Community-based initiatives in the urban realm what conditions their performance?

Abstract

A city is a place where many initiatives, people, and social and urban challenges meet. This article brings together the cumulative knowledge of eight researchers who have been studying community-based initiatives (CBIs) via case studies in various countries. In some countries, citizens were not satisfied with government-provided services, or services were lacking. Governments in other countries faced budget cuts to their public services, which led to a strong call for citizens to take matters into their own hands. There is a lack of research investigating the performance of CBIs and explaining their success and failure. The authors systematically analyze their recent case studies through qualitative comparative analysis (QCA) and try to explain under which conditions CBIs lead to high performance. One of the key findings of this analysis is that proximate conditions related to the CBIs—strong organizational capacity, democratic structure, and leadership—are important for high performance. However, these conditions are not sufficient on their own. Community-based initiatives need a conductive environment to achieve high performance; in our cases, government support and a heterogeneous community or a supportive government attitude was key.

Keywords:

• qualitative comparative analysis (QCA)
• community-based initiatives performance
• proximate and remote conditions

1. Introduction

Citizens are no longer seen as passive consumers, but co-creators of public services (Brandsen, Trommel, and Verschuere 2017; Garcia 2006; Haus, Heinelt, and Stewart 2004; O’Hare 2018; Pierre 2016). A specific phenomenon within current co-creation discourse is community-based initiatives (CBIs). CBIs have gained increasing attention in various research fields and disciplines over the last two decades, literature emphasizing citizens’ self-organizational capacity (Edelenbos and van Meerkerk 2016; Healey 2015; Haus, Heinelt, and Stewart 2004; O’Hare 2018; Van Meerkerk, Kleinhans, and Molenveld 2018). In CBIs, citizens take the lead and collectively initiate and implement projects aimed at providing public goods or services for their own community (Healey 2015; Igalla, Edelenbos, and van Meerkerk 2019a). Community-based initiatives have emerged all over the globe in developed countries, due to budget cuts and state retrenchment in various sectors (health care, social health care, energy, urban livability, etc.), and in developing countries, due to weak state and governance structures, corruption, and scarce (financial) resources (Brandsen, Trommel, and Verschuere 2017; Chaskin 2001; Teasdale 2012).

CBIs are often praised for their capacity to enhance legitimacy, solve societal problems and issues, foster (social) innovation and achieve sustainability (Attuyer 2015; Mulgan 2012; Torfing, Sørensen, and Røiseland 2019; Feola and Nunes 2014; Celata, Dinnie, and Holsten 2019). However, there are also doubts about their impact and performance (Brandsen, Trommel, and Verschuere 2017; Celata, Dinnie, and Holsten 2019; Seyfang and Longhurst 2013; 2016). The acclaimed performance often remains largely hypothetical, as literature with empirical proof is still quite scarce. Although there are some publications on the performance of CBIs (Bagnoli and Megali 2011; Bailey 2012; Ramirez 2005), these are often dominated by single-case studies. In addition, some first large N studies CBIs and their performance are available (Feola and Nunes 2014; Igalla, Edelenbos, and van Meerkerk 2019a), especially in the field of sustainability and renewable energy as CBIs are quite emergent in those fields (Seyfang and Haxeltine 2012; Celata, Dinnie, and Holsten 2019; Landholm et al., 2019; Celata and Sanna 2019). Although, more information on the performance of CBIs has become available, we still need systematic insight and knowledge into what (combinations or configurations) conditions the performance of CBIs (Igalla, Edelenbos, and van Meerkerk 2019a). We aim to unravel what explains the performance of CBIs by using fuzzy set qualitative comparative analysis (fsQCA). This method allows us to compare multiple cases — 17 in our study — and find paths that explain the performance of CBIs in the urban realm. We make a split between remote conditions, which are context conditions, and proximate conditions, which are conditions close to the study object, including the CBIs themselves. The following main research question guided this study: In which configurations of remote and proximate conditions are community-based initiatives more likely to reach performance?

In section two, we develop our conceptual framework. This framework will define CBIs, as well as their performance and potential explanatory factors. In section three, we discuss the research methodology, which is a fsQCA, and provide an overview of the cases involved. In section four, we provide the results of our QCA and show patterns in factor configurations explaining the performance of CBIs. We end our article by drawing core conclusions and outlining avenues for future research in the field.

2. Conceptual framework

In this section, we elaborate on the core ideas that constitute our conceptual framework. We first define CBIs, then we discuss performance and how it can be measured. We conclude this section by highlighting explanatory factors that could potentially influence performance.

2.1. Defining CBIs

CBIs can be approached as a specific kind of co-production and civic engagement (Van Meerkerk, Kleinhans, and Molenveld 2018). In general, different forms and degrees of civic engagement can be distinguished, often displayed by participation ladders like the ones Arnstein (1969) and Fung (2006) use to indicate the depth of participation and, therefore, degree of influence citizens can have in decision-making (Van Tatenhove, Edelenbos, and Klok 2010). Stakeholder consultation or participation implies that any group or individual who can be, or is, affected by policy programmes, plans, and projects is invited to have a say in the decision-making process (Irvin and Stansbury 2004). Citizen involvement is often conditioned by rules and procedures set by government institutions in these kinds of consultations (Edelenbos and van Meerkerk 2016); in other words, people are invited in an arena structured by the government. CBIs are a different kind of stakeholder engagement, as citizens take the lead to determine the rules and procedures by which they collectively initiate and implement initiatives aimed at solving societal problems and issues (Igalla, Edelenbos, and van Meerkerk 2019a). This makes them distinct from other forms of co-production and stakeholder engagement.

