Opening the Black Box of Integrated Urban Development Strategies: On Causal Mechanisms and Policy Theories

Abstract

This chapter is devoted to analysing the policy theory used by local plans to implement integrated policy mixes. Based on the theory-driven evaluation and social mechanism perspectives, the chapter analyses the causal process established by local plans to ensure the link between objectives and their expected results; and, therefore, the relation between goal and implementation tools. From an analytical point of view, four main causal processes are proposed as combinations of two main aspects. First, goals according to the classical distinction in urban policies between context and actors. Second, the causal mechanism behind policy tools is distinguished according to two main types: oriented to modify opportunity structures and to change actors’ preferences or resources. These causal processes are identified by analysing each policy action implemented in local plans portfolios. The comparison between URBAN and URBANA programmes shows a progressive change from causal processes based on motivational mechanisms targeted at specific groups to re-equilibrate social inequalities to causal processes centred on context improvements to enhance their attractiveness.

Introduction

Integrated urban development strategies promoted by the EU aim to improve the living conditions of residents within the urban areas targeted by them. But how do they intend to do that? The general policy frame indicates that integratedness between policy sectors and the involvement of different agents are essential to attain the sustainable development objectives established in this kind of initiative. However, it does not specify how these objectives will be achieved. Knowing the relative importance of different objectives or actions across different policy sectors is very significant because it allows knowing what challenges are intended to address, and, therefore, how the project defines improvement in living conditions. However, it does not outline how the project implementation intends to achieve the proposed objectives or how policy actions will achieve them. Such approaches turn the projects into ‘a black box’ in which the causal mechanisms that link objectives and results are unknown (Astbury & Leeuw, 2010), limiting the possibility of appropriate evaluation processes and policy learning.

In this regard, the theory-driven evaluation perspective sustains that it is necessary to know the theory behind the projects, the explanatory mechanisms underlying their policy actions connecting goals and outcomes (Weiss, 1997). More specifically, this involves analysing which causal processes link objectives and expected outcomes, and, specifically, how the instruments (or policy tools) used to implement the projects will activate the behaviours or situations that will make it possible to achieve the proposed objectives.

This task faces at least two challenges. On the one hand, a methodological challenge since analysing the projects as a whole does not allow identifying these causal mechanisms. Therefore, more specific units must be ‘broken down’ and analysed. This is the strategy developed by the comparative urban policy portfolio analysis approach (CUPPA), which we will apply here at the level of the policy actions undertaken by all projects included in the URBAN and URBANA Initiatives. On the other hand, an analytical challenge, meaning the perspective we shall adopt to analyse (reconstruct) the logic that links objectives and results through actions. For this, we will adopt the situational perspective.¹ Basically, this assumes that social phenomena, in our case, the expected outcomes of projects in terms of improvements in residents’ quality of life, should be analysed from the targeted population or stakeholders’ point of view (their interests, beliefs, and resources) situated within a given context or structure of opportunities (be it physical, social, or cultural). The action taken by the targeted population involves a combination of these factors. Therefore, the policy instruments proposed to produce the situations or behaviours necessary to achieve the objectives of the projects suppose acting on some of these elements. Thus, as the behavioural turn in public policy proposes, analysing theories behind policies involves ascertaining their premises regarding how the policy instruments deployed will influence some of these elements (interest, resources, beliefs, opportunities) and lead to the achievement of the objectives set (Scheneider & Ingram, 1990).

