1988: 76; Leleur, 2005: 72—73, 107).

The SP exploration and learning cycle makes it possible to deal with a complex problem in a much more explicit way. The example below describes some findings relating to the Øresund Fixed Link, where the set of four methods in Table 2 was applied (Leleur, 2005: 132—136).

Overview and application example

The Øresund Fixed Link—open since July 2000—can be regarded as one of the most complex transport investment decisions made in Scandinavia (Leleur, et al., 2004; Leleur, 2005: 119—137). The case work demonstrated how the huge amount of information produced in studies and so on over the years could have formed part of an ex-ante examination of how to apply the SP approach. In this respect the case has functioned as a kind of evaluation research methodology laboratory (Leleur, 2005). The idea of the ex-post study, undertaken three to four years after the opening of the fixed link, is to consider and review the impacts and the ex-ante evaluation methodology to examine whether the latter was appropriate. Therefore, the study aims at informing planning and evaluation methodology and possibly updating it. However, the ex-post study cannot give certain results concerning the ex-ante methodology stemming from the beginning of the 1990s, with the decision to implement being taken back in 1994. Development could have been otherwise if, for example, new issues of high relevance of various types had arisen: Danish and Swedish legislation being counterproductive for integration across the Øresund, an oil supply crisis, and so on. However, saying that a narrow cost—benefit analysis for large infrastructure planning is at best insufficient is a generally relevant finding, which is exemplified by the wider approach presented below that makes use of systemic planning (SP) ideas, which are briefly reiterated.

With an emphasis on a search—learn—debate process that develops around contrasting and interpreting the upcoming intermediate findings and insights of the planning problem, a group of four complementary methodologies was selected after some scrutiny; see Table 2. These SP principles for guiding the search—learn—debate process do not draw on an overarching kind of rationality; in fact, applying German sociologist Niklas Luhmann’s view on selection and complex processes, the theme of rationality in the SP process,

“disintegrates into a typology of distinct rationalities, whose relations to one another can no longer be subsumed under the requirements of rationality—in, for example, some sort of ranking.” (Luhmann, 1996: 171)

Then in SP there is no general rationality blueprint when proceeding and doing this or that in the planning process. This is part of the theoretical underpinnings of systemic planning (SP); further detail behind the formulation of SP relating to, among other things, Luhmann’s perception of social systems and his contingency/complexity thinking—perceived as representing third-wave systems science—is given in the book Systemic Planning (Leleur, 2005: 40—48, 83—94).

The OR methods that were made use of in the Øresund Fixed Link (see Table 3) were critical systems heuristics (CSH), scenario analysis (SA), futures workshop (FW), multicriteria analysis (MCA), and simulation (SI). Basic methodology references are Jackson (2000); Midgley (2000); Flood (1999); Drewes, et al. (2004); and Goodwin & Wright (1999).

In brief, CSH mapped decision coalitions ( “players”) and their motives and different responses at certain stages, whereas SA and FW provided a set of interrelating framework and trend scenarios. In that way several future images were constructed and each of these was examined using MCA and SI. These latter methods produced some quantitative expressions that illuminated some aspects of the complex investment project. In addition, the application of MCA and SI also made it relevant to reconsider some of the CSH and FW analyses and results concerning, for example, the integrative role of the new bridge linking not just two major cities across a strait but also two countries and, furthermore, giving all Nordic countries access from Scandinavia to the central part of Europe.

There was general agreement among the participants—representing both researchers and (some of) the identified stakeholders—at a seminar in 2004 where the results were presented that the assessment insights found could not have been achieved by making use of the standard cost—benefit approach that would normally be applied for such a study (Leleur, et al., 2004). In this respect it should be noted that the SP approach has been conducted as a kind of comparative study as, among other things, it has been possible to make comparisons with the actual examination process before the construction work began in 1994. To illustrate the iterative, nonlinear planning process prescribed by SP, Table 4 presents some intermediate findings. As a general characteristic, it can be noted that certain insights gained with one of the applied methods in a particular category trigger new examinations in one or more of the categories. New stakeholder preferences revealed in the systemic assessment category may, for example, lead to new scenarios being relevant in the systematic scanning category and so on. In this way the total process becomes one of exploration and learning.

