Learning from Other Places and Their Plans: Comparative Learning in and for Planning Systems

In this thematic issue we pursue the idea that comparative studies of planning systems are utterly useful for gaining a deeper understanding of learning processes and learning capacity in spatial planning systems. In contemporary planning systems the pressures towards learning and continuous self-transformation are high. On the one hand more and more planning is needed in terms of integration of expertise, policy, local knowledge, and response to long term environmental challenges, while on the other hand the value of planning systems is increasingly questioned and many places witness an erosion of planning institutions. The issue brings together a diversity of contributions that explore different forms of comparative learning and their value for any attempt at reorganization, adaptation and improvement of planning systems.


Introduction: Tough Learning in Policy and Planning
This thematic issue aims to enhance the understanding of comparative learning in spatial planning. Comparison can be useful within planning practice, and it can be useful in the scholarship of planning, whether it is in the discipline of planning, or carried out in other disciplines. Some of that scholarship can then benefit planning practice.
In their seminal essay on learning in policy, Bennett and Howlett (1992) present learning as diverse, sometimes aiming at policy change, at changing policy tools, rethinking organizations, or altering discourses. They ar-gue that different forms of learning, by different actors, in different networks, and at different levels, take place with different aims. They then invoke different logics of learning, which will resist integration into a neat typology under one concept of policy learning (although they present one themselves). In other words, 'learning' may be very different things, which might have to lead to formation of new concepts before jumping to too many conclusions regarding learning. Bennett and Howlett (1992) point out that, if we are talking about learning at levels higher than the individual, the only clear outward signal of learning (besides self-reporting) is some form of change in behavior, which then has to be ascribed to learning. They usefully remind us that other paradigms of understanding policy change have not lost their value: Competition and conflict are still useful to grasp much of what is going in policy and administration. The parallel revival of Foucauldian studies of planning and policy (see e.g., Hillier, 2002) tends to find much value in that old alternative, while looking much more skeptically at learning as a process that can be used and abused and which can go recognized and unrecognized (Burchell, Gordon, & Miller, 1991). Hood and Peters (2004) further elucidate that the evidence prompting organizations and other actors in policy to learn is never neutral and entirely transparent, while, crucially, they bring up that managers of public organizations often have structurally little motivation, interest, and the possibility to change; and whatever change does take place might derive from 'learning' much outside transparent arena's supposedly devoted to learning. Alvesson and Spicer (2012) colorfully speak of "stupiditybased organizations" and develop a rather persuasive theory of "functional stupidity" where, at the level of organizations, what is learned is often not to think, not to argue, not to question, not to deviate, and not to actually deliberate. For policy learning, for learning in governance systems including a variety of governmental and non-governmental actors, there is the additional issue of complexity in interaction and complexity in implementation (Klijn & Koppenjan, 2015). This includes limits to the steering of the web of actors after learning, in case learning leads to policy change. 'Muddling through' (Lindblom, 1959) is still on the menu (e.g., Marsden, Ferreira, Bache, Flinders, & Bartle, 2014), and the more intricate the governance configuration, the harder learning and implementing change. Organization studies, in sync with systems theory (e.g., Seidl, 2016), would further emphasize the limits of steering as emanating from limits of observation, with organizations never fully able to reconstruct each other's logic; but also management, never entirely capable of steering an organization, because of an internal opacity which is partly deliberate.

Back to Square One? Not Really, and Certainly Not in Planning
The scientific literature on learning is characterized by revivals of modernist analysis and revivals of their postmodernist critiques (including a critique of jumping to normative conclusions). On the one hand, learning is often promoted in a normative, teleological, unproblematic manner, as an extension of promotion of good governance, deliberation, of evidence-based policies. While on the other hand, learning might appear as eminently abusable, a product of questionable power relations and hidden assumptions, as performance of management and captured by ideology. In our view, the middle ground between these two perspectives has gained much less attention. Yet, such middle ground does exist in practice, e.g., in the practice of planning. Spatial planning, urban planning, regional planning, land use planning, urban and landscape design, or whatever name one might prefer for the expertise on the coordination of spatial organization (Van Assche, Beunen, Duineveld, & de Jong, 2013), was from the start oriented towards learning. Theorists and practitioners alike looked at older and other forms of organization, at model cities and situations they wanted to avoid (industrial slums, revolutions), or at the fate of administrative and creative experiments going on elsewhere. The practice and the discipline had to legitimize themselves through continuous reference to other places and ideas, and it had to adapt itself in series of successes and failures, with political overlords quick to point out what counted as a failure (Sandercock, 1990).
In this thematic issue, we want to start a new conversation about learning in planning and policy, and about learning from systems that transcend routine distinctions between overly positive and hypercritical approaches. Indeed, if we want to learn from other planning systems, then it is essential to map out how we can learn and adapt (Smith & Stirling, 2010).