Integrated Urban Development Projects as Policy Mixes: Objectives and Causal Processes

Urban integrated strategies suppose policy mixes combining different types of objectives and policy instruments to cope with the complexity of urban development, especially the sustainable model proposed by the EU. To understand the proposal projects made to attain these objectives, their ‘policy theory’ should be studied, understood as the set of causal processes that link the established objectives with their expected outcomes (Chen, 1990; Weiss, 1998). This link is established through the policy actions planned and implemented. These involve using specific policy tools to activate the behaviours or situations necessary to achieve the objectives and produce the expected outcomes. Therefore, different policy tools are identified according to the causal mechanism that would explain such activation, according to their behavioural assumption about how they will influence and produce these behaviours or situations (Bemelmans-Videc et al., 1998; Scheneider & Ingram, 1990). Therefore, each policy action could be analysed as a specific combination of policy objectives set and causal processes to achieve them, and these, in turn, according to the policy tool used and its underlying causal mechanism. These mechanisms would explain why and how a policy tool will trigger behaviours or situations that enable the achievement of the proposed objectives, at least in the way proposed or implemented by the project. A classical example, to achieve the aim of improving the economic activity of the neighbourhood, financial support is given to economic agents, underscored by the idea that such support will modify their pay-off about creating or improving an economic activity by reducing their costs, causing them to do so, and, by aggregation, this would increase business density in the neighbourhood or reactivate the economic activity of existing businesses.

From this perspective, the analysis of causal processes of the policy actions included in the projects would enable a ‘reconstruction’ of the ‘policy theory’, the underlying intervention strategy, as specific combinations of objectives and policy tools proposed in policy actions included in the integrated project understood as a policy mix. This theory would establish the framework to tackle their evaluation: did the causal processes established achieve the proposed objectives? (Rogers, 2008). Similarly, the overall strategy of a programme could be ‘reconstructed’ through this analysis of policy actions carried out by the local projects developed (Navarro & Rodríguez-García, 2020). Therefore, our proposal implies:

$$\begin{aligned} {\text{Intervention}}\, & {\mkern 1mu} {\text{Strategy}}\,{\mkern 1mu} ({\text{policy}}\,{\text{theory}})\,{\mkern 1mu} {\text{of}}\,{\mkern 1mu} {\text{a}}\,{\mkern 1mu} {\text{project}} \\ & = {\text{policy}}\,{\mkern 1mu} {\text{mix}}\,{\mkern 1mu} {\text{as}}\;{\mkern 1mu} {\text{a}}\,{\mkern 1mu} {\text{combination}}\,{\mkern 1mu} {\text{of}}\,{\mkern 1mu} {\text{policy}}\,{\mkern 1mu} {\text{actions}} \\ \end{aligned}$$

where:

$${\text{Policy}}\,{\text{action}} = {\text{f}}({\text{objective}},{\text{causal}}\,{\text{processes}})$$

$${\text{Causal}}\,{\text{process}} = {\text{f }}[{\text{policy}}\,{\text{tool}}\,({\text{a}}\,{\text{causal}}\,{\text{mechanism}})]$$

Basic Policy Actions in Integrated Urban Development Policy Mixes: Combining Objectives and Policy Tools

Which types of objectives and causal processes are common to place-based integrated urban development projects? There are two major types of objectives, which are somewhat independent of the sector of public policy involved. On the one hand, projects often pursue objectives related to neighbourhood improvement (the urban area where they are applied), based on the understanding that this sets up a specific structure of opportunities for their residents. There is a ‘neighbourhood effect’ justifying the area-based project; therefore, its objectives can be geared towards modifying it, both in its spatial dimension, as well as in the socio-economic composition, communitarian life, or environmental dimension. For example, this objective could be improving environmental quality, employment opportunities, public space or pedestrian zones, accessibility to the whole city, signage for heritage or tourist sites, or creating or improving commercial areas or new centres to provide services.

On the other hand, objectives could try to improve the situation of residents or specific groups. For example, employment training and skills, help and information for integration into the job market, developing participatory skills or promoting certain habits or lifestyles regarding health, academic education, community life, or the environment. Thus, although the policy frame of the programmes focusses on specific urban areas, their objectives can be oriented both towards socio-spatial (neighbourhood) improvement and towards residents directly (their resources, lifestyles,…). In other words, analytically, the target of policy actions could be the residents or other agents developing their activities in the targeted area (business, associations,…) or neighbourhoods as a structure of opportunities for residents of these other actors. Therefore, this distinction between neighbourhoods and actors is similar to the classical difference between place and people in urban policy analysis (Holland, 2015).