These initiatives have become a marked trend in many countries all over the world (Bailey 2012; Healey 2015; Feola and Nunes 2014; Seyfang and Longhurst 2013; 2016). Citizens control the aims, means, and actual implementation of their activities (Healey 2015), from running a community center, setting up a cooperative or charity to provide community-led care services for the elderly in the area, or creating environmental initiatives for local renewable energy (Edelenbos and van Meerkerk 2016). Community-based initiatives are fundamental bases to establish communities. There are many cases of CBIs over the world, in which communities are doing their own initiatives to improve their neighborhoods, which are parts of the everyday life of communities (Seyfang and Haxeltine 2012; Bartels 2013).

CBIs can be defined in various ways, also depending on the context in which these CBIs take place for example grassroots in sustainability initiatives (Seyfang and Longhurst, 2016). In this same field Celata, Dinnie, and Holsten (2019, 910) have defined CBIs as “…a collective action initiated and managed by groups of individuals that feel they share a connection—whether of interest, place, lifestyle, culture, or practice—and have self-organized in order to implement projects to serve their community”. In this paper, we have based our definition on CBIs on a systematic literature review by Igalla, Edelenbos, and van Meerkerk (2019b; 2019a) revealing the following main characteristics of CBIs:

1. Community-based initiatives are often locally oriented, which means that local residents are the (current) driving force behind the initiatives. They mobilize volunteers from within the community and focus on community needs.

2. The CBIs provide and maintain an alternative form of traditional governmental public services, facilities, and goods;

3. They strive for autonomy, ownership, and control regarding decision making;

4. They are often linked to formal institutions, such as local authority, governmental agencies, and NGOs, especially for funding and facilitation;

5. They often develop their own business model to increase financial stability, but do not aim for private profitmaking (see also: Bailey 2012; Lepofsky and Fraser 2003; Llano-Arias 2015; Ornetzeder and Rohracher 2013).

Based on the systematic literature review by Igalla, Edelenbos, and van Meerkerk (2019a) and in line with conceptualization by others (Healey 2015; Celata, Dinnie, and Holsten 2019), we define CBIs as a form of self-organization in which citizens mobilize energy and resources to collectively outline and implement projects aimed at providing public goods or services for their community. We approach CBIs as a form of community engagement in which citizens mobilize capacities and resources to collectively define and carry out actions aimed at providing public goods or services for their community; citizens control the aims, means, and actual implementation of their activities.

2.2. Indicating performance

With the rise of CBIs, academic attention has also increased. Scholars wonder what these new initiatives mean and entail. Are they a different phenomenon than other forms of civic engagement (Healey 2015)? Other scholars question whether these initiatives actually perform well (Brandsen, Trommel, and Verschuere 2017; Feola and Nunes 2014; Celata, Dinnie, and Holsten 2019). Since CBIs have become a political alternative for governmental public services, performance is an important outcome to ensure the availability, legitimacy, and accessibility of these services to citizens. Igalla, Edelenbos, and van Meerkerk (2019a) revealed what kind of outcomes are the focus of studies on CBIs, which is shown in Table 1. Igalla, Edelenbos, and van Meerkerk (2019a) further found that many of the studies on CBIs focus on performance (70.8%). There is also lot of attention given to performance, especially in the (urban) governance network literature (Klijn, Steijn, and Edelenbos 2010; Meier and O’Toole Jr., 2002). Also, in the field of sustainability and transition we see various publications on conceptualizing and measuring output and performance. Feola and Nunes (2014), for example, have used both subjective and objective criteria to measure performance (which they call success of transition projects). For the latter, they use criteria such as: the number of members or people involved in the transition initiative (i.e. critical mass) and the duration of the transition initiative.

Table 1. Performance of CBIs.

Type of outcome	N
Social improvements of the community or its members	36
Economic improvements of the community or its members	15
Physical improvements of community	8
Environmental improvements	11
Increasing durability	18
Achieving organizational outcomes	11
Fostering safety	6
Enhancing level of education	8
Achieving legitimacy outcomes	5
Achieving (goal) effectiveness	4
Achieving public outcomes	3
Other	5

Source: Igalla, Edelenbos, and van Meerkerk (2019a).

Total: N = 130 (100%) – most studies identified more than one outcome.

Performance is often measured by a multi-categorical indicator combining aspects of effectiveness, efficiency, innovation, democracy, etc. (Igalla, Edelenbos, and van Meerkerk 2019b). Also Celata and Sanna (2019) conducted a multi-dimensional assessment of the environmental and socioeconomic performance of several community-based sustainability initiatives, with criteria such as: social capital, social inclusion, innovativeness, carbon reduction and efficiency and economic impact.

Our study focuses on the aspect of effectiveness, defined as the extent to which a CBI achieves its intended objectives (e.g. implementing services, fulfilling a societal demand, finding a solution to a collectively defined problem, etc.). Such an indicator is widely accepted in the literature (cf. Klijn and Koppenjan 2016). Non-performance or failure means that objectives were not reached, and no positive side-effects were gained.

Identifying explanatory factors

Knowledge about CBI is broad, and there have been few publications that distinguish different conditions for performance. Feola and Nunes (2014) have developed many (categories of conditions explaining the success of transition initiatives, such as: characteristics of the initiative, organization, resources, and context. Igalla, Edelenbos, and van Meerkerk (2019a) also revealed, through their systematic literature review, important factors that are mentioned in the literature explaining CBI performance. In their selection of publications, 57.3% of the studies mentioned explanatory factors. An overview is provided in Table 2.

Table 2. Factors influencing the performance of CBIs.

Type of factor	N
Network structure	22
Organizational capacity	20
Support of government	18
Support of non-profit sector	13
Leaders and leadership	11
Democratic structure	8
Strategies of initiatives	7
Pre-existing conditions	7
Motivation	4
Other	10

Source: Igalla, Edelenbos, and van Meerkerk (2019a).

Total: N = 130 (100%) – most studies identified more than one outcome.

