Mechanisms influencing mainstreaming of adaptation in spatial development: case studies in three Dutch municipalities

Abstract

Despite adopting a mainstreaming approach to spatial adaptation, few cities actually implement it. Why this is so is disputed, as the implementation stage has been under-researched. Inspired by the “mechanism-based” strand, this paper analyses three implementation projects in the Netherlands. Mechanisms are generalizable patterns yielding more plausible explanations for phenomena than “superficial” barriers or drivers. We found one overarching mechanism: the “institutional void” – an absence of policy frameworks with clear norms and agreements on adaptation. It led to three hampering mechanisms (avoidance of administrative responsibility, aversion to innovation and avoidance of private responsibility) and three stimulating mechanisms (window of opportunity exploitation, efficient coupling and policy entrepreneurship). We demonstrate that, although hampering mechanisms constrain the solution space for mainstreaming adaptation, the stimulating mechanisms prove that room remains for committed actors to use the limited available space efficiently.

Keywords:

• urban adaptation
• spatial development
• mainstreaming of adaptation
• implementation
• mechanisms

1. Introduction

Urban areas worldwide are faced with climate change impacts that directly influence the human environment, such as extreme drought, heavy precipitation and sea-level rise. Even optimistic climate models project with high confidence that these impacts will continue beyond 2100 (IPCC 2018). In this light, cities are facing the major challenge of adapting to climate change to ensure livability and the wellbeing of their citizens now and in the future (Uittenbroek et al. 2014).

Usually, local or regional governments are responsible for ensuring the quality and safety of public space (Mees 2017). Yet, paradoxically, in contrast to climate mitigation, climate change adaptation, hereafter referred to as “adaptation”, often does not have its own sectoral policy domain with direct political commitment and allocated resources (Uittenbroek et al. 2014). One way to still bring the cross-cutting policy issue of adaptation to the fore is by integrating adaptation measures into other policy sectors. Within this so-called “mainstreaming approach”, adaptation is not the primary goal of policy action, but is embedded “across […] compartmentalised, fragmented, traditional or siloed policy subsystems” (Biesbroek 2021, 75). Although adaptation has gained increased attention from public actors, a growing body of literature shows that local and regional governments often struggle to move from including adaptation in various policy plans to actually integrating measures into existing spatial renewal projects or programmes, such as infrastructural renewal or the replacement of a sewer system (Biesbroek et al. 2014; Pieterse, Du Toit, and Van Niekerk 2021; Runhaar et al. 2018; Uittenbroek et al. 2014; Uittenbroek 2016). Many of these studies indicate challenges of responsibility, since from a mainstreaming approach, the power to act is not assigned to specific entities (Runhaar et al. 2018; Uittenbroek et al. 2014). The same studies, however, do not necessarily explain why implementing actors do not take on responsibility for mainstreaming adaptation. Consequently, interventions to improve the implementation remain poorly understood. The current paper draws inspiration from Beach and Pedersen (2013) and Biesbroek et al. (2014), who used a so-called “mechanismic” approach (hereafter: mechanism approach), to explain this phenomenon. We present the results of an empirical study in three Dutch municipalities in the form of mechanisms (i.e. underlying patterns of interaction between actors) that plausibly explain why mainstreaming adaptation in existing projects is often challenging in a context of ambiguous responsibilities.

To the best of our knowledge, so far, only Biesbroek et al. have provided mechanism-based explanations of challenges to the mainstreaming of adaptation in spatial renewal projects. We seek to deepen this work by specifically focusing on mechanisms related to responsibilities for mainstreaming adaptation, while we broaden it by identifying both hampering and stimulating mechanisms. We thereby hope to contribute usefully to the existing literature on what works and what does not when implementing adaptation from a mainstreaming approach. Besides its scientific relevance, we also expect the findings to be useful for local practitioners involved in adaptation as well as for policymakers at higher governance levels.

This paper is structured as follows. In Section 2, the key concepts of this research are clarified. The methodology used for this study is discussed in Section 3. Section 4 presents the research findings in the form of seven mechanisms. In Section 5, the findings are discussed and policy strategies are suggested, leading on to the conclusions in Section 6.

2. Mechanisms for understanding challenges of responsibility

Over the past decade, mainstreaming as an approach for implementing adaptation has gained a prominent place on the research agenda of environmental governance (Runhaar et al. 2018; Uittenbroek 2016; Uittenbroek et al. 2014), urban planning (Biesbroek 2021; Biesbroek et al. 2014; Lehmann et al. 2015; Wamsler and Pauleit 2016) and natural hazard management (Thaler et al. 2019). The resulting body of literature has introduced terms such as “implementation gap” (Runhaar et al. 2018) and “adaptation deficit” (Eisenack et al. 2014) to indicate that adaptation policy and plans are often not translated into actual outcomes. The systematic literature review by Runhaar et al. (2018) suggests that this is mainly due to a lack of sustained political commitment from higher levels of government, and to a lack of effective coordination and cooperation between sectoral departments within and across policy domains. These barriers indicate a wider context of ambiguity about responsibility for the implementation of adaptation, which can be considered typical for a mainstreaming approach for several reasons. First, adaptation is a matter of governance. The various modes through which adaptation can be governed (Driessen et al. 2012), such as public–private governance (e.g. Mees, Driessen, and Runhaar 2012), decentralized governance (e.g. Biesbroek et al. 2014; Uittenbroek et al. 2014; Uittenbroek 2016) and self-governance (e.g. Thaler et al. 2019), reflect that responsibilities for adaptation are not assigned to one level of government or group of actors. However, in practice, adaptation is typically left to local governments to implement (Runhaar et al. 2012, 2018; Mees 2017). Second, a mainstreaming approach to implementation characterizes indirect rather than direct political commitment to adaptation (Uittenbroek et al. 2014). Although adaptation objectives often overlap with, and should be integrated into, various policy domains such as spatial development, water management and public health, departmental commitments to adaptation tend to be limited (Runhaar et al. 2018; Uittenbroek et al. 2014). Finally, as argued by Uittenbroek (2016), the implementation phase faces the most barriers because the implementing actors are different from those developing the policy plans.