Regarding policy tools, two major types could be defined according to their underlying causal mechanism; their assumptions of how the behaviours or situations needed to achieve the proposed objectives will be activated. On the one hand, those that try to do so by modifying the opportunities for action available to residents or other agents without altering their capacities, interests, or beliefs. These would be contextual mechanisms that essentially involve the provision of unconditional incentives, as new opportunities are offered (a social centre, more buses, a park, traffic rules,…), but their use—or not—is subject to the reasons the stakeholders have for this (Dowding, 1991). Thus, these mechanisms seek to modify the physical, social, or decision-making context to expand—or limit—the repertoire of actions residents and other agents in the urban area can develop.

On the other hand, policy tools could try to modify residents’ motives or reasons for activities that would achieve established objectives, whether through persuasion processes about the value of certain behaviours or lifestyles (healthy habits, inter-ethnic relations, the importance of education, gender equality, environmental quality,…), or through resources that facilitate an action development or modify the order of preferences with respect to it (competences and skills for those who want to increase their employability, subsidies to initiate or improve a business, for example). In this case, these would be ‘motivational mechanisms’, as they seek to modify the motives or reasons residents have to develop certain attitudes and behaviours, influencing their ‘mental state’ (their interests and beliefs) or their capabilities (resources of various kinds: economic, information, cognitive, skills,…..), without taking action on their context (expanding or limiting their opportunities for action).²

This approach assumes that the achievement of project objectives, and their potential impact, depends on the exposure of residents or other agents to the causal processes underlying project policy actions. Contextual mechanisms involve contextual exposure to the project. The underlying premise is that changes—improvements—in the neighbourhood will generate changes—improvements—among residents because they have more—or better—opportunities at their disposal that they can use, or even because exposure will produce changes without needing to create a specific action. For example, improving public transport means more mobility opportunities available to use. Improving the environmental quality or reducing physical and social disorganisation (better urban furniture, lighting, less social conflict) can impact residents’ health or sense of security.

Motivational mechanisms, on the other hand, involve changing the ‘motives’ of residents or other agents to develop—or not—certain behaviours or lifestyles. Therefore specific exposure is required to the instruments that promote these. While contextual mechanisms involve unconditional incentives, motivational ones come closer to the idea of selective incentives, which could be ‘hard’ (resources of different kinds) or ‘soft’ (ideas, information, persuasion,…) in nature. For example, subsidies granted to certain companies or initiatives, the acquisition of job skills and competencies, or awareness-raising on certain issues among those attending courses or activities included in project activities.

There might be some affinity between the pursued objective and the policy tool used, so neighbourhood-oriented objectives tend to use contextual mechanisms, and residents-oriented objectives apply motivational mechanisms. However, analytically speaking, it does not have to be this way. Motivational mechanisms might aim to improve the neighbourhood. For example: the extension of pro-environmental behaviours among residents could be considered a means to improve the environmental quality of the neighbourhood; raising awareness of inter-ethnic or inter-generational relationships can improve the level of social cohesion in the neighbourhood (as neighbour relations); support for businesses could expand employment context—opportunities—in the neighbourhood. In the same way, contextual mechanisms can target directly residents, such as actions to improve social integration through a new centre or infrastructure in the neighbourhood to older people or children, or regulating vehicle access to specific areas. Therefore, analytically there are four major types of policy actions in area-based integrated development projects according to how causal objectives and processes are combined (Table 6.1).

Table 6.1 Policy actions as causal processes: objectives, policy tools, and causal mechanisms (Examples of policy actions in each type)

The presence of these four types of actions, and in particular their causal processes, would show the policy theory proposed by the projects as policy mixes, their strategy to improve the quality of life among residents (see Chapter 1). In addition, the combined analysis of all projects would allow reconstructing the programme’s intervention strategy, that is, the policy frame by which they are actually developed, somewhat independently of their normative or programmatic proposal (see Chapter 2).