Creative Comparison and Assessment for Reinvention
In their framing article, Van Assche, Beunen, and Verweij (2020) start the conversation by placing comparative learning in the context of systems of planning that are embedded in systems of governance (cf. Van Assche, Beunen, & Duineveld, 2014). Comparative learning is also situated in a context in which it can interact with other modes of learning: learning from experts (inside and outside the system), learning from the past, i.e., self-reflection and self-analysis, and learning as building new insights through discussion (dialectical learning). They present reflexivity and its cultivation as a precondition for learning from the past, and for the other forms of learning, while they see comparative learning as ultimately and ideally serving dialectic learning. This is the case because simple 'input' of ideas from elsewhere would likely fail, because of a lack of contextualization. Furthermore, dialectical learning is needed for comparative learning to reach its potential to produce something new, something able to capture opportunities for coordination and value creation in the receiving spatial context. Hence, the authors advocate for 'creative comparison.' D'hondt, Van Assche, and Wind (2020) take on the major challenge of reinventing planning systems across the world, for which, they argue, the need is pressing. Many countries do not have a functioning planning system, or they are saddled with colonial legacies which create new inequalities. In other places, planning systems are structurally hampered by their original problem focus and ideological assumptions. The authors argue for comprehensive forms of assessment. Assessment has to be context-sensitive to enhance context-sensitive reforms of planning, and this means they have to be largely self-assessment and strongly participatory. Restructuring planning systems is thus understood as a learning process, where learning from other places, from the past and from discussion, can easily find a place. Comparative learning plays a different, more indirect role here, as experiences across the world have underpinned recommendations for planning reform coming from UN Habitat and other international organizations. 'Best practices' might not be easily replicable, but more general principles for planning can be formulated, based on the bad experiences in many countries with particular models of planning and planning reform, on success stories where a context-sensitive explanation is available, and based on the shared goals of democracy, participation, sustainability, economic development, and stability.

Learning and Comparison in High Complexity Environments
Willems, Molenveld, Voorberg, and Brinkman (2020) focus on complex projects and associated learning processes and conduct a comparative study of nine European cities aiming to develop new green infrastructure with an eye on climate-proofing the urban environment. They studied different tools and models of community involvement, in the understanding that a more participatory approach to such projects was the only way to make them possible, and to encourage learning for adaptation, to the governance context, and the context of changing climates. The authors observed that more ambitious authorities developed new instruments for participation (living labs, project organizations, new departments), yet the relations between such institutional experiments and the organizations they were supposed to coordinate could be complex and disappointing. The rest of the governance system, the routines, rules, and expectations in place, did not disappear. The study indicates that the possibilities and limits of participation in different places hinge on not only the ideas regarding participation, but also on the ways planning is structured and how it is embedded in broader governance configurations. The authors also suggest that, at least for many European countries, a transition might be going on towards more participatory governance through network steering, marked by more opportunities for learning and adaptation. Existing, more centralized systems and their modes of learning and adaptations cling on and influence learning modes in specific ways.
De Groot, Leendertse, and Arts (2020) stay within the realm of complex public projects and their learning potential. They focus on transport infrastructure networks, which are under a variety of pressures. These pressures do not allow for easy integration into an optimal design and management strategy, and they are highly dynamic, which prompted calls for more adaptive management. Learning is understood as enhancing adaptive capacity. In the vein of the aforementioned Hood and Peters (2004), the authors ask how agencies responsible for large infrastructure projects learn and how this contributes to their adaptive capacity. Indeed, as Willems et al. (2020) also point out, and in line with classic analyses such as (Scott, 1998), once engineeringdominated public organizations are in charge of complex projects, they are hard to dislodge, and the associated discourses on expertise, steering, and participation are hard dismantle. De Groot et al. (2020) give central place to the project level of organization, with in some cases projects serving as new entities coordinating various organizations, and in other cases as more rudimentary information exchange platforms coordinating actors within the organization. The authors observe the general success of projects in terms of local adaptation but also the distance between project discourse and the mother organizations, or higher management (echoing the Luhmannian analysis of Seidl, 2016; see also Van Assche & Verschraegen, 2008).
De Groot et al. (2020) and Willems et al. (2020) both bump into the central issue of complexity in current governance. Complexity is both necessary and problematic in the search for answers to big problems in democratic societies. Planning, as coordination in the organization of space is bound to encounter complexity, because people want to do many things in space and project many competing meanings on it. Planning is thus faced with an intricate web of expectations, interests, forms of knowledge, actors, institutions, pasts, and futures. Learning from planning systems is therefore not only learning about different contexts, ideologies etc., but also learning about distinct modes of creating and managing complexity. This becomes even more important because complexity is increasing with ongoing differentiation in society (Seidl, 2016), and because contemporary sustainability issues demand unprecedented levels of coordination (Patterson et al., 2017). Complexity, again, is a double-edged sword, as planning complexity is required to deal with external complexity, and as planning complexity renders smooth adaptation and learning difficult (de Roo & Silva, 2010). Neither expert-driven systems nor highly participatory and decentralized systems have a distinct advantage in the abstract here. The devil, as usual, is in the detail.