Although adaptation barrier literature (using different terms, see Table 1) leaves little doubt that responsibilities for the mainstreaming of adaptation are ambiguous, systematic explanations of why actors do not take on responsibility for adaptation are still lacking (Eisenack et al. 2014; Biesbroek et al. 2014). Several scholars have explored the root causes of barriers, however. For instance, Moser and Ekstrom (2010) have suggested that “interconnected structural elements” help to explain why barriers arise, while Lehmann et al. (2015) refined these into “second-tier barriers”, such as actor-specific characteristics, the institutional environment and the natural and socio-economic environment (see Table 1). Although these frameworks do provide more explanatory knowledge, Eisenack et al. (2014) noted that the literature is contradictory in that empirical cases sometimes indicate different root causes for the same barrier. Hence, they promoted a comparative, actor-centred research approach that is able to identify common causal patterns and interdependencies of barriers. Biesbroek et al. (2014) responded by identifying mechanisms that go beyond describing barriers and their root causes, instead aiming to explain more generically and systematically how and why they have occurred. Whereas barriers are typically static variables, factors or conditions, mechanisms are recurring processes consisting of “entities (with their properties) and the activities that these entities engage in either by themselves or in concert with other entities. These entities bring about change…”. ¹ (Biesbroek et al. 2014, 109, citing Hedström 2005). Processes that are similar and are “identified in different cases, … can be generalized and captured into mechanisms” (Biesbroek et al. 2014, 109). A mechanism approach thus puts actors at the heart of explanations for the occurrence of governance challenges. In this paper, we adopt the same approach to explain why actors do, or do not, take on responsibility for mainstreaming adaptation in urban renewal projects. However, we use the term “mechanisms” (Biesbroek et al. 2014) rather than “causal mechanisms” (Beach and Pedersen 2013) to clarify that we do not necessarily prove causality, but rather draw inferences about causality.

Table 1. Overview and comparison of concepts used in adaptation literature.

Concept	Explanation in the context of this study	Knowledge type
Barriers to climate adaptation (Moser and Ekstrom 2010)	Variables indicating that actors do or do not take on responsibility or feel responsible for the mainstreaming of adaptation in urban renewal projects	Descriptive
First-tier barriers and opportunities for urban adaptation planning (Lehmann et al. 2015)
Superficially identified barriers (Biesbroek et al. 2014)
Mainstreaming drivers and barriers (Runhaar et al. 2018)
Structural elements of adaptation (Moser and Ekstrom 2010)	Variables indicating the root causes of why actors do or do not take on responsibility or feel responsible for the mainstreaming of adaptation in urban renewal projects	Explanatory
Second-tier barriers and opportunities for urban adaptation planning (Lehmann et al. 2015)
Mechanism (Biesbroek et al. 2014)	Recurring process (consisting of actors and the activity they engage in) that can be considered to account for why actors do or do not take on responsibility or feel responsible for the mainstreaming of adaptation in urban renewal projects	Explanatory
Causal mechanism (Beach and Pedersen 2013)

3. Methods

3.1. Research design and Cases

We opted for an embedded case study design to examine the mainstreaming of adaptation in three urban renewal projects in municipalities in the Netherlands: (1) project Betere Buurten Schansen Noord (Nieuwegein municipality); (2) project Proeftuin (Houten municipality); and (3) projects in Kanaleneiland Noord and Overvecht (Utrecht municipality). All projects addressed densely populated, rundown neighborhoods with limited public space. They were also implemented in a similar institutional setting of decentralized governance (Driessen et al. 2012), with the local municipality as implementing actor.

The projects had been selected as pilots by the Dutch National Delta Programme on Spatial Adaptation based on their intention to mainstream adaptation measures. The Delta Programme subsidized the pilots to monitor the implementation process and use the knowledge in the Delta Plan Spatial Adaptation, and it also subsidized the research underlying the present study (Van den Ende et al. 2021). Thanks to their frontrunner position in adaptation mainstreaming, the pilot projects can be characterized as “revelatory cases” (Yin 2009) or “favorable cases” (Flyvbjerg 2006). Pilot schemes have access to additional resources; therefore if our study were to find that it had been difficult to implement adaptation, it seems likely that in “non-revelatory cases” or “intermediate cases” implementation would be equally or more difficult (Flyvbjerg 2006; Yin 2009). Any stimulating mechanisms identified by us could provide relevant information for cases initiating similar processes. This guaranteed that the mechanisms identified in this paper would have external validity.

3.2. Data collection methods

This paper builds on empirical data we collected for a project monitoring and evaluation report (see Van den Ende et al. 2021), using various sources and methods. First, to gain a basic understanding of the three urban renewal projects, research analyses conducted by law students at Utrecht University were studied. This information was complemented with a document study of publicly accessible project reports. To develop a theoretical knowledge base, the literature on adaptation challenges was reviewed, focusing on studies on mainstreaming and the implementation stage of the policy cycle. The project analyses, document study and literature study were used to formulate interview questions. Second, semi-structured interviews were conducted with five civil servants from the three case municipalities: two adaptation policy advisors, an urban water and sewerage technical advisor and two programme managers.