The design and implementation of place-based integrated strategies as policy mixes are given by at least three aspects or ‘starting conditions’ of the projects (Navarro, 2016). Firstly, the ‘repertoire’ of objectives and policy tools established by each policy or programme through its policy frame, since projects must show a degree of external coherence with the framework they establish. This ‘starting condition’ should explain programme differences (Navarro & Rodríguez-García, 2020). Secondly, on account of the characteristics of the territory where the intervention takes place (local needs, culture, and capacities, stakeholders present, previous experiences, …), the policy frame must be adapted to the territorial area where policy actions take place. This factor could explain differences between local projects and, thus, within programmes (Navarro et al., 2019). And finally, the institutional environment—the local government system—in which local authorities are situated. This gives them different capacities or even guides their preferences, agendas, and interactions with other institutional and non-institutional actors to conform muti-level governance processes (Navarro, 2009; Navarro et al., 2008; Sellers, 2002). These sources of diversity are a consequence of the place-based approach adopted by the EU cohesion policy and its ‘meso-level approach’ combining top-down and bottom-up approaches to design and implement nested integrated urban strategies in the framework of EU policy (Crescenzi & Rodríguez-Pose, 2011; Newig & Koontz, 2014). This chapter will focus on the differences between URBAN and URBANA programmes. The repertoire of objectives and instruments included in their policy frames is fairly broad. In addition, they share a similar policy frame based on the urban integrated strategy proposed by the EU, except for the importance of environmental sustainability as an objective in the case of URBANA versus URBAN (see Chapters 1 and 4). We aim to show how to analyse the strategy of the projects and whether our proposal can show differences between their policy frames understood as the policy theory applied to improve the quality of life in urban areas.

Integrated Urban Development Strategies in the URBAN and URBANA Programmes: Changes in Their Policy Theory

Which strategy do URBAN or URBANA Initiative projects deploy? Are there any differences between them? To provide evidence to these questions, we have studied the policy actions included in the projects applying the CUPPA approach. We have analysed all policy actions implemented according to the evaluative reports. Therefore, we will examine the actual strategy implemented by projects. We have excluded those policy actions concerning project management, as they are generally referred to somewhat generically and similarly across all projects. We will analyse a total of 514 policy actions for 82 projects, 205 in 39 URBAN projects, and 309 in 43 URBANA projects.

The most relevant objectives and policy tools for each action have been identified, following the classification proposed in the previous section. Most of the measures analysed are more or less equally focussed on residents and the neighbourhood (49% and 51% of the total actions, respectively). However, there is a greater tendency to use policy tools that involve contextual mechanisms than motivational ones (61.1% and 38.9%, respectively). In addition, there are differences between the two programmes (Fig. 6.1). Actions taken within URBANA projects tend to focus somewhat more on the neighbourhood than in URBAN projects (54.4% and 45.9%, respectively), and causal processes that apply contextual mechanisms are also more prevalent than in URBAN projects (64.4% and 56.1%).

Fig. 6.1

(Percentages of the total number of actions in each programme. Source Own elaboration based on Urban Impact Project databases)

Objectives and mechanisms of policy actions implemented in URBAN and URBANA projects

But which instruments are used to achieve which kinds of objectives? How are objectives and policy tools combined in policy actions? As indicated above, there is likely to be some affinity between objectives and policy tools. When the objectives are resident-oriented, motivational and contextual mechanisms (56.7% and 43.3%, respectively) are often used, more or less equally. However, contextual mechanisms are clearly dominant when the objective is to modify the neighbourhood as an opportunity structure for residents (21.8% and 78.2%, respectively). The pattern is quite similar between URBAN and URBANA projects.

Nevertheless, the relationship between contextual objectives and contextual mechanisms is more pronounced in the latter: in the case of URBAN projects, 68% of the neighbourhood-oriented actions apply contextual mechanisms, whereas this percentage climbs to 85% in the case of URBANA projects. Thus, comparing the two programmes, we find similar objectives (improving the neighbourhood) are sought using different causal processes. Therefore, knowing project content—such as policy sectors, priorities, or objectives—is not enough to understand the strategy deployed by a policy action and integrated urban initiatives at a more aggregate level.

