“100 Resilient Cities”: Addressing Urban Violence and Creating a World of Ordinary Resilient Cities

Although the use of resilience in international relations and urban planning has given rise to a growing body of critical research, this contested concept continues to feature prominently in the conversation on the development of cities. Taking the 100 Resilient Cities (100RC) network pioneered by the Rockefeller Foundation as a case study, this article exposes some of the challenges inherent in the implementation of a global model of resilience. Exploring initiatives related to violence prevention in the member cities of Medellin, Cali (Colombia), Chicago (United States), and Belfast (Northern Ireland), this study will look at the practices of resilience officers, a position created by the 100RC network, and determine whether it can be considered as a new profession in the field of resilience planning. It will also use urban resilience to question the category of global cities, by suggesting that networks centered on resilience can serve as globalizing agents for “ordinary cities” (Robinson 2006). Finally, this article maintains that although the flexible and elusive definition promoted by 100RC facilitated a global circulation of the concept, its one-size-fits-all approach implied significant challenges and led in some cases to its depoliticization.

ven as the use of resilience in international relations and urban planning has given rise to a growing body of critical research (e.g., Reid 2012;Fainstein 2015;Chandler and Coaffee 2016;Naef 2020), this contested concept continues to feature prominently in the conversation on the development of cities. In the last decade, urban resilience was also increasingly seen as an innovative tool for global governance, giving birth to new "transnational municipal networks" (Kern and Bulkeley 2009). The most important are 100 Resilient Cities (100RC), running from 2013 to 2019, and its legacy, the Resilient Cities Network (R-Cities), officially launched in 2020. Additionally, C40 is the main network focusing on resilience to climate changes, and the campaign Making Cities Resilient is guided by the United Nations Office for Disaster Reduction. These resilience-based networks aim to enhance transnational collaborations between cities, as well as public-private partnerships. Some of them are funded by foundations, like Rockefeller in the case of 100RC, or Bloomberg Philanthropies (among others) in the case of C40. Each offers a definition of resilience reflecting the idea of a global urban model. The objective is clearly acknowledged in the case of 100RC, which aims to "institutionalize a global model of urban resilience" through the appointment in member cities of a chief resilience officer (CRO).
Since urban resilience has been integrated in the discussion in social sciences, several commentators (Leitner et al. 2018;Roberts et al. 2020;Nielsen and Papin 2021;Fastiggi, Meerow, and Miller 2021) have pointed out the absence of empirical research on the topic and especially the lack of engagement with issues of power. Geographers and anthropologists thus have an important role to play by offering in-depth qualitative analysis to understand how resilience is promoted globally and adopted or contested locally. The objective of this study is to contribute to this emergent debate by providing some insights into the use of this concept in the governance of four 100RC member cities in three different regions of the world. By looking closely at the 100RC network until its end, this contribution sheds light on the challenges and limits the implementation of a global model of urban resilience can involve. This multisited study will also present some original elements of comprehension on how networks centered on resilient cities help call into question the category of global cities.
In line with King's (2004) claim that all cities are world cities and Amin and Graham's (1997) proposal to study ordinary cities, Robinson (2006) offered an alternative framework to the one opposing global and peripheral cities. She suggested that all cities are better understood as "ordinary." Robinson criticized a Western and neocolonial conceptualization of modernity that categorizes many cities as lacking the qualities of "city-ness." She called for a deconstruction of these widespread categories of global, world, or-in opposition-third-world cities. Calling into question the world cities literature's view that considers some cities as "out of the game" and irrelevant to theoretical reflections (Robinson 2006, 102), she advocated a "world of ordinary cities." The deconstruction of dominant frameworks sustaining what she saw as "the fantasy of (Western) urban modernity" (Robinson 2006, 13) implies the questioning of international rankings categorizing cities from the traditional (or worse, primitive) to the modern. In contrast, she saw difference as diversity rather than a hierarchical ordering. In Robinson's postcolonial vision of city-ness, the many ways of being modern suggest that all cities can inform urban theory. Looking at architecture and urban design in places like Rio de Janeiro, New York, or Kuala Lumpur, Robinson clearly demonstrated that modernity is borrowed and invented in both wealthier and poorer cities. As she put it, there is a need to address "the harmful habit of viewing the embrace of novelty as 'innovative' in Western contexts, but 'imitative' in many others" (Robinson, 2006, 66). Hence, by questioning the assumption that global flows are believed to originate from the West, whereas others are invisibilized or considered as copies (Robinson 2006), the objective of this article is to establish whether a network like 100RC contributes to valorizing practices and ideas born in cities deemed secondary.
The relation between urban resilience and globalization has been thoroughly addressed in social sciences (e.g., Leitner et al. 2018;Beauchamp et al. 2019;Nielsen and Papin 2021). Although cybercriminality and pandemics have recently exposed the weakness of highly interconnected centers, the integration of cities in international networks is usually seen as an asset for their resilience. Almost ten years ago, Djament-Tran and Reghezza-Zitt (2012) reminded us that, historically, cities relied on their backcountries when they were hit by a disaster. As they put it, globalization created new inequities, because the better connected territories were more likely to benefit from resilience programs. In this contribution, this relation is turned around: If world cities or global cities are more likely to be resilient due to their transnational connections, how does the affiliation of peripheral cities in resiliencebased networks contribute to the global integration of these less connected cities? To use Robinson's words, how can a network like 100RC help to create a world of ordinary cities?
To address these questions, I offer some empirical elements on the practices and discourses of resilience practitioners involved in 100RC, by focusing on the position of CRO, created during the 100RC program and continued in R-Cities. I suggest that with the global flows they generate (e.g., through conferences, field visits, workshops, policies exchanges, technical consultancies), CROs act as transfer agents who contribute to the "world city-ness of their city" (Robinson 2006, 112). By analyzing the position of CRO, I also examine how 100RC contributed to creating a new type of city-practitioner, reflecting Rockefeller collaborators' initial suggestion that a community of practice, and even a new resilience profession, could emerge from the network (Urban Institute 2018).