Due to the COVID-19 restrictions, most interviews were held online, in Dutch. They lasted approximately one hour and comprised three parts. The first part investigated the governance context; interviewees were asked to give a chronological account of governance interactions, as well as important moments, events, or milestones for the process of mainstreaming adaptation in projects. The second part entailed describing the factors, conditions or events that influenced mainstreaming attempts, either positively or negatively. Interviewees were then asked follow-up questions on why they think things happened in certain ways. In the final part of the interview, interviewees were asked to evaluate the project in terms of adaptation goals reached.

3.3. Identification of mechanisms

A mechanism can be understood as the explanatory link between the independent variable ‘X’ (comparable to the concept of superficially identified or first-tier drivers and barriers) and the dependent variable or outcome ‘Y’ (Beach and Pedersen 2013). In this paper, X is considered to be a context of ambiguity of responsibilities for the implementation of adaptation, while Y is the extent of mainstreamed adaptation in urban renewal projects. To identify mechanisms, we opted for a process-tracing method (Beach and Pedersen 2013). In particular, we chose theory-building process tracing because we wished to infer generalizable mechanisms from recurrent processes in empirical cases. The data analysis was structured along three steps proposed by Beach and Pedersen (2013). Step 1 was to collect, for each case, detailed empirical narratives about the process of mainstreaming adaptation measures in the urban renewal project (see the technical report by Van den Ende et al. 2021). We selected only the narratives relating to responsibility or accountability. Step 2 was to infer, from these empirical narratives, observable manifestations of an underlying mechanism. We framed the manifestations in terms of an entity and an activity and we placed them in a chain of events. The last step was to deduce mechanisms from these sequences of observable manifestations. Similar processes traced in at least two cases were labeled as a mechanism. For this creative and intuitive process we drew inspiration from scholarly studies in public administration and policy sciences (e.g. Hajer 2003; Howlett 1998; Huitema and Meijerink 2010; Kingdon 1993; Meijerink 2005). Although linearly formulated, the three steps were highly iterative. One particularly important element in this process was the internal validation of outcomes with interviewees via email and during online meetings. With regard to the external validation, it is important to note that the mechanisms are generalizable, but the observable manifestations are not. Other researchers may observe different manifestations, yet identify mechanisms similar to those we identified. They might also find different mechanisms.

4. Results

In this section, we discuss the four mechanisms that hampered and the three mechanisms that stimulated the implementation of mainstreamed adaptation. Tables 2 and 3 show, respectively, how each mechanism was identified based on empirical manifestations. These manifestations form an explanatory link between the same independent variable, i.e. a context of ambiguity of responsibilities for the implementation of adaptation, and the dependent variable, i.e. the extent of mainstreamed adaptation, which can differ per mechanism.

Table 2. Process tracing of the mechanisms hampering the mainstreaming of adaptation in the three empirical cases.

1. Observable manifestations of the ‘institutional void mechanism’ in a chain of events			Observable manifestation of Y
1.1	The urban design handbook for spatial development used by the municipality was unsuitable for innovative adaptation (Nieuwegein)	Programme managers referred to this handbook to reject adaptation measures (Nieuwegein)	No mainstreaming of adaptation in urban renewal (Utrecht, Nieuwegein)
1.2	In the absence of an administratively determined adaptation policy framework, sectoral departments were not confronted with hard agreements and enforceable adaptation goals (Utrecht)	Sectoral departments prioritized regulated tasks over unregulated adaptation ambitions (Utrecht)
1.3	The municipality had ambitions but had not prioritized any in the project requirements (Utrecht)	All decisions about which adaptation measures to include in the spatial design were made by urban designers (Utrecht)
1.4	The municipality did not specify technical aspects of adaptation (Utrecht)	Urban designers struggled with translating adaptation into elementary measures (Utrecht)
2. Observable manifestations of the ‘avoidance of administrative responsibility mechanism’ in a chain of events			Observable manifestation of Y
2.1	“Green” and “grey” departments had a siloed mentality (Utrecht)	Departments worked at cross purposes without exploring the benefits of spatial measures with multiple functions (e.g. green space, ecology and water retention) (Utrecht)	No mainstreaming of adaptation in urban renewal (Utrecht, Nieuwegein)
2.2	Sectoral departments and policy advisors lacked experience with unintended effects of adaptation measures (Utrecht)	Sectoral departments were reluctant to risk any unexpected responsibilities for maintenance (Utrecht)
2.3	Municipal departments did not consider adaptation as part of their job (Nieuwegein)	Sectoral departments were unwilling to spend their budget on adaptation measures or additional maintenance that adaptation measures require (Nieuwegein)
3. Observable manifestations of the ‘aversion to innovation mechanism’ in a chain of events			Observable manifestation of Y
3.1	Sectoral departments were familiar with business-as-usual measures for spatial development (Nieuwegein)	Sectoral departments rejected more climate-proof measures, as these deviated from existing norms and standards (Nieuwegein)	No mainstreaming of adaptation in urban renewal (Utrecht, Nieuwegein)
3.2	Policy advisors lacked experience with possible unwanted side-effects of innovative solutions (Utrecht)	Sectoral departments were unwilling to take the risks that accompany more innovative solutions (Utrecht)
3.3	The municipality received a grant from the national Delta Programme for experimentation (Houten, Nieuwegein)	Policy advisor and programme managers did not depend on the commitment or approval of sectoral departments (Houten, Nieuwegein)	Mainstreaming of innovative adaptation in urban renewal (Houten, Nieuwegein)
4. Observable manifestations of the ‘avoidance of private responsibility mechanism’ in a chain of events			Observable manifestation of Y
4.1	Citizens considered the municipality to be responsible for adaptation (Houten)	Citizens did not feel an urgent need for adaptation (Houten)	Private actors barely contributed to the mainstreaming of adaptation in urban renewal (Houten, Nieuwegein, Utrecht)
4.2	Citizens prioritized legal obligations and day-today matters (Houten, Utrecht)	Citizens did not feel an urgent need for adaptation (Houten, Utrecht)
4.3	No enforcement of the requirement not to exceed a maximum paved area in private gardens (Utrecht)	Citizens did not feel an urgent need for adaptation (Utrecht)
4.4	Citizens lacked knowledge on adaptation (Nieuwegein)	Citizens feared potential risks of measures such as green façades (Nieuwegein)
4.5	Housing corporations were unable to quantify the benefits of adaptation (Utrecht)	Housing corporations wanted to avoid higher rents (Utrecht)
4.6	The national government has no legally binding performance agreements for adaptation (Utrecht)	Housing corporations prioritized legally binding performance agreements (e.g. for climate mitigation) over adaptation (Utrecht)
4.7	Housing corporations held tenants responsible for adaptation (Utrecht)	Tenants held housing corporations responsible for adaptation (Utrecht)