But what is the weight of each combination of objectives and causal processes in the policy mix defined by the project portfolio? What kind of strategies do they deploy? A greater proportion of all policy actions analysed seek to improve the neighbourhood as a structure of opportunities through contextual mechanisms (39.9%). Second are those seeking to train or persuade residents to take a specific action that supposes an objective accomplishment or facilitates its achievement (27.8%), followed by those seeking to do so by changing the opportunities available to do the action needed to accomplish an objective (21.2%). Finally, a small percentage of the policy actions aim to improve the neighbourhood by acting on the reasons of residents or other agents to develop the activity necessary to achieve the proposed objective (11.1%). This portfolio means that the logic of the intervention lies primarily in contextual exposure to the improvements that the project can bring to the neighbourhood.

The comparison between the URBAN and URBANA shows that this contextual logic is more relevant among the latter. There are changes in their objectives and, especially, in the mechanisms used to achieve them. Firstly, the weight of actions based on contextual mechanisms to improve the neighbourhood increases from 31% in URBAN projects to 46% in URBANA projects (Fig. 6.2). Secondly, there is a clearer specialisation of the URBANA programme around two types of actions: motivational mechanisms for resident-oriented actions and contextual mechanisms for neighbourhood-oriented actions. The former has more or less a similar weighting in URBAN and URBANA projects (29% and 26%, respectively). However, in URBANA, causal processes that seek to improve the neighbourhood using motivational mechanisms have a lower weighting (from 14.6% to 8.7%), as do those seeking to facilitate certain attitudes or behaviours through contextual tools (24.8% to 18.8%).³

Fig. 6.2

(Percentages for each type of action [objective and mechanism] within the total number of actions in each programme. Source Own elaboration based on Urban Impact Project databases)

Strategies deployed in URBAN and URBANA projects: objectives and mechanisms

These results mean URBAN projects deployed different types of causal processes in a more balanced way to achieve their objectives than URBANA projects. These tries combined to a greater extent the effects that would be derived from specific exposure involving motivational mechanisms and from contextual exposure to neighbourhood improvements. Hence, URBANA projects place greater emphasis on the second type of exposure: improving residents’ living conditions will result from their contextual exposure to improvements promoted by projects in the neighbourhood as opportunity structures. This implies a significant change in the intervention strategy for integrated urban development that is not directly evident from a normative analysis of the integrated strategy policy frame defined by the European Union or the study of programmes implemented in Spain. Although their goals and implementation preferences (policy frames) may seem similar, the strategy actually deployed appears to be different.

In this regard, as indicated above, the contextual strategy essentially involves the provision of unconditional incentives in the form of contextual exposure. The central idea is that there will be more and new opportunities for residents that will eliminate or mitigate the negative consequences of the ‘neighbourhood effect’, which justified the implementation of the project. However, contextual exposure does not necessarily ensure that residents will ‘use’ such opportunities or, more generally, that all residents are exposed in the same way to the new ‘positive neighbourhood effect’ promoted by the projects. This strategy does not pay attention to the social composition of the neighbourhood, the fact that resources, interests, or beliefs are not equally distributed among its residents, or even that their lifestyles involve different degrees and forms of exposure to the neighbourhood. More contextual opportunities do not mean that all residents use (or take advantage of) them equally.

For example, a new service facility in the neighbourhood equally reduces the costs of use for all residents (at least, in terms of displacement), being able to influence usage and the resulting benefits. That is the main causal mechanism underlying this widespread type of policy action in urban integrated strategies. Nevertheless, this does not mean that the reasons for using such services change: those related to the need for them or others about the relationship between citizens and public services, such as trust in them or those who provide them, their reputation, or information barriers that could explain the use of public services regardless of their territorial localisation or accessibility. Some analyses show that the effects of contextual exposure to the neighbourhood and its changes due to, for instance, place-based policy actions, are different according to the recourses, capabilities, and lifestyles of different social groups (Zapata & Navarro, 2017; Zapata et al., 2019). Other studies show that the effects of the projects become more evident when contextual exposure and specific exposure are combined, involving simultaneous exposure to contextual and motivational mechanisms (Navarro, 2016).