100RC offers a pertinent case study for looking at the way resilience can serve as a globalizing agent. Characterized by an important diversity of cities, ranging from secondary cities (e.g., Vejle or Semarang) to the most internationally connected (e.g., Los Angeles or London), 100RC extends the study of resilient cities beyond the Global North and especially beyond obvious (and global) resilient cities. Moreover, considering such a network as a lever to access the world stage offers an opportunity to better understand how urban resilience can address (or perpetuate) global inequalities. To question the relations between globalization, resilience, and inequality, I examine various initiatives involving resilience and urban violence.
100RC defined resilience on the belief that "flexibility was essential for cities and communities to thrive in the face of shocks and stresses" (100 Resilient Cities 2013). As Sabatier and Reghezza (2021) commented, "The objectives are fixed, but the approach is very flexible and targets are negotiated" (17). I maintain, however, that although this broad and flexible definition facilitated the global circulation of the concept, this approach, coupled with a one-size-fits-all methodology, also implied significant limits in the rollout of projects and sometimes led to the depoliticization of resilience. 100RC included urban violence in its broad definition of resilience, but in some cases, a rigid methodology based on a technical and ecosystemic conception of resilience did not leave room to effectively address issues tied to violence. Being resilient and prepared for future flooding is, for instance, quite different from building resilience toward gang violence. Frameworks based on the adaptation to natural stress and shocks, such as the construction of resilient infrastructures to "live with water" (in 100RC cities like Rotterdam, New Orleans, or New York), are not always suited to addressing issues of violence (in 100RC cities like Ciudad Ju arez, Medellin, or Cali). Most city dwellers do not want to live with violence; they refuse to adapt to a context of endemic violence framed as unavoidable. In contrast to inevitable events like floods or earthquakes implying adaptive and engineering-based strategies, residents of violent neighborhoods seek transformative dynamics to address the causes of violence.
This study thus stresses the need to contextualize resilience and to consider the multiplicity of its deployments (Simon and Randalls 2016). It also offers a novel contribution to the scarce research on how urban violence is addressed in resilience-based practices and discourses. 100RC, one of the first programs to consider urban violence through the lens of resilience, provides an opportunity to understand whether such approaches offer innovative means to tackle urban violence or whether they merely represent old wine in a new bottle. Finally, looking at the networks formed by different cities affected by violence sheds some light on how an international program like 100RC can help create a world of ordinary resilient cities.

Methods
To respond to the call for more empirical and qualitative research on the role of resilience in urban planning and city governance, this article offers a complementary approach to existing research, which is still largely based on content analysis of the literature produced by resilience-based networks (Leitner et al. 2018;Fastenrath, Coenen, and Davidson 2019;Fitzgibbons and Mitchell 2019). Examining some of these member cities on the ground, through an analysis of the discourses of the main stakeholders, provides an in-depth understanding of the challenges associated with their governance. Exploring the practices and narratives of the actors involved in these international networks gives a better understanding of how globalized ideas are adopted or contested locally. This analysis is part of a multisited research project conducted in five cities that joined the 100RC network at different stages of the program: Medellin, Cali (Colombia), Chicago, New Orleans (United States), and Belfast (Northern Ireland). In the context of 100RC and more broadly of international city networks, this multisited ethnography (Marcus 1995) represents an original empirical contribution to previous qualitative research focusing mainly on single case studies (or a comparison of two cases). It offers key understandings of a program like 100RC involving cities from all continents. Resilience-based discourses and practices did not spread in a homogeneous way around the world; this study provides an opportunity to observe their diffusion in various regions. Furthermore, the cities studied did not all join the 100RC network at the same time, and the discourses of the practitioners were significantly shaped by the length of integration of their city.
A special interest in the use of resilience in contexts of violence guided the choice of the case studies. Initially, this research focused on Medellin and New Orleans, presented as "flagships" of 100RC. The networks both cities built with other member cities, however, led me to extend my fieldwork to the other cities mentioned earlier. It appeared very soon in my research that for a good understanding of the connections that link these resilient cities, I would also need to globalize my own fieldwork. In this article, I present data collected in Medellin, Cali, Chicago, and Belfast; the case of New Orleans is addressed in another contribution (Naef 2022). After decades of being the world murder capital, Medellin became a role model in terms of violence prevention. The second city of Colombia thus featured as the 100RC flagship of resilience toward endemic violence; it is also present in other networks such as C40 and the International Urban Cooperation (IUC) program. As one the first cities to join the network, it was often presented as a "pioneer city" or a "laboratory city." Although not considered as flagships, Cali and Chicago also had violence as a main stress identified in their resilience strategy. By contrast, Belfast's strategy did not focus on violence, despite the city's troubled history of conflict.
Fieldwork included eighty semidirective interviews with resilience practitioners (in 100RC, C40, Making Cities Resilient, and IUC), 100RC partners (public and private actors), municipal officials (elected representatives and staff), and civil society in general (nongovernmental organization [NGO] collaborators, urban practitioners, violent nonstate actors, and community leaders). Fieldwork lasted four months; interviews in the case studies were conducted in English and Spanish without the filter of a translator and data were then coded. In parallel, extensive content analysis was conducted throughout the research, including key reports published in the frame of 100RC and other networks; promotional material linked to social media and blogs; and specific media coverage. Furthermore, participant observation in several conferences, field visits, and workshops organized in the frame of 100RC, United Nations Office for Disaster Reduction, and R-Cities brought additional insight to the study.

A Journey Toward (Global) Urban Resilience
Despite Robinson's criticisms, the designation of global city is still widely used by actors involved in urban planning and policymaking. Numerous indexes categorizing the "global-ness" of cities have surfaced in recent decades, based on the connectivity, industry, finance, or cultural activity of urban centers. Sassen (1991) coined the label of "global cities" to describe centers that shape the global economy. She nevertheless pointed out that only specific parts of these cities contribute to this process, what she called "global-city functions"; for instance, a business district like Wall Street. The indicators used to measure (and rank) global-ness have since been questioned, and for Robinson (2006), "location-bound indices cannot offer any useful assessment of the economic significance of 'worldliness' of most cities" (106). Yet, as Taylor (2004) commented, even if globalization implies much more than finance and business, these two domains are still seen as the dominant "networkers" (as cited in Robinson 2006) that participate to raise some urban centers as global.