We use the explanatory factors identified in Table 2 and conceptually develop these factors for the purpose of our comparative case study. As the cases described in this paper come from many different countries and contexts, we will follow Carsten and Wagemann’s (2006) two-step approach:

1. First, we will look for relevant remote conditions (in italics in Table 2), which are context conditions almost out of reach for conscious intervention by the actors, and more at a distance and a given.

2. Second, we will dive into proximate conditions (underlined in Table 2), which are closer to human action, both actor-based and process-related events. These often display the causal mechanisms more clearly and more closely describe the circumstances of CBIs (Carsten and Wagemann 2006, 760–761).

We will first explain the remote conditions that show a causal connection to the success or failure of the CBI, then we will discuss the proximate conditions that relate to the CBI itself. By combining the remote conditions with the proximate conditions, we can build a stronger causal ‘model’. All the conditions and their operationalization can be found in Annex 1 (online supplementary material).

2.3.1. Remote conditions

In terms of country characteristics, we can say that some countries and their respective governments are more open to self-organization and supportive of these initiatives than others (Jepperson 2002; Schofer and Fourcade-Gourinchas 2001). Jepperson (2002) explains civic engagement by means of an important dimension: statism. Countries that score low on the statist dimension locate purpose and authority in society at large, with government seen more as an instrument and expression of society. Society retains collective agency and legitimacy, whereas the government has less independent legitimation (Jepperson 2002). The UK is typically seen as scoring low on statism, with its belief in small government and big society. New forms of civic engagement might encounter more difficulties trying to evolve in countries with a high level of statism, where the government is more likely to interfere in the private sphere and society (Schofer and Fourcade-Gourinchas 2001). This can have both positive and negative effects for the performance of CBIs. On the one hand, it is likely to bring about red tape that might hinder them; on the other hand, government support may be stronger.

Bearing the abovementioned dimension in mind, we define our first two remote conditions as government attitude. One of the factors that Igalla, Edelenbos, and van Meerkerk (2019a) show as being important for CBIs is government support. As CBIs often operate in the institutionalized public domain with public rules and regulations, they often interact with governments (cf. Healey 2015; Perkins, Brown, and Taylor 1996). Seixas and Berkes (2009) found – in their comparative case study on 10 ‘successful’ CBIs in different South-American countries – that most of these CBIs had supportive relationships with government organizations. They describe whether the political context is favorable for CBIs in terms of attitude of (local) public institutions and conducive legal and policy frameworks (e.g. Korosec and Berman 2006; Llano-Arias 2015).

In this paper, we divide government support into government attitude and conducive legal and policy frameworks. In terms of attitude, governments may provide a range of services and support functions for CBIs, including seed money and grants, networking and marketing, technical and managerial expertise, and advisory services to navigate through bureaucratic tangles (Korosec and Berman 2006; Llano-Arias 2015). A stable boundary spanner supporting a CBI increases its chances of success, because he or she can advocate for the initiative or move resources (Nederhand, Bekkers, and Voorberg 2016). The supportive action of governments is important for the potential future of CBIs, as this can boost their start and growth in scale and scope (Healey 2015). The same argument goes for the legal and policy frameworks that institutionalize the open attitude within the political-administrative government system that likely contribute to the success of CBIs. Conducive means that there are institutions and frameworks that support consensus-building and resource-sharing (Börzel and Risse 2000).

The third and fourth remote condition have to do with the neighborhood characteristics. First is the degree of urbanization; some studies associate ‘the urban’ as a breeding ground for socially innovative CBIs (Brandsen et al. 2016; Moulaert et al. 2010). On the other hand, there are also examples of (innovative) rural community development initiatives (e.g. Global Ecovillage Network, as in Kunze and Avelino (2015), and Via Campesina, as in Juarez et al. [2016]). It seems useful to assess if and to what extent this ‘degree of urbanisiation’ influences the outcomes of CBIs. Urban is typically contrasted to rural, but in reality, there is a scale, and typologies have been developed to define the continuum from rural to urban. Losada et al. (1998), for instance, use such a typology in the following manner: urban, suburban, peri-urban, and rural. Characteristics that can be used to define what category a certain defined space or area falls into are as follows (Iaquinta and Drescher 2000):

• Demographic component: the population size and density; larger and denser is associated with more ‘urban’.

• Economic sectoral component: an agricultural labor force is associated with rural, a non-agricultural labor force is associated with suburban, while peri-urban is a mix of both.

• Social-psychological component: a consciousness of the residents that they live in a rural area, something in between and urban and rural area (peri-urban), on the edge of the city (suburban), or in the central part of the city (urban).

Second, we use the heterogeneity of the community. The heterogeneity of stakeholders and community members is associated with a negative impact on participation, as the bonding and binding capacity of communities is not that strong: ‘In heterogeneous communities people are often less likely to participate due to divisions of language, tenure, income, gender, age or politics, than in less diverse communities’ (Botes and Van Rensburg 2000, 40). However, other scholars see more value in the diversity of cultural, social, and personal backgrounds of participants, which could lead to more creativity and innovation (Bekkers, Edelenbos, and Steijn 2011).

2.3.2. Proximate conditions

In terms of proximate conditions, we distinguish actor- and process-related conditions. Actor-related conditions are: network strength, leadership, motivation, and capacity.

Network strength is based on Putnam’s terminology of social capital (Putnam 1995, 2000). Social capital is defined by Putnam (1995, 664-665) as “features of social life – networks, norms, and trust – that enable people to act together more effectively to pursue shared objectives”. Social capital deals with the connections between members within a specific community, within a CBI, and in its relationships with actors outside the CBI that form a network that facilitates the achievement of collective goals (Coleman 1988; Purdue 2001). Social capital has been categorized into the common distinction between bonding, bridging, and linking social capital (Newman et al., 2008). Bonding social capital refers to both “trusting and cooperative relationships between members of a network who see themselves as being similar, in terms of their shared social identity” (Woolcock 2001, 654–655). Bridging and linking social capital refer to relationships that lead out of the CBI. In the case of CBIs (Igalla, Edelenbos, and van Meerkerk 2019b), bridging ties can refer to ties connecting target groups or other associations operating in the community (cf. Somerville and McElwee 2011; Szreter 2002). Linking ties are often present through connections with (local) government (agencies) and other institutions, such as funding agencies (cf. Dale, Ling, and Newman 2010; Szreter 2002).