From Policy Frames to a Theory of Integrated Urban Development: Some Analytical and Methodological Foundations

In this chapter, we have shown another possible application of the CUPPA approach: analysing the theory underlying integrated urban strategies. The analysis has shown differences in the intervention strategies that cannot be captured by analysing the policy frame of the programmes applying the normative or the programme analysis approaches. The analysis has not been based on the economic weight or importance of different areas or sectors of public policy. Instead, strategies have been defined in response to the causal processes established to achieve the proposed objectives and their assumptions of how they would affect residents’ quality of life. Therefore, we now know in more detail what the projects have sought to do and how they have sought to achieve their pursued outcomes, thereby indicating how the expected effects and impacts might occur. The policy theory behind integral strategies has been reconstructed by analysing their causal mechanisms and processes.

This CUPPA application has helped analyse the character of EU-integrated urban initiatives in Spain and their change since 1993. In this respect, the analyses show that, in the Spanish case, the integrated urban development strategy promoted by the European Union has changed from a more varied mix of different types of policy actions to greater specialisation in a contextual strategy. Put another way, the focus has been placed on changes in the neighbourhood as a driving force for improving residents’ quality of life than on policy actions oriented to modify actors’ capabilities, resources, or beliefs. This could indicate a change in policy orientation from a redistribution approach to a more distributional and developmental approach in urban policies. This could coincide with the shift from convergence towards competitiveness policy goals in EU cohesion policy during the period analysed (McCann, 2015) and the move away from the anti-poverty approach of the EU urban initiatives in the 90s (Zimmermann & Atkinson, 2021), which could mean a shift from the ‘revitalising the neighbourhood’ towards ‘creating competitive places’ policy frames mentioned in previous chapters.

In this chapter, analytical bases and methods have been provided—and applied—to analyse these changes from a comparative perspective and at the local integrated strategy scale (and their policy actions). Those are based on the proposal to study urban policies as multi-level policy mixes and the CUPPA approach as a research methodology (Navarro & Rodríguez-García, 2020). These elements provide some analytical foundations to elaborate a theoretical perspective of integrated urban strategies promoted by the EU that could be analysed empirically from a comparative perspective beyond the study of good practices or the normative analysis of policy frames and their limitations (see Chapter 1).

This trend towards the contextual strategy in the policy theory of urban integrated strategies might face risks arising from potential heterogeneity in the socio-spatial contexts (neighbourhoods) in which it is developed. If the ‘neighbourhood effect’ that justified the intervention might not affect all its residents equally, the same could be concluded regarding the ‘neighbourhood effect’ promoted by the contextual exposure underlying this strategy. Instead, in the case of URBAN projects, this contextual strategy is combined with policy actions focussed on residents’ motives, interests, beliefs, and resources. These actions could address better the heterogeneity existing in neighbourhoods or other socio-spatial scales—municipalities and functional urban areas—in which the integrated strategy could be applied. The mechanism behind this ‘contextual strategy’ in integrated urban policies could promote heterogeneous effects at the territorial level—differences between targeted territories—and at the individual level—different types of residents in targeted places according to their resources, lifestyles, ….

Nevertheless, more comparative analyses are needed to conclude this trend in the policy behind the EU-integrated urban initiatives and their effects on quality of life and socio-spatial cohesion across Europe. These analyses could also include other explanatory sources previously mentioned as starting conditions: the traits of places where projects are applied and the institutional framework in which municipalities define their strategy, including the planning traditions or the new urban policies that state members are launching across the EU (Nadin & Stead, 2008; Zimmermann & Fedeli, 2021).