Among the cities analyzed here, Chicago possesses many characteristics profiling it as global, such as its size and its cultural influence. Although Medellin, Cali, and Belfast can certainly be categorized as less globally integrated, they had nonetheless already acquired a certain international aura before 100RC was launched. Medellin has recently won several awards in urban design and has acquired international fame after being judged a world model in urban planning and violence reduction. As Kauark Fontes (2018) stated, Medellin reinvented itself with "ordinary measures." Inspired by the Global South, "it benefited from and set an example of abandoning formal western comparisons by tapping into its creativity" (Kauark Fontes 2018, 1). Belfast also increased its global aura, attracting many companies and investments in the last two decades: In 2019, the Financial Times ranked it second in the world's top midsize cities for business friendliness (O'Muirigh 2019). In contrast, Cali is generally considered more peripheral and much less connected worldwide. Yet, as the following results will show, 100RC significantly contributed to its global integration, beyond its international image as a salsa capital.
The four cities under study will also illustrate some of the limits attached to the rollout of a global model of resilience. Exploring the narratives of CROs and partners, reaching beyond the self-celebratory discourses often present in the 100RC literature provides a better understanding of the challenges linked to the governance and the global resonance of these resilient cities. Identifying these challenges does not imply that I reject the use of resilience in urban discourses, nor that I consider it simply as an "empty or floating signifier" as studies on sustainable cities suggest (Brown 2016). If a flexible and elusive definition of resilience can be considered as having a floating significance, I maintain nonetheless that empirical work is needed to clarify how resilience is heterogeneously understood, what resilience makes us do, and how resilience-based networks affect our cities. What a network like 100RC did first was to assemble a wide variety of cities, as it built on a broad and systemic approach to urban resilience: "the capacity of a city (individuals, communities, institutions, businesses and systems) to survive, adapt and thrive no matter what kinds of chronic stresses or acute shocks they experience" (100RC 2013). Besides natural challenges like earthquakes, floods, or heat waves, it also included social issues, such as endemic violence or unemployment.
100RC was funded by Rockefeller with approximately US$164 million, used mainly to provide two years' salary for the CROs of 100 member cities (selected from a cohort of more than 1,000). The objective was for municipalities to then assume funding the position with their own budget. Seniority and proximity of CROs to the mayor varied, depending on the administrations, and the status of the CRO within the administrative and political structure of municipalities was subject to debate. Cities oversaw the nomination of a CRO, by either advertising the post or making a direct appointment; they also decided on its management level. The main task of the CRO was to coordinate all municipality action associated with urban resilience and implement a resilience strategy for the city. Additional funds were available for knowledge exchanges (through conferences, workshops, and field visits), which constituted a basis of the program. To provide support for CROs, a large platform of partners was set up, composed of private corporations, public bodies, NGOs, and universities, among others. CROs were referred to as "resilience champions" in the 100RC literature, a status they would legitimize by their capacity to network and break the administrative silos in city departments. They were generally supported by a team of resilience officers, ranging from one assistant (as in Medellin, Cali, and Belfast) to a wider team of collaborators (as in Chicago). Medellin, Cali, and Belfast were guided by the same CRO during the whole 100RC process, whereas Chicago had various CROs during the two years' funding.
According to Rockefeller and the Urban Institute in charge of the assessment reports, the program contributed to put in motion 6,000 resilience practitioners; to create more than fifty resilience strategies; to leverage close to $900 million of pledged support from platform partners, philanthropies, and private sources; and to launch more than 1,800 projects (Rockefeller 2019; Urban Institute 2019). Results gathered in my own research call for caution regarding the large number of projects mentioned earlier.
Although resilience strategies were indeed composed of dozens of projects, many had already been implemented before 100RC started (or were already in the process of being implemented) and some others never happened. As some CROs commented, resilience strategies were often the continuation of preexisting plans. In these strategies, 100RC member cities were mobilized to provide a global model for other cities around the world. As the 100RC President Michael Berkowitz stated, "There's actually a pride in stealing solutions, city to city. … You'll run into mayors who actually brag about, 'We saw gondolas and bus rapid transit in Medellin and we went back and we're looking to implement that kind of solution here'" (Spector 2016). Robinson (2006) highlighted the potential of a city development strategy supported by United Nations agencies and the World Bank to "bring ordinary cities into view" (132). Although she acknowledged that these strategies were shaped by the power interests of international development actors, she nonetheless considered them as opportunities to respond to the diversity and complexity of cities. Accordingly, resilience strategies are considered here as the main tool for CROs to create and reinforce what Appadurai (1996) conceived as global flows in a disjunctive world order. Hence, by looking at the circulation of these flows in resilience planning through the discourses of CROs and an analysis of their strategies, I evaluate how these networks contribute to bringing less globally integrated cities into spaces where ideas and connections are forged, relations formed, and experiences packaged as best practices (Robinson 2011). S€ oderstr€ om (2012) already questioned how the designation of "smart city" enabled these urban centers to access global flows of knowledge and practices or if this label implied only virtuous rhetoric and another way of consolidating hegemonic categories. In the fields of architecture and urban planning, Dupuis (2017) demonstrated how cities from the North and the South were now more autonomous and interconnected, choosing innovative solutions from a world portfolio. Local actors travel between different cultural worlds, to access spatially dispersed resources and adapt them to specific contexts. To connect these worlds, politicians, practitioners, activists, and consultants act as transfer agents: "shuttling policies and knowledge about policies around the world through conferences, fact-finding study trips, consultancy work and so on" (McCann and Ward 2011, xiv). They help to draw mental maps of "best cities," bringing new places into conversation with each other, while pushing others apart (McCann and Ward 2011). The results presented here thus illustrate that 100RC represents an example of these world portfolios, and CROs are transfer agents, facilitating the translation and adaptation of urban models from one context to another.