Table 3. Process-tracing of the mechanisms stimulating mainstreaming of adaptation in the three municipalities.

5. Observable manifestations of the ‘window of opportunity exploitation mechanism’ in a chain of events			Observable manifestation of Y
5.1	The municipality experienced an extreme weather event (Houten)	Municipal awareness of the need for adaptation grew and the municipality decided to adapt the city hall square to climate change (Houten)	Mainstreaming of adaptation in urban renewal took place (Houten, Nieuwegein, Utrecht)
5.2	The municipality experienced an extreme weather event (Nieuwegein)	Municipal awareness of the need for adaptation grew and the municipal council provided an additional budget for adaptation (Nieuwegein)
5.3	The municipality assessed the current spatial design for planned urban renewal projects (Houten, Nieuwegein, Utrecht)	The municipality foresaw climate-related problems in the future and decided to integrate adaptation (Houten, Nieuwegein, Utrecht)
5.4	The municipality received a grant from the national Delta Programme for mainstreaming adaptation in urban renewal projects (Houten, Nieuwegein, Utrecht)	The policy advisor and programme managers were able to demonstrate the financial feasibility of some innovative adaptation measures (Houten, Nieuwegein)
6. Observable manifestations of the ‘efficient coupling mechanism’ in a chain of events			Observable manifestation of Y
6.1	The policy advisors and programme manager linked adaptation with planned urban renewal projects (Houten, Nieuwegein, Utrecht)	The policy advisor and programme managers were able to use funds from multiple sources (Houten, Nieuwegein, Utrecht)	Mainstreaming of adaptation in urban renewal took place (Houten, Nieuwegein, Utrecht)
7. Observable manifestations of the ‘policy entrepreneurship mechanism’ in a chain of events			Observable manifestation of Y
7.1	The municipality appointed policy advisors on adaptation (Utrecht, Nieuwegein)	Policy advisors actively promoted adaptation in sectoral departments (Utrecht, Nieuwegein)	Mainstreaming of adaptation in urban renewal took place (Houten, Nieuwegein, Utrecht)
7.2	The municipality appointed policy advisors on adaptation (Utrecht, Nieuwegein)	Policy advisors found and exploited policy opportunities to implement adaptation (Utrecht, Nieuwegein)
7.3	The programme manager aspired to make the municipality frontrunner in adaptation (Houten)	The programme manager found and exploited policy opportunities to implement adaptation (Houten)

4.1. Hampering mechanisms

4.1.1. Institutional void mechanism

In Utrecht and Nieuwegein, we observed that an institutional void around adaptation hampered the mainstreaming of measures in urban renewal projects (Van den Ende et al. 2021). Hajer (2003, 175) uses the term to indicate that policymaking often occurs in a context of “no generally accepted rules and norms according to which politics is to be conducted and policy measures are to be agreed on”. Empirical manifestations of an institutional void mechanism were observed in Nieuwegein and Utrecht. According to the interviewed policy advisor of Nieuwegein, existing municipal policy on spatial development was not particularly conducive to integrating adaptation; programme managers worked in accordance with an urban design handbook that provided limited scope for innovation that deviated from municipal norms and standards (see Table 2). Examples of measures rejected for this reason are wood chip paths and grass pavers. In Utrecht, also, the local policy advisor noted that without a clear policy framework for adaptation, “… [adaptation] only remains an ambition and the moment things start to get tight, the ambitions that have not been agreed upon will be dropped. Then, a parking norm is prioritised over adaptation, as it’s written down in the rules”. A third manifestation can be recognized in the lack of municipal guidance to urban designers during the design process. The policy advisor mentioned that the municipality of Utrecht did not emphasize priorities in project requirements and left the choice of adaptation measures to designers, most of whom lacked experience in adaptation. Finally, designers were not guided on how to translate the ambition of adaptation as a “catch-all term” into concrete goals and technical requirements, such as the amount of green space, the number of trees, or the type of street profile. These four empirical manifestations suggest that, in a context of ambiguity about responsibilities for adaptation, a lack of institutional guidance negatively influences the extent to which adaptation is mainstreamed in urban renewal projects (see Table 2).