The outcome of the type of examination outlined above showed that the Øresund Fixed Link was a feasible project from a societal point of view if—as it turned out—different strategic impacts such as European and regional


Intermediate findings and specifications based on SP exploration and learning

Some Intermediate Findings and Specifications that Can Feed Back into the Process
Systemic scanning: Issues of identification and demarcationGeneral concerns:• Øresund region one of several spheres• The meaning of national barriers• Drivers: market, clusters, culture, etc.• Infrastructure and developmentSpecific concerns:• Limitations of cause-effect model• Interpretation of expressed expectationsSystematic scanning: Issues relating to scenariosRegional scenarios:• Economy, regulation, transport, etc.• Local integration vs. nonintegration• Baltic Sea development: trade etc.• Competitive transport developmentEU-wide scenarios:• Economy, regulation, transport• Trends: resources and technology• Trends: modal policies etc.
Systemic assessment: Issues relating to stakeholder preferencesEx-ante:• Local pro-coalition• Local environmental anti-coalition• National interest• International pro-coalitionEx-post:• National interest• Øresund region citizens• Øresund companies• International interestSystematic assessment: Issues relating to multicriteria analysisNarrow feasibility (CBA):• Investment• Time savings• Cost savings• Local emissions and accidentsWider feasibility (MCA):• Network and mobility• Global emissions (CO2)• Employment• Logistics and goods effects
mobility, regional employment, climate effects, and so on were considered and included in the assessment. Such wider impacts are not part of a conventional cost—benefit analysis (CBA), which is why the SP approach is found more suitable considering that the actual Øresund case is much more complex than, for example, ordinary medium-sized Danish highway projects that are recurrently examined by CBA (Leleur, 2000). In the case presented here the focus has been on methodology and process aspects; a full account of specific results and a closer description of details and other study information are available in Leleur, et al. (2004) and—with emphasis on their principal interpretation as relating to systemic planning—Leleur (2005).

Findings and perspective

Below a summary is given of the general findings relating to the formulation of systemic planning principles and methodology (Leleur, 2005: 107—108, 76).

  • The systemic planner needs to assess that the actual planning task is really suitable for a systemic approach. This means that simple cause—effect relationships cannot be obtained and that the characterizing features of the problem are complex to some extent along the dimensions of means, path, and ends, expressing varying degrees of detail, dynamic, and preference complexity. Such a problem has earlier been categorized as open ended. The planner ends up perceiving that the planning task is “beyond” a conventional approach and decides to continue with a systemic one.

  • Against this background it becomes relevant to start building knowledge about the concrete “subworld” that will unfold in the course of events and to set up a planning team that can be expected to handle this. Some kind of framework to assist in structuring a search—learn—debate process could be relevant and could be developed with different purposes in mind. The generic structure presented in Table 2 makes it possible in a kind of nonlinear “iteration” to contrast subsequent findings and insights based on cross-referencing methodologically the pair scanning and assessment with systemic and systematic, “filled in” with suitable planning methods for the problem. Findings and insights can be highly different in nature: Some may be derived from mathematical modeling as optimization relating to “what-if” considerations; others may concern issues raised by different power coalitions; and still other issues will have to do with implementation and so on. The planning unfolds by making use of findings and insights, be they of first, second, or even higher order. The latter may in some cases be expected if, for example, the “wickedness” of the problem is due to politics and power issues.

  • The systemic approach can be built on a major methodological base (see Table 3), and one should expect that in real-life studies highly different methods will sometimes be applied in combination. In the case presented above, among other things, the “soft” critical systems heuristics (CSH) approach is applied together with the “hard” multicriteria methodology. Needless to say, the choice of specific methods is important and ought to be driven by the actual planning problem.

No doubt a systemic planning process should be expected to be demanding in skills, resources, and so on. For this reason alone, undertaking a systemic planning study should be contemplated ahead of its commencement. A general characterization as demanding, compared to conventional planning tasks where suitable planning routines and methods are available, follows from the successive establishment of the necessary subworld around it; from applying and combining different methods it needs to be seen how they perform and provide formation of meaning and understanding in the particular context consisting of the planning problem and its environment, stakeholders, planners, concrete interpretations, narratives, suggestions, paradoxes, and so on.

If a kind of epistemology should be sketched on the background of generalizing planning thinking, we have to move from a hierarchical, well-ordered input—output process toward a wider process that also contains what we may see as networks or heterarchies. What characterizes a heterarchy is that there is no single “monitor” and no single “highest level.” The instrumental reason for conventional planning has, furthermore, been embedded in a wider communicative rationality, which can be seen as a move from a foundational hypothetic-deductive orientation toward a non-foundational perception; that is, the communications-based agreement.