Circulating Knowledge on Resilience and Violence
The social dimension of resilience was present in most 100RC city strategies, and endemic violence was a concern shared by almost a quarter of the cities involved, many of them in Latin America and the Caribbean. Chicago and Medellin appeared as key players in issues of resilience and urban violence. The two cities are associated with large-scale endemic violence, and both developed innovative ways of decreasing it before their integration in 100RC. Chicago has been tackling issues of violence over decades through its network of NGOs and universities. Medellin became a model of violence prevention based on its ambitious program of social urbanism, where the improvement of public spaces in some of its most violent neighborhoods significantly reduced the number of homicides. Both cities hosted international forums on resilience and violence with the collaboration of 100RC and platform partners such as the U.S. Agency for International Development and the World Bank. In Medellin, participants visited emblematic neighborhoods; for instance, specific parts of the Comuna 13 considered as showcases in terms of resilience and violence (Naef 2020). In Chicago, they were presented with innovative local programs. Cure Violence, for instance, which uses community members as "violence interrupters," inspired a similar project in Charco Azul, one of the most violent neighborhoods of Cali. Like the Comuna 13 in Medellin, this area also became a showcase of Cali's resilience to violence and was often visited by external delegations.
The recovery of public space was a special focus. In this context, Medellin was presented in Cali's Resilient Strategy as a model (100RC 2018) mainly due to its international fame in terms of public space regeneration, carried out in its social urbanism program. As Cali's CRO at the time pointed out, however, these two cities approach urban planning in significantly different ways: In Medellin, they do large interventions in the public space; and from there on they start to do some community work. Our interventions in the public space are a lot smaller and simpler. (Interview with CRO, 25 October 2019) Here, Cali's former CRO is referring to the integral urban projects, which form the backbone of social urbanism in Medellin. These interventions in the public space all have a broad focus-tackling simultaneously issues linked to mobility, security, education, and sport, among others-and associated with areas as large as an entire neighborhood. In contrast, Cali favored microinterventions on specific squares or street corners (Figures 1 and 2). Actors involved in this program saw urban planning in Medellin and Cali as different, mainly due to the way public space was treated but also due to the nature of the planning itself: Medellin has a planning strategy for the twenty upcoming years. This way of thinking twenty years Despite their differences, Medellin and Cali continued to collaborate, under the guidance of their respective CROs. In November 2018, they partnered with Belfast and were awarded a grant from the IUC program City-to-City Pairings, which promotes partnerships between European and non-European cities. Delegations from each city visited the two others to exchange views on local strategies related to violence prevention, peace building, and restorative justice. During field visits, emblematic areas like the Comuna 13 and Charco Azul were again center stage. Referring to the Colombian peace process, the IUC later stressed the importance for Cali and Medellin of visiting a city like Belfast, which they saw as a model for peace building after the 1998 Peace Agreements. The Belfast delegation, however, realized that whereas their city might serve as a model for the consolidation of the Colombian peace process, this European city also had many things to learn from Colombia. As the Belfast CRO commented: We felt that we had learned more from them than they did from us. One of our conclusions was that we may have overemphasized institutions and structures over feelings. … We all came back on the plane from Colombia feeling we got the institutions right. And yet, we somehow hadn't opened the memory box. In this context, the importance of memory work and violence prevention in the strategy of both Colombian cities was considered an important lesson for the Northern Ireland capital, because violence, collective memory, and peace building were not central to their conception of resilience, which focused mainly on climate change. This city-to-city exchange also illustrates how CROs' work extends beyond their main mandate of coordination to encompass applying for additional funds and acting as catalysts to enhance city cooperation. Their role as transfer agents includes creating subnetworks related to specific thematics such as violence prevention. Looking at city networks in European climate governance, Kern and Bulkeley (2009) conceptualized these subnetworks as "pioneers for pioneers," pointing out that some member cities are more active than others, especially those that joined earlier on. The examples presented earlier also demonstrate how resilience-based networks contribute to creating spatialities of circulation (Robinson 2011). Workshops organized in city halls or cultural centers, field visits in conflict-ridden neighborhoods, and conferences grouping member cities sharing similar stresses all take part in the circulation of knowledge, best practices, and models associated with resilience and urban violence. Exemplary case studies such as those of the Comuna 13 or Charco Azul, repeatedly visited by practitioners and municipality representatives, bestow legitimacy on them.
"Not in My Term of Office" One 100RC regional director highlighted the fact that the network reached beyond governments because a resilience agenda was not a government agenda: "The challenges that a city must face in the future have absolutely nothing to do with the political party that is governing the city at a specific time" (interview with 100RC executive, 21 January 2020). Even if a resilience agenda is not a government agenda, however, my own research confirms that it can hardly ignore a given political context. Examples related to resilience and violence are enlightening. From 2013 to 2015, the shaping of Medellin's resilience strategy focused on peace building and violence prevention. When the strategy was ready for launching in October 2015, the mayor called the CRO and suggested he postpone its release: "If you launch this document [now] it will be a big triumph for me. But the next mayor will not take it into consideration" (interview with CRO, 25 January 2019). Medellin's resilience strategy was thus unveiled later, in 2016. Nevertheless, the new administration expressed very little political will to implement it, because issues related to violence prevention and peace building were no longer priorities on its agenda. Despite the lack of interest from the new administration, Medellin's CRO initiated a "hemispheric network on violence prevention" grouping cities from the United States, Latin America, and the Caribbean. Because the funding was provided by the World Bank and U.S. Agency for International Development, however, the arrival of Donald Trump as the U.S. president in 2017 jeopardized most of this support and cut short the project. The absence of backing from his own municipality and the end of international funding significantly affected initiatives associated with violence and resilience in Medellin. As the CRO stated, "How can I talk about violence prevention if my own city is not taking it into account?" (interview with CRO, 25 January 2019). Another 100RC regional director described the vulnerability of projects' continuity and the reluctance of politicians to anticipate crisis as the "not in my term of office" syndrome (Carter 2020). In other words, resilience planning goes beyond the temporality of a mayoral term of office, and although mayoral buy-in is important for the implementation of projects, it is not enough. For the resilience agenda to remain a top political priority, the strategy uptake needs to reach the whole community: the municipality staff (who might remain in position well beyond the mayoral term) as well as all other partners. Yet, as mentioned, observation and interviews demonstrated that partners, especially local NGOs listed in resilience strategies, were often barely aware or sometimes totally unaware of the existence of the 100RC program. This disconnection between 100RC and local partners has been confirmed in other contexts. For instance, in Rome and Athens, Galderisi et al. (2020) highlighted the limited timeline of 100RC as a central obstacle to adequately involving local partners. In a similar vein, the 100RC midterm evaluation report in 2018 concluded that the expectations of partners had not been met and that CROs had noted a lack of clarity about which partners they could engage with.