Comparing for Learning and Comparing the Learning
This idea is confirmed in this issue by Leinfelder and Buitelaar (2020), who analyze patterns of urban sprawl in Flanders and the Netherlands, with on first assumption the devil residing in Flanders, where sprawl rules. Leinfelder and Buitelaar (2020) use a detailed comparative study, invoking other comparatives along the way (Italy, US). The authors do not confirm the negative stereotypes on sprawl dominant in the US planning literature and present a subtle analysis of driving factors of sprawl in both Flanders and the Netherlands, with distinct forms of sprawl clearly emerging as a result of more than laissez-faire-laissez-passer attitudes. Indeed, they show that histories of governmental decisions, of institutional choices, and material legacies, as in the physical landscape resulting from earlier planning, trigger particular forms of sprawl while discouraging others. In terms of Evolutionary Governance Theory (Van Assche et al., 2014) one can speak of the interplay between path dependencies, interdependencies, and goal dependencies (effects on governance of visions of the future), which shapes the possibilities of containing sprawl and the possibilities of learning from others (e.g., from Dutch neighbors) to do so.
The study by Leinfelder and Buitelaar (2020), as the others in this issue, highlight the utility in comparative planning studies to include the learning modes in the observed systems in the analysis, which will deepen the learning from the comparative analysis. The choice of the Flemish and Dutch planning system is interesting also because it offers rich possibilities to study comparative learning in planning systems: There is a tradition of shifting images of the neighbor in each of these systems, changing interpretations, which then triggered different forms of learning, ranging from attempts to copy to learning by avoiding the neighbors mistakes.
In their commentary, Rooij and van Dorst (2020) focus on the pedagogical uses of comparative work. They report on the pattern language approach to design and design pedagogy, an approach originally proposed by Christopher Alexander in the 1960s (Alexander et al., 1977). Alexander's work in their view deserves a reappraisal and can be developed to help students quickly analyze a place, structure their design thinking without pushing it too much towards a particular solution. A plethora of comparative work underpins the pattern language approach, while it also enables quick comparison of places, their structural features, qualities and problems, and results of previous planning and design interventions. The updated approach was tested in class and found useful by students, and it points at an argument made by D'Hondt et al. (2020) in their contribution, i.e., that indeed context is almost everything, that learning from other places means adapting insights to a new context, but that nevertheless, one has to remain open to the possible travel of both problems and solutions. This is partly an issue of transcending context, partly of sharing context (spatial, economic, institutional).

Conclusion: A Long Way to Go towards Sustainability Learning
We emphasize that this thematic issue is the beginning of a conversation. Indeed, learning might be popular in various policy-related literatures but, as said, large gaps remain in the terrain between the poles of naive learning optimism and learning as necessarily captured by strategy and competition. Each planning system has its own modes of learning, with its own potential for comparative learning and for linkages with other forms of learning. More than particular methods of comparison, what counts is the location of the comparison in broader research and/or policy goals, and the location of the compared planning system in broader governance configurations and histories. Those embeddings will co-determine how to interpret success and failure in observed systems, as they will shape the possibilities for understanding and organization in a system-presumed-to-learn. More attention is also needed for non-learning, overlooking or ignoring what is learned and learning the 'wrong' things, as these mechanisms to influence the evolution of planning systems.
Much of what has been said is relevant for likely the biggest challenge for planning and governance: planning and policy for sustainability (or resilience, climate change, energy transition, etc.). Much learning, and much learning through comparison, will need to take place before people know which tools might actually work under given conditions, before they know which aims are realistic and which forms of governance might allow for the forms of coordination needed, and before they have a good understanding of which modes of balancing, integrating, and differentiating expertise and which checks and balance might be lost, and which ones have to be guarded at all cost. Much of what we just listed is a matter of politics, not science; yet where there is a role of science, the topic of learning through comparison, in policy and planning, will most likely require much more attention than it gets now. And it will need to transcend its ideological quibbles and dividing lines to present a more realistic analysis of the potential of learning to illuminate the potential of societal transformation.