The second factor, leadership, has been considered in many studies a key variable influencing the success or failure of an initiative. Emerson, Nabatchi, and Balogh (2012) describe that a leader has the potential to initiate and mobilize resources in the broad sense of the word. They can mobilize and inspire people, activate or access resources, and possesses skills that can help build collaborative and strategic alliances. If a leader is entrepreneurial (i.e. explores new projects, experiments, and develops new ideas (Gupta et al. 2010)), this will likely lead to more revenue options and performance. Leadership deals with the qualities and activities of individuals who form and manage the CBI (van Meerkerk and Edelenbos 2018). Leadership in general implies that people are “influencing the activities of an organised group toward goal achievement” (Renko et al. 2015, 54). Leadership can be focused on the internal organization and its relationship with the environment (Tushman and Scanlan 1981b, 1981a). Two well-known subsets of intra-organizational leadership can be distinguished in this regard: transformational and transactional leadership (Bass 1999). Regarding external orientation, boundary spanning leadership is often mentioned regarding external orientation (Tushman and Scanlan 1981a). Both have been distinguished in the literature on CBIs as well, and coined as community leadership (Bénit‐Gbaffou and Katsaura 2014; Purdue 2001; van Meerkerk and Edelenbos 2018).

Transformational leadership is focused on inspiring people to raising awareness, following a vision, and obtaining certain goals, values, and performances (Moynihan, Pandey, and Wright 2012). Transactional leadership, on the other hand, is focused on management strictly through clarifying and determining rewards for meeting expectations (Bass et al. 2003). Boundary spanning leadership is, moreover, considered important to creating a better fit by enhancing and adapting performance to the environment (Aldrich and Herker 1977). This type of leadership is vital for gaining necessary resources and linking the organization to external developments that might create opportunities to make performance more effective, innovative, and efficient (Tushman and Scanlan 1981a, 1981b).

The third factor, motivation, is a widely researched topic; for example, studies have been done on employees’ motivation to work for the public sector (public service motivation, Perry 1996). A distinction is made between intrinsic and extrinsic motivation, where the first relates to personal drive, and the latter relates to material rewards (money, services, and goods) attained in exchange for contributions. Some research has been conducted in the field of participation and the motivation to engage in participation processes (Alford 2002; Lowndes 2005; van Eijk and Steen 2014). When people have an intrinsic motivation to engage in CBIs, or participate in general, then people ‘like to’ get engaged (Lowndes, Pratchett, and Stoker 2006). Citizens have a sense of attachment to the neighborhood, issue, or group that reinforces commitment. This commitment is rooted in doing something they like to do and identify with. In CBIs, volunteers are more likely to be intrinsically motivated. Alford (2002) mentions various intrinsic motivations for community engagement:

• Values; expressing humanitarian values or altruistic concerns.

• Understanding; increasing one’s knowledge of the world and developing skills.

• Enhancement; developing psychologically and enhancing self-esteem.

• Career; volunteering to gain experiences beneficial to their careers.

• Social; fitting in and getting along with social groups.

• Protective; coping with inner anxieties and conflicts.

The more present these motivation factors are, the more inclined people are to devote themselves to a CBI’s cause.

The fourth and final proximate condition is organizational capacity, which implies “the ability of an organization to accomplish its mission effectively” (Eisinger 2002, 115). Two dimensions are considered to be relevant for CBIs (Igalla, Edelenbos, and van Meerkerk 2019b): financial and human resources (see also Foster-Fishman and Long 2009; Hassink et al. 2013; Healey 2015; Sharir and Lerner 2006; van Meerkerk, Boonstra, and Edelenbos 2013). Human resources are about the volunteers who are active in the CBI. Igalla, Edelenbos, and van Meerkerk (2019b) argue that volunteers provide time, energy, ideas, experience, and skills that increase the capacity of CBIs to achieve the desired outcomes (see also Healey 2015). Financial resources are also necessary to pay for buildings, fund professional staff, invest in communication and exposure, and handle other costs (Foster-Fishman and Long 2009).

In terms of process-related conditions, we can distinguish democratic structure and NGO support. Democratic structure has many meanings in terms of CBIs and networks. Sørensen and Torfing (2009), for instance, describe the ‘democratic anchorage’ of networks, which, translated to CBIs, means that they are controlled by democratically elected politicians, represent participating groups and organizations, accountable to the local community, and interact in accordance with a certain democratic code of conduct (Delmas and Toffel 2008). While the latter three-point seems to fit CBIs, the first point does not. Such initiatives often emerge through networks in which the government or elected politicians do not predominate (Boonstra and Boelens 2011, 113). What is more important in the case of CBI is the nature of representation, the source of legitimacy, and the transparency and rules of conduct within the initiatives. The amount of support an initiative receives from other established non-government organizations is also important. Organizations can help with brokerage activities via boundary spanners, increase the recognition and legitimacy when starting the initiative or later on in the process, or provide resources such as expertise, advice, or financial means.

3. Data and analysis

This article studies 17 cases from nine different countries. These cases are shown in Table 3. These cases are very different, take place in very different contexts, and are initiated for different reasons. However, they share core characteristics: they grew bottom-up; citizens decided on their course, aims, and resources; they are place-based; and outcomes are expected to be beneficial for the neighborhood. More information about specific cases can be retrieved from the researchers.¹

Table 3. Cases.