The Limits of a Global Approach to Resilience
Webber et al. (2020) emphasized the flexibility and mobility of notions such as sustainability, smartness, and resilience: "Their ambiguity means that they can be defined in different ways to suit those deploying them, enabling them to travel through the fast policy networks of propagating institutions and individuals" (343). Flexibility in the meaning of resilience, however, was not reflected in the methodology proposed to cities when they had to build their own strategy. Several CROs commented on the rigidity of this methodology, based on a one-size-fitsall approach to resilience that did not reflect the specificities of the cities involved. One CRO qualified the 100RC methodology as authoritarian: They forced us to have this resilience strategy, which I don't think is a bad thing. But the methodology did not conform to my political reality … it's good to have concepts, participatory [processes], the problem was they said: "You must have fifty goals." I don't want fifty goals, I want fourteen!" (Interview with CRO, 12 October 2019) The rigidity of the 100RC conceptual framework was also addressed in the final assessment report itself, which underlined the importance of finding a balance between the one-size-fits-all model and actions tailored to specific cities (Urban Institute 2019). Other resilience practitioners commented on the fact that this methodological rigidity led to many difficulties in the implementation of projects: I don't think any major city network, whether it be 100RC, C40, ICLEI [Local Governments for Sustainability] … none of them have cracked that code on how to go beyond planning. … 100RC was really effective for a lot of cities in planning and helping to plan. There were a lot of planners on staff so they understood planning systems, but they themselves didn't really understand their role after that. (Interview with resilience team collaborator, 31 July 2019) When it comes to the specifics, they are like: "Hey you got to get in line with everyone else." So, that's kind of frustrating. (Interview with CRO, 23 July 2019) This concern was even more striking when social issues like urban violence were concerned. As one of Medellin's officials commented, the difficulty of addressing urban violence as part of the implementation of their resilience strategy led them to consider going back to some more "environmental problems": "But in the end, neither one nor the other worked" (interview with Medellin official, 18 October 2019). Many 100RC stakeholders nonetheless viewed the integration of a social perspective into the resilience discussion as one of the most innovative contributions of the program. A nuance was nevertheless introduced by one of the regional directors, who questioned the pertinence of this approach in poorer urban centers: You could say that in places where social exclusion is a priority, perhaps urban resilience should not be a priority. Resilience might be an expensive approach to public policy. (Interview with 100RC executive, 21 January 2020) Some scholars argue that the ecological and systemic foundation of resilience is inappropriate for addressing social problems, especially those of poverty, violence, and inequality, and contributes to the depoliticization of the concept (Fainstein 2015;Davoudi 2017;Roberts et al. 2020). In contrast, as Matin et al. (2018) maintained, "equitable resilience" calls for policy and practitioners to engage with the politics of social, cultural, and political change: "equitable resilience is that form of resilience which is increasingly likely when resilience practice takes into account issues of social vulnerability and differential access to power, knowledge, and resources" (202). For them, the depoliticized language of resilience is not helpful. They consider transformation as inherently political and advocate for a more organic way to bring power and agency into resilience thinking. In the context of 100RC, this criticism was, for instance, central in the decision of the former CRO of Durban to leave the program. The Durban resilience team considered the 100RC definition to be a reductionist application of the concept from the natural sciences to the social world, failing to address questions of politics and power that were central in this South African city (Roberts et al. 2020). On another hand, Grove et al. (2020), in their study of Miami's membership in 100RC, called for caution in simply reading urban resilience as a depoliticizing imposition of external governmental rationalities: "it is a site of political potential: a space where a variety of actors reflect on the limits of existing urban practices and advance competing ethically and politically charged visions or urban governance" (1623).
Finally, the translation of the word resilience represented another important challenge for the global aspirations of 100RC. Globalization and inequality are embedded in the dominant use of English in resilience discourses, and especially in a program like 100RC comprising forty-seven countries and twentyone different languages. As Chmutina et al. (2021) commented, "While the most sustainable solutions often emerge in a local context, speakers of non-English languages are compelled to express themselves using categories and concepts developed in English" (1). For them, these norms serve to perpetuate inequalities and allow a neoliberal agenda to flourish. CROs and 100RC staff were well aware of this complexity, however. As one former CRO commented, resilience was inexistent or meaningless in most languages. Even where it exists, the English term was very often preferred because it conveys a meaning that other languages could not adequately transmit: Usually you try to find what is the closest word. In Greece, for example, resilience is translated into Greek, but the Greek word actually doesn't mean anything, and it's not used widely. It doesn't have the breadth that the English word, for example, has. Therefore, we used it because we had to use the Greek equivalent, but we ended up [with] everybody using the English word in the Greek context. Resilience became almost like a new word but was able to describe this new kind of approach. Because the Greek word had a very conservative, technical connotation. (Interview with former CRO, 22 January 2020) CROs thus played a key role of translation and storytelling: They were transfer agents who aimed to diffuse a shared way of tackling urban resilience. CROs were the linchpin in the diffusion of a common language in a city network characterized by important economic and cultural diversity. As one 100RC collaborator commented, "CROs went through the same process, the same methodology, and they all feel like part of an alumni group of people who went through a PhD" (interview with 100RC executive, 22 January 2020). CROs also had to play with words when negotiating with partners and with Rockefeller itself, because every institution had its own approach and understanding of resilience: When I sit with the World Bank or Rockefeller, they all have their own approach and I am like, "Okay, okay, … " but at the end I grab what is convenient for me and my citizens. I just play along with what they need to hear so I can get what I need. (Interview with CRO, 12 October 2019) In contrast, observation in the field and content analysis of material related to partners showed that several of them did not use the term resilience in their own discourses. Some initiatives listed in the strategy were centered on the concept, but many others focused on topics like climate change, cybersecurity, or mobility and never referred to resilience.