4.1.2. Avoidance of administrative responsibility mechanism

In Utrecht and Nieuwegein, we identified empirical manifestations of an avoidance of administrative responsibility mechanism, i.e. the tendency of sectoral departments to refrain from taking on responsibility for a more climate-adaptive approach to spatial development (Van den Ende et al. 2021). The interviewees noted that, in Utrecht, adaptation elements were barely integrated into the visions of the two municipal departments most involved in spatial development: “green” (public green space) and “grey” (infrastructure). To illustrate, the technical advisor of Utrecht mentioned that “you often see that the public green space is higher than the street level. And from a climate adaptation perspective, you want to be able to let the rainwater run to the green public space, so they have to be lowered. […] but it is still the case that new rows of trees are planted on higher ground”. Departments held onto their silo mentality and approached spatial development as separate entities rather than taking the initiative to act as one “municipality” striving for a spatial design with multiple integrated goals (see Table 2). The policy advisor for Utrecht also mentioned a lack of experience of the unintended side-effects of adaptation measures. For example, bioswales may cause discharged waste to accumulate, while lowering the ground level of public green space may encourage undesirable parking on verges. This uncertainty raised questions relating to responsibility for the maintenance. Furthermore, in Nieuwegein the mechanism manifested in that sectoral departments opted for the business-as-usual over climate-adaptive solutions to spatial development, since they did not consider adaptation to be part of their remit. For example, neither wood chip paths nor green parking space made it to the design table; the former because it was thought they would result in messy streetscapes and rotting wood, the latter because of undesirable side-effects such as muddiness. According to the policy advisor, adaptation tends to trigger “eternal discussions” between departments that feel no responsibility to spend their budget on additional maintenance work.

4.1.3. Aversion to innovation mechanism

In all cases, we found indications that municipal departments stuck to business-as-usual solutions with which they were familiar, fed by resistance to a change in working practices (Van den Ende et al. 2021). In Nieuwegein, the aversion to innovation mechanism manifested as departments opting for default measures in spatial designs, such as paving materials, that have proved to work well, but are not climate-proof (see Table 2). A more innovative measure such as permeable clay brick pavers was long regarded as non-negotiable, as it was less stable and not so neatly finished as regular paving. The adaptation measure of wood chip paths did not make it to the spatial design either, because it did not meet maintenance norms. According to the policy advisor, “departments pointed out that they did not want all that novelty”. In Utrecht, a similar manifestation was observed, as the “green” and “grey” departments integrated hardly any climate-adaptive elements in their vision. One reason for sectoral departments opting for business-as-usual solutions was, among others, a lack of knowledge and experience with adaptation and, consequently, the fear of making a poor decision. Interestingly, despite their innovative nature and high costs, the adaptation measures of trees planted in structural soil and pop-up water defence systems were successfully implemented in Nieuwegein and Houten respectively. Here, a grant from the national Delta Programme allowed the programme manager and policy advisor to experiment without needing approval from sectoral departments.

4.1.4. Avoidance of private responsibility mechanism

We identified an avoidance of private responsibility mechanism that is comparable to the avoidance of administrative responsibility mechanism. The main difference is that governments, not private actors, are formally responsible for ensuring the quality and safety of public space (Mees 2017; Uittenbroek et al. 2019). We included this mechanism because there are governance scholars who promote citizen responsibilisation to support local governments in serving public interests. An argument often used for this is that governments alone cannot adapt cities to climate change; private actors need to take action on private property (Mees 2017; Uittenbroek et al. 2019).

In all three cases, the avoidance of private responsibility mechanism manifested in the limited degree of adaptation shown by private actors, such as citizens and social housing corporations (Van den Ende et al. 2021). All interviewees indicated the lack of a sense of urgency for adaptation among citizens, presumably because of the difficulty of comprehending climate change impacts (see Table 2). Only in Houten had citizens experienced such impact in the form of flooding in the city hall square, but they considered the municipality responsible and expected it to act. Furthermore, citizens prioritized other obligations, such as the energy transition, or more mundane matters. The case of Utrecht showed that official rules are not always effective; although social housing corporations had set a maximum percentage for paved areas in gardens, this was not enforced and so citizens ignored it. Furthermore, homeowners refrained from green façades because of potential damage. This suggests that the innovative character of adaptation dissuades citizens as well as municipal departments. Social housing corporations withdrew from adaptation for economic reasons: they wanted to avoid higher rents and mentioned the challenge of making adaptation measures financially attractive. They also indicated legally binding priorities such as climate mitigation were more important than adaptation. Finally, social housing corporations considered their tenants to be responsible for implementing certain adaptation measures, such as sumps.

Despite the hampering mechanisms, all three municipalities eventually integrated some adaptation measures into their urban renewal projects. Below, we showcase three mechanisms that, despite the presence of the hampering mechanisms discussed above, stimulated the mainstreaming of adaptation.

4.2. Stimulating mechanisms

4.2.1. Window of opportunity exploitation mechanism

One mechanism that positively influenced the extent to which adaptation was mainstreamed in urban renewal is the window of opportunity exploitation mechanism. A window of opportunity presents itself “when a problem is recognized, a solution is available and the political conditions are right” (Kingdon 1993, 41). We observed that in all three cases, a problem was recognized for which adaptation turned out to be a fitting solution (Van den Ende et al. 2021). For instance, in Houten, the frequent pluvial floods in the paved city hall square were damaging the city hall (see Table 3). Nieuwegein and Utrecht had not experienced any negative impacts of climate change, but the planned urban renewal made them aware that the new spatial design should be climate-proof. Nieuwegein reached this conclusion after a mandatory climate risk assessment indicated streets that were vulnerable to impacts from climate change. When an extreme precipitation event occurred a few years later, municipal awareness of the risks increased and the local council eventually decided to allocate extra funds for adaptation. In all cases, the grant received for adaptation improved the political conditions.