Case	Country	City	Main aim
Broekpolder	the Netherlands	Vlaardingen	Protest movement - against large country houses in the Broekpolder area, and to develop and maintain multiple-purpose area: leisure, sports, education, nature reserve, and ecology.
Delfshaven	the Netherlands	Rotterdam	Cooperation established by residents aiming to improve the economic development and resilience of the Delfshaven area by linking larger businesses and institutions to local initiatives
Caring Neigbourhood	the Netherlands	Eindhoven	Initiative of elderly residents aiming to make their neighborhood an appropriate place to grow old while staying independent by taking care of each other rather than (fully) depending on formal elderly care organizations.
Caterham Barracks	UK	Caterham	After closing the army barracks in 1995, community members aimed to protect the site and prevent the demolition and replacement of the historical building by making the site a conservation area.
4 tunnel	Italy	Milan	An initiative focused on the environmental, social, and cultural upgrade of areas close to Milan Central Station. Mainly focusing on infrastructure.
Defending Debar Maalo	Macedonia	Skopje	Initiative that aims to safeguard the public spaces in the neighborhood (under threat of proposed urban renewal) by mobilizing citizens to find solutions to common problems in the neighborhood: increase of traffic, public spaces occupied by cars, increased noise, and air and solid waste pollution.
We are Karposh 4	Macedonia	Skopje	A protest group aiming to stop the destruction of the parks, which are to be replaced by parking lots. Furthermore, this group advocates for the transformation of small, temporary buildings to communal homes or small retail shops.
Park defenders	Macedonia	Skopje	An initiative to protect and help sustain green public spaces in the neighborhood quarter around the hotel ‘Bristol’ where new administrative buildings were planned.
I am Xochimilco	Mexico	Mexico City	The aim of this initiative is to avoid continuous deterioration of Xochimilco’s lake and its biodiversity caused by unplanned urbanization, illegal settlements, and illegal wastewater discharge into the lake, as well as inefficient water management and bad governance from the local government.
Water to drink	Mexico	Juárez City	The initiative aims to provide low-income communities with water services, free water storage containers, educational materials, and water treatment kits.
Buklod Tao	Philippines	Barangay Banaba, San Mateo, Rizal,	Buklod Tao aims to strengthen the capacities of the community in disaster preparedness, awareness, and environmental protection.
Catholic relief services	Philippines	Tacloban City	This recovery project provides shelter and latrines to the Typhoon Haiyan-affected households in the island of Leyte.
Tanauan Barangays	Philippines	Municipality of Tanauan, Eastern Visayas	This initiative aims to rebuild and recover the community after the Typhoon Haiyan disaster and to create awareness and preparedness to cope with future disasters. The local government collaborates to teach disaster risk reduction and management in the barangays.
DAMPA	Philippines	Quezon City	The general aim of DAMPA is to create practical solutions to poverty problems, especially among the urban poor communities in the Philippines. For example, adequate and affordable housing, protection from evictions, relocation assistance, and the provision of basic services (e.g. water, health care, and education).
Anybody can sail	Sweden	Gothenburg	Initiative to give the unused inner harbor a new function: an accessible sailing school for children with disabilities.
Climate congress	United States	Cambridge, Massachusetts	Activists called for a community congress, which was presented as an opportunity to build community-wide consensus around the kind of paradigm shift needed to radically reduce greenhouse gas emissions.

Fuzzy-set Qualitative Comparative Analysis is used to systematize the comparison, a specific method that enables comparison of cases from various contexts in a qualitative way (Ragin 2000, 2008a, 2008b; Rihoux and Lobe 2009). The cases were thoroughly compared after defining and describing all the individual remote and proximate conditions in the conceptual part of this paper. Researchers used their case and theoretical knowledge to assign cases a membership score from 1 to 0 based on particular conditions. For example, if a case took place in an ‘urbanized neighborhood’, it was considered as having a high membership score or a high degree of this characteristic and therefore it scores a 1 for this condition. Each case was subsequently assigned a fuzzy value for each condition. This process is called calibration (Ragin 2008b; Schneider and Wagemann 2012). With these scores, we indicate:

• 1 = case is full member of a set and can be characterized by a high degree of this condition

• 0.67 = case is more in than out of a set and can be characterized by a moderate degree of this condition

• 0.33 = case is more out than in a set and can be characterized by a low degree of this condition

• 0 = case is fully out of a set and can be characterized by the absence of this condition

Each of these anchor points refers to a specific qualitative label, which can be found in Annex 1 (online supplemental material) After assigning all cases a score on all conditions, the researcher does the same for the ‘outcome’ (in frequentist methods: dependent variable). The outcome in this paper is whether or not a CBI performs (i.e. whether it reaches its intended objectives or not).

Necessity and sufficiency analyses can be done with the completed dataset. We provide an example to explain these two notions. There are many explanations for why a human is feeling well. Some people might need the status of a large house and a stable job, others might need a broad circle of friends and a large family. For human well-being, water is a necessary condition. However, water alone does not suffice for well-being. People need a combination of water, food, air, social contacts, etc. to be well. So, water is not a sufficient condition for well-being (Rihoux and Ragin 2008). This simple example explains how QCA unravels data. By assigning membership scores to cases ranging from 1 to 0, a researcher determines what necessary and sufficient conditions explain CBI performance.

There are multiple parameters of fit to check the relevance of the analyses. First and most important: consistency. A result is perfectly consistent if a certain condition is present whenever a particular outcome is present. If the consistency parameter drops, it means that there are cases that show the same combination of conditions but a different outcome. In technical terms, we can say that consistency is the proportion of the total membership scores that the cases have in a particular outcome (Ragin 2006; Schneider and Wagemann 2012).

We conducted a carefully designed scoring and coding process in which:

• Each author was a case holder (for sometimes several cases) based on prior research conducted by this researcher and the therefore available knowledge of the case(s);

• Each case holder drafted a case template in which the case was described on the variables and a scoring sheet for the variables was developed;

• Two sessions with the researchers/authors were organized in which the case templates were critically discussed and checked (calibration, validation).