Global Resilience and Local Challenges
100RC and its inheritance R-Cities brought an important diversity of cities into the global arena of resilience. From 2013 onward, the Rockefeller Foundation established a network of cities characterized by high diversity in terms of size, wealth, and international connections. Very simply, despite an overrepresentation of northern (and especially U.S.) cities (Tierney 2015;Nielsen and Papin 2021), 100RC put "global" and "peripheral" cities together in the world game. Resilience could thus act as a globalizing agent, by reinforcing international flows in some of these urban centers. The four cities analyzed here joined 100RC at different times between 2015 and 2018 and present different perspectives about the impact of the program on their governance and their global reach. Whereas several 100RC stakeholders and partners in Chicago and Medellin expressed skepticism about the effects of the program in their city, the views in Cali were much more optimistic. Belfast was still in the process of building its strategy during the study, and the resilience team was enthusiastic about its future implementation. In terms of resilience and violence, Chicago played a significant role in diffusing knowledge on violence prevention. Many of the projects eventually implemented, however, focused on water management. In contrast, Medellin, presented in the 100RC literature as the "resilience to violence" flagship, struggled to live up to its role model status. The Medellin resilience office eventually spun off as a nonprofit, and its focus is now more local, centered on its home department of Antioquia. In Cali, several innovative projects were developed in lowincome neighborhoods, principally the construction of modern schools and the implementation of violence prevention initiatives. These projects clearly reinforced the position of Colombia's third city on the global map. Cali was nominated in 2019 as one of the five finalists for the United Cities and Local Governments peace prize. 1 Furthermore, the school construction program of Cali, labeled My Community Is School, enabled the city to join the United Nations Educational, Scientific and Cultural Organization World Learning Cities Network. In 2019, the city hosted the United Nations Educational, Scientific and Cultural Organization International Forum on inclusion and equity in education. Cali is now catching up with its neighbor Medellin as an international model of social investment and peace building in Latin America. It was, for instance, praised in the 2020 UN-Habitat "World Cities Report" for its work with informal networks and communities to raise awareness about the COVID pandemic (UN-Habitat 2020, 214).
This study also demonstrates that the flows generated by a network like 100RC are far from being unidirectional; they do not travel solely from the Global North to the Global South. Whereas Medellin has been a Southern urban model for two decades, the rising importance of a city like Cali in resilience thinking is revealing. The influence that the two Colombian cities had on a European city like Belfast in terms of memory work and peace building exemplifies the growing impact of cities considered to be peripheral. Several 100RC stakeholders also introduced the idea that implementing resilience strategies was less complex in what they called secondary cities-often a reference to cities from the Global South-where urban planning was considered more flexible. One regional director, for instance, commented on the different sizes of the cities involved in the program: "Resilience is not for the wealthy and big cities. We have seen more success in secondary cities, because of their size and their ability to experiment, learn, iterate and locally institutionalize" (Carter 2020). This focus on smaller cities can be found in the newborn R-Cities, which in November 2020 announced a partnership with the National League of Cities. 2 The executive director of R-Cities stressed the importance of introducing resilience thinking into midsize cities and declared that R-Cities would assist the cities of the National League of Cities with network expertise (field notes, 15 November 2020). For some practitioners, when they are not "wealthy and big," cities are considered more flexible for resilience planning. Networks like 100RC or R-Cities can therefore offer a breeding ground for the globalization of these smaller-and maybe-"not so ordinary" cities.

CROs: Transfer Agents and Gatekeepers
If this analysis demonstrates the central role played by CROs and their collaborators in the globalization of peripheral cities, it remains ambivalent on how the CRO can be considered as a new profession. All cities analyzed here are now also part of R-Cities, and they all had an appointed CRO when this article was written. In R-Cities, CROs often have a different title (resilience directors, heads, or commissioners) and often wear a double hat in their administration, usually in departments such as planning, strategic management, or sustainability. In Cali, for instance, the CRO was also the territorial subsecretary for planning and opportunity. Hence, although the institutionalization of these "resilience champions" in local governments was seen as one of the major achievements by the 100RC staff, their double roles suggest that the continuation of dedicated funding for a CRO position remains a significant challenge. Moreover, the status a CRO should have in his or her city is an important issue. Political turnovers are considered both in academia and in practice as putting important limits on their work. Yet, although most of the resilience practitioners interviewed here favored a position as close to leadership as possible, some expressed doubts about this view, pointing to the political vulnerability such status implies. Even among CROs, there was debate about the position of a CRO elected as a deputy mayor or, in contrast, assigned to middle management. Some suggested moving out of the government orbit, as did the resilience office of Medellin, to gain more independence. The case of Medellin, however, showed that this approach did not insulate the position from political turnover. In general, this study demonstrates the importance of mayoral support for rolling out resilience planning, confirming that despite the risk of political turnover, CROs close to their mayor seemed more successful in implementing their strategies. One 100RC regional director drew attention to the evolution of the CRO position as a new profession in resilience practice by pointing to its replication in nonmember cities. In a different viewpoint, another CRO pointed out that, in theory, a CRO would not be needed if resilience was embedded within all of the city's systems. Nevertheless, although this article brings an ambivalent response to the vision of the CRO as a new profession, it concurs with McCann and , who urged us to seriously consider the role of transfer agents and to view "transfer" as a sociospatial, power-laden process, implying changes and struggles: "the expertise of various think tanks, consultants, gurus, and mediators has become central to the day-to-day governance of cities" (xix).
CROs act as transfer agents by breaking down silos within local government, applying for additional funds, and building new networks. Through mundane practices such as organizing workshops and field visits, they act as catalysts, spreading a global narrative on urban resilience. Ward and McCann (2011) called for a careful analysis of these practices, which they consider "at first glance mundane" but significantly influential in the way cities are globally positioned. Taking up this idea, S€ oderstr€ om et al.