4.2.2. Efficient coupling mechanism

The efficient coupling mechanism differs from the window of opportunity exploitation mechanism in that adaptation is initiated as an improvement to an existing solution, rather than as a solution to a problem. The efficient coupling mechanism reflects proactive endeavors to let adaptation efficiently “piggyback” on other projects. More precisely, all municipalities consciously linked adaptation with planned urban renewal projects (Van den Ende et al. 2021). In Nieuwegein, for instance, a large urban renewal programme for outdated postwar neighborhoods was exploited, to efficiently integrate adaptation elements such as water-retentive road surfacing (see Table 3). In Utrecht, adaptation was linked with the redevelopment of public space, the replacement of sewer systems, and the renovation of social housing buildings. In Houten, adaptation measures were efficiently integrated with the redevelopment of the entire city hall square, which the municipality initiated in response to citizens’ requests to provide a meeting spot for the elderly and to create an office garden. All interviewees indicated that the efficient coupling of adaptation with other work increased the financial feasibility.

4.2.3. Policy entrepreneurship mechanism

Finally, all cases yielded empirical evidence for the existence of a policy entrepreneurship mechanism (Van den Ende et al. 2021), which refers to the positive influence of adaptation ambassadors on the mainstreaming of adaptation in spatial designs. Kingdon (1984, as cited in Huitema and Meijerink 2010, 25) described policy entrepreneurs as “people willing to invest their resources in return for future policies they favor. They are motivated by […] their straightforward concern about certain problems, […] their promotion of their policy values, and their simple pleasure in participating”. The mechanism manifested in the decisive role of policy advisors and programme managers in making their municipality a frontrunner in adaptation (see Table 3). Among other things, they exploited windows of opportunity that arose, such as extreme weather events. During the design phase, they promoted adaptation measures – from very simple measures to more innovative ones. For instance, in Utrecht, it was the programme manager, policy advisor and technical advisor who actively tried, sometimes successfully, to convince sectoral departments to develop climate-proof measures. In Nieuwegein, the adaptation policy advisor also insisted that sectoral departments consider adaptation aspects when developing their measures. Although several attempts such as wood chip paths failed, after much persuasion the measure of permeable clay brick pavers was accepted and has since become a standard element in spatial design. It was this policy advisor who applied for and received a grant, which was spent on more innovative techniques. The programme manager in Houten was also highly committed to adaptation; he aspired to make the municipality a frontrunner in urban sustainability and adaptation. He too applied for and was awarded a grant to experiment with innovative measures.

5. Discussion

In this section, we briefly discuss the seven mechanisms presented above. In terms of sequence, we found one logical relationship, namely that the institutional void mechanism as an overarching mechanism shapes the context for the three other hampering mechanisms and the three stimulating mechanisms. This affects the solution space to mainstreaming adaptation. According to Haasnoot et al. (2020, 2), “the solution space represents the boundaries of what adaptation is socially accepted, economically affordable, politically feasible, and technically possible and is, therefore, the ‘room to maneuver’”. Given this definition, it can be assumed that the solution space for the mainstreaming of adaptation is constrained by the hampering mechanisms. However, the stimulating mechanisms suggest there is still some room for maneuver that policy entrepreneurs can make smart use of. It can be assumed that the opposite of an institutional void – a strong legal policy framework for adaptation – would set clear social, economic, political and technical boundaries for adaptation measures, yet thereby possibly exclude more innovative measures that fall beyond the boundaries.

Pursuing this line of thought, an institutional void can both restrict and liberate the mainstreaming of adaptation. It allows for weak political commitment to adaptation, while also providing room for experimentation. However, it should be remembered that limited commitment to adaptation can pose problems if there are no external, often incidental, events to be seized on, or if these options run out, or if there is no policy entrepreneur to make smart use of the available solution space. The latter is more likely the case for small municipalities that oftentimes struggle with a lack of capacity.

5.1. Interpretation of mechanisms in terms of policy stability and policy change

To better understand the potential of mechanisms to enable policy change in favor of adaptation, we further discuss them in light of existing theories on policy stability and policy change (Hajer 2003; Howlett 1998; Huitema and Meijerink 2010; Kingdon 1993; Meijerink 2005).

Interestingly, we noticed a certain ambiguity in the literature concerning the institutional void mechanism, as it can be interpreted by theories of policy stability and of policy change (Hajer 2003; Loorbach 2007; Meijerink 2005). On the one hand, the punctuated equilibrium framework (PEF) perceives the lack of formal institutions for adaptation as a stimulant of policy stability, as it maintains the policy monopoly of traditional spatial planning (Meijerink 2005). On the other hand, Hajer (2003) argues that policymaking on new policy issues initially takes place in a void of commonly accepted rules and standards. In fact, as “institutional arrangements might be challenged by new developments” (Hajer 2003, 177) the notion of an institutional void predicts that changes will occur in institutions, to fit the “actual practice of policymaking” (Hajer 2003, 176). Also, transition scholars emphasize the time dependence of institutional voids, since the institutionalization of new policy issues typically takes a long time (Loorbach 2007). Since the mechanisms apply to a policy topic in an early transition phase, it may be that, in five years from now, the institutional void has filled up with policy, rules and norms for adaptation, which may dissolve the hampering mechanisms and limit the relevance of the stimulating mechanisms.

