*3.1. Methods*

To achieve our research objective we used social network analysis (SNA), a standard tool to study social structures, relations and flows [35]. A growing body of work has applied SNA to understand the relational patterns between firms and cities, mostly at the global/world level [17,22,36,37], but also at the regional [31] and multi-scalar levels [8]. Despite some critiques and limitations of networks as a metaphor to understand relational patterns and the resulting hierarchies in world cities [38–40], SNA has been providing relevant insights for assessing and modeling urban networks through corporate networks [41]. We draw on this body of literature, opting for a two-mode (also known as affiliation) network of cities and firms. This option allows us to explore how both groups of actors are associated to each other by means of Smart City projects in Spain. In this sense, the analysis conducted here expands the available evidence about Smart City networks using SNA to compare cities and firms at a national level, a dimension that has not received attention in the existing literature.

The data on the cities and the corporations involved in urban development through the RECI were compiled through Internet searches of the websites of the RECI, of the city councils, and of news and press releases between July and November 2016. The cities were identified by consulting RECI's official website; at the time of data gathering, the RECI was made up of 65 cities. The firms involved in Smart City projects in these 65 cities were identified through Internet searches of each city's official website (some cities, like Barcelona or Santander, have specific websites for the smart city development projects) and by using key search words such as the cities' names followed by "smart city", "smart city projects" or "urban innovation". Overall, we followed a broad inclusion criterion, searching for corporations that were taking part in Smart City related projects or actions in each city. This strategy resulted in 553 corporations involved in Smart City projects of RECI members, including firms taking part in Smart City actions at a local, regional and international level, for example as part of European consortiums. Of these 553 initially identified firms, 66 were participating in Smart City initiatives in at least two cities of the RECI and were included in the analysis. By doing so, we removed pendant nodes that would be connected to the graph by only one tie. Moreover, these pendant nodes corresponded mostly to cities participating in EU intercity projects, so this decision allowed us to focus on cases actually connected by the RECI association [42], which may help to understand the processes within the RECI at a local level. The number of initial firms was also reduced, as we consider different branches of the

same firm as a single firm. For example, Acciona Agua, Acciona Energía, Acciona Infraestructuras and Acciona Instalaciones were all considered as Acciona, for clarity purposes. As a result, the final network consisted of 47 cities and 66 firms.