(2014) demonstrated that smart cities, like sustainable and creative cities, are part of contemporary language games played around urban management and development: "These games involve experts, marketing specialists, consultants, corporations, city officials, etc. and frame how cities are understood, conceptualized and planned. Although we might consider this discursive activity with some skepticism, it often makes a difference" (307). This study confirms that mundane practices and language games related to urban resilience also affect the cities where they are developed. In 100RC, CROs, but also consultants, architects, city officials, or scholars, played the role of transfer agents and contributed to create what Robinson (2011) defined as spaces of circulating knowledge. They traveled, for instance, from the United States to Colombia to share best practices in violence prevention. A program like Cure Violence inspired resilience-based practices in some of the most violent districts of Cali, where a neighborhood like Charco Azul became a living laboratory in urban resilience planning and violence prevention.
In line with this vision, CROs and other transfer agents also act as gatekeepers by selecting what and who will be at the center of resilience discourses and practices, as well as how they will be integrated. Indeed, my own research demonstrated that among the vast number of initiatives listed in city strategies, not all were finally implemented and only a limited number of the partners listed benefited from them. Although CROs do not have complete control over this process, they help direct funds to specific projects, select which organization will be promoted, and choose what neighborhood or program will serve as a showcase for their city.
Resources in terms of funding, knowledge, or technical support are not unlimited, and some communities or places might benefit more than others. Some of them, because of their history or their strategic location, might attract a disproportionate amount of resources and so compromise the broader development of the whole city. In this study, urban areas such as the Comuna 13 in Medellin or Charco Azul in Cali have been continuously spotlighted in recent years during field visits and international meetings promoting urban resilience. These places can be considered, using the metaphor of synecdoche proposed by Amin and Graham, as representing entire cities by featuring just parts of their incredible diversity (as cited in Robinson 2006, 171). Such a process can lead to inequalities, because the hypervisibility of some places or communities could self-perpetuate cycles where resilience champions have major access to resources, in contrast to others. Inequality can thus be seen in hypervisible islands of exception: urban areas where influential communities, support organizations, advocacy groups, NGOs but also social investment and innovative urban projects are concentrated, synecdoches that are often used to illustrate the resilience of a city as a whole.

The (Post)Politics of Resilience
Interviews in all case studies confirmed what has been partly illustrated by the previous debate on the political vulnerability of CROs: Political turnover was seen as a significant limit to the implementation of resilience-labeled projects. Robinson (2006) emphasized the importance of a stable and consistent political context for city development strategies to be replicable instruments of urban development. In the context of 100RC, the impact of political changes on the rollout of resilience projects has been confirmed by the program assessment itself (Urban Institute 2019), as well as by other researchers in Rome, Athens (Galderisi et al. 2020), and several U.S. cities (Fastiggi et al. 2021). In my own research, CROs and 100RC staff all agreed on the challenges political turnover represented. In 2020, a couple of months after the election of the Chicago mayor, a former CRO, for instance, commented on the importance of reflecting on the mayoral transition: It is one of the tricky parts for 100RC. … And that's something that every organization that thinks about urban issues has to think seriously about, because just because one mayor is into it doesn't mean the next mayor is. (Interview with CRO, 23 July 2020) This statement was acknowledged by one of the regional directors, who nevertheless insisted on the importance of embedding resilience thinking in the city administration, to ensure its continuation: Of course, this is tricky because it means that the person that has been appointed has the trust of the mayor, and when the mayor leaves and a new mayor comes, he or she might not have the same level of trust with this particular person. We don't defend names or persons here. We have seen many transitions in our program over the past six years, CROs come and go, and the resilience agendas continue. (Interview with 100RC executive, 21 January 2020) Moreover, questions related to the rigidity of the methodology, also addressed in academia (Roberts et al. 2020;Webber et al. 2020;Nielsen and Papin 2021), were of significant concern. The potentialities and limits of the institutionalization of a global and one-size-fits-all approach to urban resilience also generated vigorous debate among 100RC collaborators and partners. I stated in this article that a broad and flexible definition of the concept favored its circulation. As Simon and Randalls (2016) commented, the flexibility of resilience "is what makes it an effective and slick frame for mobilizing interventions (or not) without really having to articulate agendas, values and desires" (15). Other scholars have used examples tied to 100RC to describe what they considered as a flexible, elusive, or ambiguous approach to resilience. In New Orleans, for instance, Tierney (2015) commented on the "useful ambiguity" of the concept, describing how it became more and more elusive and abstract over time. In line with this observation, my own previous work (Naef 2020) has also demonstrated that the 100RC elusive representation of resilience in Medellin facilitated the construction of an efficient city brand; nevertheless, it created tensions due to the lack of inclusiveness of the communities' various understandings of the concept. Here, I aimed to show that although 100RC's flexible and elusive definition of resilience-its useful ambiguity-allowed it to circulate smoothly in global discourses, it also challenged the implementation of associated projects. I also suggest that this approach to resilience, tied to a rigid methodology embedded in an ecosystemic framework, contributed in some cases to the depoliticization of resilience. Simon and Randalls (2016) proposed a notion of "resilience multiples," suggesting that there are no single resilience arguments. For them, to politicize resilience implies deuniversalizing the concept and contextualizing it. Both authors defined as "postpolitical" the lack of antagonism and the making of decisions by consensus and technocratic management. In the debate on world city rankings described earlier, Davidson and Iveson (2015) similarly characterized as postpolitical the discourses of policymakers aiming to situate their city at the heights of global city league tables. They viewed this process as a "regulating fiction" where world-class rankings become the metric for evaluating the performance of cities. In the context of smart cities, S€ oderstr€ om et al. (2014) also considered the "smarter cities" campaign as an apolitical approach leading to the creation of "fictitious cities" (Marcuse 2005(Marcuse , as cited in S€ oderstr€ om et al. 2014. They demonstrated how corporate discourses determine who shapes imaginaries on cities: "who has the power to define the smartness of cities" (S€ oderstr€ om et al. 2014, 310). After the International Business Machines Corporation (IBM) officially registered "smarter cities" as a trademark in 2011, S€ oderstr€ om et al. (2014) explored how the corporation's discourse tended to ignore history and presented urban affairs as an apolitical matter: "In the smarter cities campaign, causes of urban problems are associated with demographic trends, such as an estimated doubling of the world urban population by 2050, climate change and tight municipal budgets. Never with politics" (317).