• In this way we believe a robust scoring process was followed in which the scoring was checked, double checked and finalized.

For example, the scoring on performance, went through a process in which the corresponding researcher described the output (in the case template) and outcome reached (up until that moment) in the case(s). They then gave their score expressing to what extent the objectives in the project were reached (high, moderate, some, absent). These scores were checked and validated in panel sessions among the researchers to convince the other members that the scoring was right. This process has led sometimes to slight adjustments in scoring. This process continued until the point the researchers reached common ground in the scoring of all cases.

The coding process resulted in a raw data matrix, shown in Table 4. A few restrictions must be explained. The data in general are highly skewed toward high scores. This means that the conditions are mostly present, and most of the cases are success cases (i.e. the outcome is almost always present as well). A few measures were implemented to deal with this skewedness. First, we took some of the conditions out of the dataset. The conditions motivation, network structure, and urbanized area are trivial necessary conditions for the presence and absence of the outcome (Goertz 2003). Second, we used the conservative solution, which fits our modest aims, to describe the data. Such a solution bases its conclusions only on the data that is in the dataset and not on logical remainders (combinations of 1-0 which were not empirically proven).

Table 4. Raw data matrix.

	Netstruc	High organizational capacity	Motiv	Highly democratic structure	Strong leader-ship	Govattitute	Invitational government culture	Strong NGO support	Urban	Heterogenic neighborhood	Performance
NLBroek	1	0.67	1	0.67	1	0.67	0.33	0.67	0.33	0.67	1
UKCater	1	1	1	1	1	0.67	1	0.67	1	0.33	1
USCCEC	0.33	0.33	0.33	0	0.33	0.33	0.33	0.33	1	0.33	0
NLDelfs	1	0.67	0.67	0.67	0.67	0.67	0.33	0.67	1	0.67	0.67
MADebar	1	1	0.67	1	0.67	0.67	0.33	0.67	1	0.33	0.67
MAKarposh	0.67	0.67	1	1	1	0.67	0.33	0.67	1	0.33	1
Maparkdefen	0.33	0.67	0.67	0.67	1	1	0.33	1	1	0	0
SWSegla	1	1	1	0	0.67	1	1	1	1	0.33	1
IT4tunnel	0.67	0.67	1	0.67	0.67	0.67	0.67	0.33	1	0,33	0.67
PhBuklod	1	1	1	1	1	1	0.33	1	0.33	1	1
PhDampa	1	1	0.67	1	1	1	0.67	1	1	1	1
ThCreative	0.33	0.33	0.33	0	0.33	0.33	0.33	0.67	1	0.67	0.33
MxIamXoxhi	0.67	0.33	1	1	0.67	0	0.33	0.67	0.67	1	0,33
MxWater	0.67	0.67	1	1	1	0.33	0.33	1	0.33	0	0.67
Nlcareneig	1	1	1	0.67	0,67	0.67	0.67	0.33	1	0.33	0.67
PhHaiyan	1	0.67	1	1	0.67	0.67	0.33	1	1	1	1
PhTanauan	1	1	1	0.67	0.67	0.67	0.33	1	0.67	0.67	0.67

4. Results

The analysis was conducted with the fsQCA package for R². As recommended by Schneider and Wagemann (2010), we first performed an analysis of necessary remote conditions. No single condition passed the necessity threshold for the presence of the intended goals. We next looked for necessary conditions among the proximate conditions. The analysis revealed four necessary conditions (threshold 0.90). However, we can also see that two conditions are trivial necessary conditions, as indicated by Relevance of Necessity (RoN < 0.6, in Table 5 – see Schneider and Wagemann 2012, 236–238): network structure and motivation. We exclude these conditions from the rest of the analysis. We see that these conditions are highly skewed and almost always present, even when a CBI does not perform well (a necessity analysis of the absence of the outcome was also performed). This is due to a lack of variety in the data. Hence, we exclude these conditions from the sufficiency analysis. However, what still remains is that the CBIs that reach their intended goals show strong organizational capacity and leadership. These are necessary conditions.

Table 5. Necessity analysis of remote and proximate conditions Presence of outcome: intended objectives reached.

	Consistency	RoN	Coverage
Government attitude	0.83	0.82	0.88
Government culture	0.63	0.82	0.88
Urban	0.85	0.38	0.70
Heterogeneity	0.66	0.86	0.85
Network structure	0.97	0.59	0.83
Organizational capacity	0.92	0.69	0.84
Motivation	0.97	0.47	0.79
Democratic structure	0.86	0.71	0.83
Leadership	0.94	0.67	0.85
NGO support	0.86	0.62	0.79

This paper only presents the paths toward success (i.e. the cases in which the intended outcomes were achieved). After the necessity analysis, we began the analysis of sufficient conditions with the sufficient remote conditions. Schneider and Wagemann (2006) recommend choosing the parsimonious solution for this two-step approach, which tries to minimize the solution even further than the conservative solution. The latter only considers the rows on which we have data. The parsimonious solution shows three paths to high-performing CBIs, which adhere to a consistency cutoff of 0.85. This threshold was chosen because this is good practice (Schneider and Wagemann 2010), and because it is high enough to prevent cases which are different in kind (cases in which the outcome is absent). In total, these paths are consistent and cover 10 of the total 13 success cases. PhDampa is part of both paths. These paths show CBI-enhancing contexts; in other words, contexts that are favorable for CBI performance:

1. The presence of an open government attitude (cons: 0.92, PRI: 0.88, cov: 0,63, unique cov: 0, 26, N = 5, IT4tunnel, Nlcareneig, SWSegla, UKCater; PhDampa)

2. The presence of a positive government attitude AND the presence of a heterogenic neighborhood (cons: 0.96, PRI: 0.94, cov: 0.60, unique cov: 0,10 N = 5, NLBroek, NLDelfs, PhBuklod, PhHaiyan, PhDampa, PhTanauan).