In Table 5 a new outlook for planning as systemic planning is presented in an overview by comparing issues that characterize it—ranging from problem type to the view above on epistemology—with those of conventional planning. In this respect it can be noted that a consequence of applying the thinking of Morin, Luhmann, and Habermas leads to biperspectivism based on both simplicity and complexity orientations as concerns the paradigm of thinking, and to a nonfoundational, so-called sympoietic orientation based on communicative action as concerns epistemology. The concept of sympoiesis—inspired by seeing autopoietic systems forming co-evolutionary networks—has been introduced with the formulation of systemic planning to indicate the phenomenon and outcome of reciprocal relationships between individual entities and ensembles (Leleur, 2005: 97). Whereas Habermas is well established in current systems science theory and practice—see for example Midgley (2000) and Jackson (2000)—the reception and application of Morin and Luhmann are less so. To point to this situation—and to the fact that a


Comparison of conventional planning with a new outlook for planning

Issues of CharacterizationConventional PlanningA New Outlook for Planning
Problem typeSimple, defined as noncomplexComplex, not just complicated
Paradigm of thinkingSimplicity representing monoperspectivismBoth simplicity and complexity representing biperspectivism
Rationale of planningMainly proactive and optimizing, with emphasis on modelsMainly enabling and mediating, with emphasis also on learning
Professions involvedEngineers, economists, geographers, etc.Also sociologists, political scientists, etc.
Planning practiceA linear process of activities ( “tasks”), dominated by first-order findings that can be combined to produce a planA nonlinear (self-organizing, autocatalytic) process of activities ( “events”), both first- and second-order findings
EpistemologyFoundational, hypothetico-deductive ( “hierarchical input—output”), based on instrumental reasonNonfoundational, sympoietic ( “heterarchical networks”), based on communicative action
complexity orientation could provide managers and planners with other insights than those associated with metaphors derived from deterministic, chaotic models ( “chaos management”; see Stacey, 1993)—the conceptual underpinnings of systemic planning relating especially to the work of Luhmann are considered further in Leleur (2006).

Whether systemic planning is warranted or not depends on the actual circumstances and our interpretation of these. Clearly, there have been cases where conventional planning has failed as conditions were not right for a systematic approach; and clearly also there can be no guarantees that widening planning into what we have called systemic planning, comprising both “hard” and “soft” methodologies, will be successful. The applications so far, however, are promising. Therefore we argue that SP holds potential as guidance for planning in a context of open-ended, complex problems necessitating proactive decision making. Increasing complexity in society in general and in the professional spheres of administration and business more specifically, combined with the flexibility and adaptability of SP, make it relevant to pursue a further development of the current SP principles and methodology.



Dreyfus, H. L. and Dreyfus, S. E. (1988). Mind Over Machine, ISBN 9780029080610.


Drewes Nielsen, L., Homann Jespersen, P. and Hartmann-Pedersen, K. (2004). “Future workshops on freight transport: A methodology for actor involvement,” World Transport Policy and Practice, ISSN 1352-7614. 10(3): 36-41.


Flood, R. L. (1999). Rethinking the Fifth Discipline: Learning within the Unknowable, ISBN 9780415185301.


Goodwin, P. and Wright, G. (1999). Decision Analysis for Management Judgment, 2nd edn, ISBN 9780470861080 (2004).


Jackson, M. C. (2000). Systems Approaches to Management, ISBN 9780306465062 (2005).


Leleur, S. (2000). Road Infrastructure Planning: A Decision-Oriented Approach, 2nd edn, ISBN 9788750208242.


Leleur, S. (2005). Systemic Planning: Principles and Methodology for Planning in a Complex World, ISBN 9788750209652. See www.systemicplan-


Leleur, S. (2006). “Systems science and complexity: Comparing a complexity-based orientation with other current research orientations,” UK Systems Society Annual Conference 2006, September.


Leleur, S., Holvad, T., Jensen, A. V. and Salling, K. B. (2004). Development of the CLG-DSS Evaluation Model, ISBN 9788791639012.


Luhmann, N. (1996). Social Systems, ISBN 9780804726252.


Midgley, G. (2000). Systemic Intervention: Philosophy, Methodology, and Practice, ISBN 9780306464881.


Morin, E. (1974). “Complexity,” International Social Science Journal, ISSN 0020-8701, 36(4): 555-582.


Morin, E. (1986). “Kompleksitetens bud,” Paradigma, ISSN 1600-0285, 1(1): 18-20.


Stacey, R. D. (1993). The Chaos Frontier: Creative Strategic Control for Business, ISBN 9780750609500.

Article Information (continued)

This display is generated from NISO JATS XML with jats-html.xsl. The XSLT engine is Microsoft.