Accordingly, some resilient cities, or are least some parts of them, can be considered as being in the realms of postpolitics. As in the apolitical framework described earlier, social issues such as chronic violence, endemic poverty, or historical segregation were often sidelined in favor of a more technical and engineering approach to resilience. This was the case in Chicago, which identified violence as one of its main stresses but ended up focusing primarily on water management, as well as in Belfast, which centered its initiative mainly on climate change, despite its important history of conflict. Similarly, in Medellin the new administration cut short the discussion on resilience and violence, because peace building was no longer on its agenda. In contrast, the resilience team of Cali managed to keep a strong focus on violence and inequality, through some innovative interventions in some of the city's most deprived neighborhoods.
Finally, as highlighted by S€ oderstr€ om et al. (2014), in the consensus of postpolitics, minorities are portrayed as disturbing factors in the quest for policies benefiting the whole population. The status of minorities in the (post)politics of smart or resilient cities raises the question of how resilience strategies address inequalities, a dynamic considered with skepticism by a growing body of literature on 100RC (Leitner et al. 2018;Fitzgibbons and Mitchell 2019;Roberts et al. 2020;Webber et al. 2020;Fastiggi et al. 2021). Although this study partly supports this view, the diversity of cities taking part in 100RC and the multidirectionality of the flows it generates can nevertheless also include what Grove (2018) saw as potentialities in resilience thinking and planning. Hence, resilience cannot be considered merely as a dominant mode of Western intervention in the Global South. 100RC shows that equality and resilience need to be considered within city strategies (how they address equality and what elements of the city should be the focus) and also within the global framework of the program (how cities are deemed to be resilient). On one hand, these results and previous work on 100RC demonstrated the challenges that such an approach to resilience implied when social issues like inequality, violence, or poverty were at stake. On the other hand, this study showed how 100RC helped bring less globally integrated cities into the picture. Only 100 selected cities from among more than 1,000 applicants were considered, however. From this standpoint, Nielsen and Papin (2021) considered 100RC to be an "elite club." As this study illustrates, the increasing influence of global agents on urban affairs, such as international organizations, multinationals, and philanthropies, contributes to situating some cities as well connected and others as outside the world grid. When they described IBM's trademark of "smarter cities," S€ oderstr€ om et al. (2014) used the concept of "obligatory passage point" (Callon 1986) to demonstrate how IBM created a discourse enabling it to figure as an inescapable actor (and contractor) when urban centers aimed at becoming smart cities. Although the Rockefeller Foundation is significantly present in the conversation on urban resilience, the context of 100RC is different, of course, because it does not imply any trademark. Nevertheless, the work of S€ oderstr€ om et al. casts light on the power relations that determine how a city is designated as resilient. My own research demonstrated that although 100RC initially proposed a flexible and integrative definition of resilience, the dominant use of English, the hypervisibility of some particular sites, the challenges of promoting alternative representations of the concept, or the need to adapt to a rigid methodology eventually paved a rather unidirectional way to urban resilience.

Conclusion
"Fictitious," "postpolitical," or "tied to a fantasy of modernity," all cities need to be further studied to better understand their complexity, the ways in which they interact, and how they are shaped by international processes. One objective of this study has been to shed some light on the power dynamics generated by the practices and discourses of transfer agents: "the disciplining role that these seemingly mundane practices reveal" (Ward and McCann 2011, 176). These resilience practitioners were also portrayed as gatekeepers, because they help determine what initiatives and ideas will circulate; what representations, knowledge, and discourses are diffused. Borrowing Tsing's (2000) topographical metaphor of globalization, they contribute to the contested making of channels and landscape elements that generate circulation. In this article, I have shown how some new paths were created, allowing cities to move more efficiently toward globalization, and contributing to the creation of a world of ordinary resilient cities. The implementation of city strategies was also constrained, however, by resistances associated with changes in politics and funding, as well as competing representations of resilience in different contexts.
Although the ending of the 100RC in 2019 offers only a short retrospective view of the impact it could have had on the global integration of the cities studied here, some of the discourses of its main stakeholders provide insights into the networks, models, or exemplary sites that are being created, disseminated, and promoted internationally. The globalization of these cities shows that their experience is increasingly shaping our urban world. It also demonstrates that the mundane dimension of resilience makes a difference. The discourses produced and the practices introduced, such as the many conferences, workshops, and visits organized in these resilient cities, have contributed to their globalization. The lens of violence and resilience brought into focus cities where, at least partly, hardship and adversity have been the breeding grounds for their international resonance.
Finally, the flexible and elusive definition of resilience developed by 100RC enabled these cities to produce a variety of discourses and practices in relation to the stresses and shocks they endured. Ultimately, though, it also led to some limits on the ground. Promoting a definition encompassing natural as well as social issues is certainly innovative, but it can also be challenging, especially when it involves issues of inequality and violence. Indeed, although many cities in 100RC identified violence as an important stress, the complexity of addressing such a problematic in their strategy eventually led many of them to shift their resilience agenda toward more conventional problematics like climate change or water management. Borrowing from some authors introduced in the last section, I have considered, through the lens of postpolitics, the technical-driven approach to resilience that some cities have developed, despite their initial focus on urban violence. In some cases, the disputed meanings triggered by a social approach to resilience did not fit into municipal agendas. Moreover, the innovative dimension of violence prevention projects in resilience strategies should also be treated with caution, because several of them were the continuation of preexisting ones.
As this study demonstrates, resilience is not an empty signifier; on the contrary, it is full of conflicting meanings and it needs to be put into context. As mentioned in the introduction, being resilient to future flooding is quite different from building resilience toward urban violence. To live with natural threats is unavoidable, but inequalities and violence are not inalienable in urban contexts. Just as cities have different ways of being modern, resilience has many shades of meaning, and seeking to standardize it might lead to a loss of its various nuances. In other words, the globalization of resilience risks reifying, romanticizing, and depoliticizing it, transforming it into a noncontroversial and postpolitical concept.