In sum, we can say that these two paths are consistent (solution consistency: 0.94) and together cover 10 of the 13 success cases (solution coverage: 0,83), making them empirically relevant. The paths also show that none of the three remote conditions are redundant (Schneider and Wagemann 2006).

We next combined the knowledge about the remote conditions with the sufficiency analysis. First, we combined the sufficient remote paths with the proximate conditions. In the paths underneath, we combined the knowledge we have from the remote conditions with a new combined analysis. We highlight only a few of the paths in which the remote conditions described above play a prominent role. These paths adhere to a consistency cutoff point of 0.9, to prevent cases that are different in kind. These are paths that cover many cases, and therefore have a high coverage and are empirically relevant.

Path 1: open government attitude AND strong organizational capacity AND highly democratic structure AND strong leadership (cons: 0.95, PRI: 0.92, cov.: 0.51, unique coverage: 0,09, N = 4, IT4tunnel, NLcareneig, PHdampa, PhTanauan, UKCater) -> high performance.

One of the cases that can be characterized by this path is the Caterham Barracks in the UK: a strong CBI with many resources, money, people, capacities, etc. The organizational structure of the CBI is in place, with a general board in which various stakeholders (private, public, societal) have their seat. The Trust’s Board makes decisions. This case contains people with high connective leadership skills, but these leading figures are also focused on getting results as well as building and maintaining internal and external connections. The legal frameworks from the government side are in place, with, for example, the S106 agreement. The community leadership can be considered as a connective style toward the neighborhood community and governmental institutions. This connective style of leadership evokes an open government stance toward the CBI. With many brokerage activities by various boundary spanners, the CBI is actively supported by, and receives, public recognition and resources. The government, for example, strongly encourages citizen involvement in developing the overall plan for the Caterham Barracks (S106 Agreement). The leaders showed vision and transformational leadership and attracted followers and volunteers which led, in turn, to organizational capacity. The CBI was organized as a trust. A board of directors represents public, private, and societal organizations, with the community leader as the chairperson. This was considered a democratic structure, leading to enforced organizational capacity with many committed volunteers.

Path 2: However, there are also many cases (N = 11) in which a positive government attitude is not present or part of the recipe toward performance. In those cases, a combination of organizational capacity AND democratic structure AND Leadership AND NGO supports leads to performance (cons: 0.93, PRI: 0.90, cov.: 0.72, unique coverage: 0.29). It seems that CBIs need at least some support, from either government OR NGOs.

Path 3: a supportive government attitude AND a heterogenic neighborhood AND organizational capacity AND highly democratic structure AND strong leadership (cons: 1, PRI: 1, coverage: 0.54, unique coverage: 0.32, N = 6, NLDelfs; NLBroek, PhBuklod, PhDampa, PhHaiyan, PhTanauan).

The combination of a supportive government attitude, a heterogenic neighborhood with great organizational capacity, a representative and legitimate process, and strong leadership leads to high performance for five of the cases. One of these is Buklod Tao from the Philippines. Buklod Tao is loosely translated as ‘people bonded together’ and aims to strengthen the capacities of the community in disaster preparedness and environmental protection. It is a people’s organization that originated in Banaba, a village in San Mateo, a peri-urban municipality. The organization has its roots in the church-based Basic Ecclesial Community (BEC). The founder and president, Manuel Abinales, was a tagadiwa (i.e. animator) for the BEC. He organized six buklod or ‘cells’ composed of 10-12 neighbors, mostly women, in North Libis and South Libis. They had weekly meetings not just on liturgical discussions but also issues of community importance. The organization started with about 70 members but inflated from 150 in 2009 to about 700 in 2010. Buklod Tao had 756 members in 2012. The organization addresses different local needs that require linkages with the formal institutions, umbrella organizations, and government institutions. Individuals and organizations from local, provincial, regional, national, and international levels supported Buklod Tao early in its development. Buklod Tao depends on funds to maintain its activities and organization, which it gained initially through member contributions and then through external funding. Their biggest financial contributors are the international organizations Christian Aid and International Disaster Volunteers.

Another case is Broekpolder in the Netherlands. CBIs are not often widely supported internally in the municipal organization, so a lot depends on the spokesman or boundary spanner at the municipal side. Broekpolder has many different people of various backgrounds involved. There are many volunteers with many resources, such as time, energy, experience, and knowledge. However, financial resources are often absent, and there is a constant search for additional funds. The CBI lacks transparency, as the three members of the board often work closely together but do not share this with the wider community. The case shows strong leadership with clear vision and ability to connect internally and externally. We see strong connective leadership in this case, not only at the community organization but also at the municipal organization, which infused an overall supportive stance from the local government. Support from the local government was dependent on the representative strength of the CBI; it had to prove that it was able to reach and involve citizens from different backgrounds. The connective leadership style of the community leaders was also important in this goal. The CBI had a high reach in getting many volunteers on board, which was also stimulated by the conditional support of the local government and the connective community leadership style. The connective leadership style also led to keeping everyone informed about what was going on and what important decisions were being made. Although the three board members were quite dominant in making decisions, the community leader was often out there to sense and pick up important issues, views, and standpoints. Various working groups were operative, leading to decision-making input on the board level.

However, we also see that one case is still successful despite dealing with the absence of a stimulating government attitude AND the absence of a heterogenic neighborhood: ‘water to drink’ in Mexico. There are multiple CBIs in Ciudad Juárez’s low-income neighborhoods. Access to water in these neighborhoods depends greatly on citizens and non-profit civil society organizations that provide low-income communities with water services, free water storage containers, educational materials, and water treatment kits. These communities are irregular settlements (e.g. illegal land tenure) where the local government cannot collect taxes. The residents, moreover, have a strong distrust of political actors, and there is a cold relationship with the political system. In addition, government bodies have proved to be ineffective in terms of water management, even though there have been efforts from the Mexican local, regional, and national governments to invest in water and sanitation infrastructure projects. Initiatives such as ‘water to drink’ work together with the US cross-border institutions to make positive changes in the region.