The avoidance of administrative responsibility mechanism reflects a situation of policy stability. The advocacy coalition framework (ACF) would recognize in each sectoral department a coalition of civil servants that shares “beliefs concerning problems, causes, and remedies […] and thereby upholds certain core aspects of a policy programme” (Meijerink 2005, 1062), which, in this study, are the traditional values of spatial planning. The cases show that many departments acted from a “silo mentality”, whereas it has been argued that collaboration and coordination are essential for implementing policy issues, such as adaptation, that are not part of a policy domain but transcend sectoral boundaries (De Waal et al. 2019; Hajer 2003; Runhaar et al. 2018; Uittenbroek et al. 2014). The avoidance of private responsibility mechanism also reflects a situation of policy stability, maintained by two key processes: there is no legal obligation for private actors to adapt to climate change and, in this context, they refrain from taking on responsibility for adaptation for various reasons. However, it can be argued that the relevance of this mechanism is limited to urban areas in developed countries and where climate change impacts have not yet been highly disruptive, because in less developed countries prone to climate change it is often not so much the question whether private actors are willing to adapt, but to what extent they have the capacity to do so (Ogato et al. 2017).

The aversion to innovation mechanism can be explained by multiple theories of policy stability, but also reflects prospects for policy change. Again, from a PEF and ACF perspective, sectoral departments can be seen as coalitions trying to secure institutionalized ideas about traditional spatial planning by continuing business as usual and suppressing novelty (Meijerink 2005). Another relevant perspective is provided by the epistemic communities framework (ECF), which explains the extent of policy stability in terms of how much decision-makers trust their capability to deal with complex and uncertain policy problems and, thereby, feel they do not need “a network of professionals with recognized expertise and competence in a particular policy domain […]”. (Meijerink 2005, 1063). This might explain why the expert knowledge of policy advisors did not convince departments to adopt climate-adaptative measures. Interestingly, it seems that sectoral departments rejected adaptation measures based on Kingdon's (1993) three criteria for useful policy solutions of technical feasibility, financial feasibility and acceptability - where technical feasibility relates to uncertainty (e.g. unknown options and outcomes) while acceptability has more to do with political risk (e.g. known options and their likely outcomes, Osborne and Brown 2011). The pilot projects showed that financial feasibility was achieved by being awarded a grant, as this enabled experimentation with measures whose effectiveness (i.e. technical feasibility) was still uncertain, such as the pop-up water defence systems and the trees planted in structural soil. However, obtaining a grant did not solve the issue of responsibility for maintaining measures such as wood chip paths (i.e. acceptability). Hence, it can be argued that it requires not only more knowledge on the effectiveness of measures or a more steady financial basis, but also critical reflection on the municipal culture in which the political risk of being blamed for poor policy decisions may cause departments to avoid innovation (Howlett 2012).

The multiple streams framework (MSF) helps to explain why, despite a context of policy stability, adaptation was nonetheless integrated into spatial designs (Kingdon 1993). According to MSF, policy change occurs when a problem stream, solution stream and political stream converge (Howlett 1998; Meijerink 2005). The three stimulating mechanisms identified in the empirical cases seem to represent these streams. For instance, the window of opportunity exploitation mechanism shows that, supported by a grant (the politics stream), politicians initiated adaptation (the solution stream) to situations they considered problematic (the problem stream). However, although in Houten pluvial floods in the city hall square increased municipal awareness of the need for adaptation, it was citizens’ complaints that eventually accelerated the spatial redevelopment project. Furthermore, Utrecht had not experienced any problematic events, while in Nieuwegein the topic of adaptation was already on the political agenda when pluvial floods occurred. This suggests that, although climate change impacts may have been perceived as problematic, the events were not a reason for the municipality to change policy suddenly. The efficient coupling mechanism, in turn, reveals a pattern in which adaptation is only perceived as financially feasible if measures are efficiently linked to existing projects. Seeking win-win solutions can be seen as a way to pick the “low-hanging fruit” options and thereby avoid difficult political choices in a context of institutional void and avoidance of responsibility. However, the risk associated with this pattern is that it only works until the easy targets run out. Furthermore, both the window of opportunity exploitation mechanism and the efficient coupling mechanism reflect an opportunity-driven approach in which adaptation depends not only on external conditions but also on the presence of an actor that pushes forward solutions to a perceived problem. However, Howlett (1998, 499) quotes Kingdon who warns that: “[…] open windows are small and scarce. Opportunities come, but they also pass. Windows do not stay open long. If a chance is missed, another must be awaited”. In this study, the politics stream of the MSF comprises contexts of institutional void in which policy entrepreneurs play a crucial role in promoting adaptation in spatial planning. In other municipalities, however, the national mood, elected council members or citizens may push the municipality to opt for climate-proof neighborhoods (Kingdon 1993). Despite their important role, policy entrepreneurs do not necessarily renegotiate existing norms and rules for spatial planning. Cracking what Howlett calls the “pervasive bureaucratic mentality” of sectoral departments should not be directed at individual policy entrepreneurs, as “individual agency by itself is a necessary but not a sufficient condition for innovation in public services – organizational and political mandate for innovation (including of course the risk of failure) is essential also […]” (Osborne and Brown 2011, 1342). This suggests that although, as our study shows, policy entrepreneurs can make smart use of available policy space in a context of the institutional void, they are unlikely to solve the institutional void. We believe that as long as there is an institutional void around adaptations, the MSF continues to be relevant to explain the mainstreaming of adaptation in urban renewal projects, or a lack thereof.