5. Discussion and conclusions

In this article, we conducted a fsQCA of 17 cases and analyzed which configurations of remote and proximate conditions explain CBI performance in the urban realm. We were interested in finding combinations of factors and conditions (configurations) that explain the performance of CBIs. With this study we aim to contribute to the (emerging) literature on CBI, especially on the impact and performance of CBIs. This study is not without limitations, so we must be humble in formulating the results and conclusions. One limitation is that we tried to include many cases, but the number of 17 cases is of course not enough to achieve conclusive and generalizable results. In order to become more conclusive, we need additional research and (large N) case studies. Moreover, the cases took place in different countries and different (sector, political) contexts. We didn’t explicitly take this contextualization into account in our study. We focused on urbanized areas, but in these areas various topics (social, economic, infrastructure, etc) also take place with certain political sensitivity. This context is, for example, discussed more extensively in other publications by, for example, Feola and Nunes (2014) who indicate the important role of mass media, private companies and variety of public authorities in reaching success. Moreover, sector context (food, energy, etc) and political sensitivity and politicization of the subject also seem to matter (Celata, Dinnie, and Holsten 2019). These context variables were not extensively considered in our study. Therefore, we must treat our conclusions with care. Despite these limitations which were caused by choices made for feasibility of the study, we arrived at some relevant and interesting insights.

In the first ‘model’, we only took remote conditions into account. A positive government stance toward CBIs is sufficient for performance of CBIs, which confirms the positive relationship between government support and CBI performance (Dale, Ling, and Newman 2010; Seixas and Berkes 2009; Igalla, Edelenbos, and van Meerkerk 2019b). Governments that have both legal and policy conducive frameworks in place are conditioning successful CBIs. We found this is a sufficient condition in almost half of the high-performing cases. This finding indicates that the way governments relate themselves to CBIs (i.e. their attitudes can function as an important catalyser). However, there are also many CBIs which perform well without these frameworks, they seem to find an alternative source of support, that of NGOs. It seems that CBIs need at least some support, from either government or NGOs. Some studies further stress the importance of the collaboration process between CBIs and other stakeholders (private, public, societal) to reach performance (Feola and Nunes 2014). In the wider literature on collaboration we see that relational quality and trust are important elements of collaboration that explain performance (Klijn, Steijn, and Edelenbos 2010). Future research could also include these factors to explain the performance of CBIs.

A second conclusion regarding the remote conditions is that a heterogenic community in combination with a supportive government attitude (e.g. in terms of brokerage activities) is a key to success. This heterogeneity leads to capacity variety, which enhances the performative capacity of CBIs in combination with government brokerage activities. This conclusion has different theoretical implications. Heterogeneity does not show the negative effects expected in literature on community engagement (cf. Botes and Van Rensburg 2000). However, although heterogenic community seems to matter to condition CBI success, the analysis shows that organizational capacity and leadership are necessary conditions. Other studies have already indicated the importance of human and social capital and inclusive and diverse communities (Feola and Nunes 2014; Celata and Sanna 2019). Celata, Dinnie, and Holsten (2019) also show that CBIs (in the field of renewable energy, transition and sustainability) develop many relationships inside the CBI but also with external stakeholders (especially public authorities). Our study shows that the combination with other factors, especially boundary-spanning and brokerage leadership activities, are key in gaining performance. It is important to learn more specifically how this interplay of factors work and which skills stakeholders (representatives from CBIs and related (public, private, societal) organizations need in order to cope with a multitude of relationships and therefore show good connective leadership.

To conclude, when we look into the model including both proximate and remote conditions, we observe that having a strong network structure or a strong motivation are trivial necessary conditions. Our sample is biased toward CBIs that have strong networks and high motivation, even if they are not performing well. Therefore, in this article, these conditions seem to have less explanatory power regarding success. Taking the proximate conditions into consideration, we found two recipes that explain CBI performance. CBIs that have strong organizational capacity, strong charismatic leaders, and a sound democratic structure supported by a government with conductive governmental frameworks seem to perform well. A final insight from our study is that CBIs with the above-mentioned characteristics – organizational capacity, strong leadership and sound democratic structure – need to be based in a heterogenic community and receive hands-on government support to perform well. In other words, CBIs need certain proximate conditions to perform well, but these need to be complemented with remote conditions for good performance. Therefore, government-CBI relationships remain an important research object for scientists.

Based on this study, we know that an open government attitude, and support from both government and NGOs enhances performance in different ways, but we still do not know which and how different supporting activities contribute to performance (Van Meerkerk, Kleinhans, and Molenveld 2018). Policy support can take different forms, such as public grants, taxation schemes, adaptation or regulation or land-use policies to allow CBIs to use vacant public spaces and buildings, etc. The actual support depends on the specific issue, sector specifics and nature of the reciprocal relationship between CBI and public authorities (Celata, Dinnie, and Holsten 2019). We need to know more about which types of support work in what kind of circumstances. This again pleads for a more contextual approach (type of issues/services, specificity of sectors/domains, etc.) to learn more about which (combinations of) factors explain under which conditions CBIs perform best.

Disclosure statement

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

Supplemental data

Supplemental data for this article can be accessed here.

Notes

Notes

1 Please contact the corresponding author: [email protected] to obtain more information, for example by detailed case templates in which the cases are described and scored.

2 Ioana-Elena Oana and Carsten Q. Schneider. 2018. SetMethods: an Add-on Package for Advanced QCA.

The R Journal. https://journal.r-project.org/archive/2018/RJ-2018-031/index.html