5.2. Overcoming hampering mechanisms

The mainstreaming strategy upon which this paper focuses is what Wamsler and Pauleit (2016, 73) refer to as “programmatic mainstreaming: the modification of the implementing body’s sector work by integrating aspects related to adaptation into on-the-ground operations, projects or programmes”. According to them, for adaptation to become a standard element in municipal planning, there must be mainstreaming strategies at other levels too. One strategy that all our empirical cases at least partly adopted is managerial mainstreaming: the modification of job descriptions and provision of funding to employ adaptation ambassadors, such as policy advisors. However, this also involves assessing whether the municipal organization maintains a culture of avoidance of political blame. If necessary, the municipality could, instead, adopt a culture that is open to experimentation and stimulates learning from mistakes. An apparently necessary strategy for overcoming departments’ avoidance of responsibility is intra-organizational mainstreaming of adaptation, which encourages departments to collaborate and coordinate adaptation actions. However, in our study, this strategy proved to be inadequate. A strategy that most likely leads to more direct commitment of departments is regulatory mainstreaming, which involves changes to municipalities’ spatial planning procedures, regulations, and policies in favor of adaptation. However, this still requires municipalities to commit to adaptation. Directed mainstreaming of adaptation reflects a more hierarchical strategy in which the national government addresses the institutional void around adaptation by making clear who has the responsibility for ensuring appropriate regulations, policy and legislation. The findings show that grants for innovation did, indeed, enable policy entrepreneurs to overcome mechanisms of avoidance of administrative responsibility and aversion to innovation. Directed mainstreaming, such as legal performance agreements for green roofs and green gardens, may also overcome citizens’ and social housing corporations’ avoidance of responsibility, as current strategies of inter-organizational mainstreaming of adaptation focused on collaboration and networking has proved inadequate. These mainstreaming strategies imply that governments at national and sub-national level need to provide rules and guidance for mainstreaming at the local level. It can, therefore, be argued that for adaptation to be successfully mainstreamed on the ground also requires aspects of a more dedicated approach (Wamsler and Pauleit 2016).

6. Conclusion

Mainstreaming is seen as an efficient approach to implementing adaptation, but it also comes with challenges of responsibility (Runhaar et al. 2018; Uittenbroek 2016; Uittenbroek et al. 2014). Although barriers have been widely researched, little is known about why implementing actors often do not take on the responsibility for mainstreaming adaptation. We have attempted to address this knowledge gap, using a mechanism approach: a specific sub-debate in the adaptation governance literature that we believe has the best prospects of improving understanding and using the implementation phase more effectively. We applied the mechanism approach to the implementation of mainstreamed adaptation in three municipalities in the Netherlands. This yielded seven mechanisms that explain the extent of mainstreamed adaptation in a context of ambiguity about responsibility better than explanations based on superficially identified drivers and barriers.

The cases showed that a lack of institutions for mainstreaming adaptation (i.e. the institutional void mechanism) resulted in uncertainty about what should be done and who should do what. Because of their silo mentality and resistance to change, municipal departments turned out to be not very receptive to initiating adaptation without assigned responsibilities (i.e. the avoidance of administrative responsibility mechanism; aversion to innovation mechanism). Citizens and social housing corporations refrained from adaptation action for various reasons (i.e. the avoidance of private responsibility mechanism). If, in this context, policy entrepreneurs were requested to implement adaptation (i.e. the policy entrepreneurship mechanism), they tried to exploit policy windows to push adaptation through (i.e. the window of opportunity exploitation mechanism), while also making efficient use of existing initiatives and budget sources (i.e. the efficient coupling mechanism). The findings suggest that the first four mechanisms limit the solution space for mainstreamed adaptation, while the latter three mechanisms proved that actors can still make smart use of the limited solution space available.

Mechanism-based explanations, as we have shown, are easier to connect with mainstream literature on policy stability and change. This offers opportunities to arrive at more solid explanations from which it is also easier to derive action perspectives for effective implementation of mainstreamed adaptation. In this paper, we identified two routes of action perspectives. One route is for policymakers to address the existing institutional void around adaptation. This involves state and regional governments developing local policy frameworks, assigning clear responsibilities and providing more guidance. A second route is for implementing actors to work within the existing institutional void – that is, ensuring that windows of opportunity are recognized and exploited, trying to make smart connections and employing adaptation ambassadors. The risks associated with the second route relate to the dependence on external factors, such as shock events, or to focusing solely on win-win options or low-hanging fruit. Hence, we argue that the first route is crucial for moving forward in the long term. Further research is needed to explore possible pathways to effectively and legitimately assign responsibilities for the mainstreaming of adaptation to implementing actors.

Paradoxically, the generic mechanisms identified in this paper indicate that the empirical reality of mainstreaming adaptation is more nuanced than often appears from the literature. This nuance has possible implications for the organization of future policy evaluations. Broadly speaking, the findings raise the need to consider the specific context in which adaptation is mainstreamed in policy evaluations. For instance, a more realistic picture of the extent of adaptation could be generated if implemented measures, rather than the presence of strategic policy and planning documents, are the unit of analysis (e.g. Reckien et al. 2018). Evaluations of the latter are likely to show a more positive picture of the extent of adaptation than those that distinguish between policy plans (i.e. policy output) and implemented measures (i.e. policy outcomes). Another suggestion for policy evaluations is to assess the adaptations achieved by municipalities given the available solution space. Was there policy for adaptation, or did the implementation depend on a policy entrepreneur exploiting windows of opportunity? And did any rules and standards influence the scope for experimentation with innovative measures? Ultimately, it is a better understanding of the local that will increase the mainstreaming of adaptation in general.

Acknowledgements

The authors thank Joy Burrough for professionally language editing a near-final version of the text. We also thank our respondents for the interviews. Finally, the Dutch Delta Programme is thanked for subsidizing the research underlying the present study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 We note that entities are actors, such as individual programme managers or municipal departments; examples of activities are interactions, habitual actions and routines.