https://orcid.org/0000-0001-5569-858X
![Sample our Built Environment journals, sign in here to start your access, Latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5926/TOC-Subject-Sample-2021-Built-Environment.jpg"})
ABSTRACT
The government of large metropolises has become increasingly complex during what many authors have characterized as a transition from government to governance. This paper tackles an unexplored facet of this complexity, namely the circulation of local policy elites within government structures through time in different institutional and political contexts. We analyse the networks created within the state by the circulation of individuals between the top positions of local governmental agencies in three large cities: Sao Paulo, Paris and Milan. The cases demonstrate dense networks in each city connecting policy sectors, administrations and government levels. Contrary to accounts maintaining that governance has reduced government structures, all cities showed constant increases in network size and connectivity, mirroring the expansion and specialization of local bureaucracies. The cases also indicate the resilient salience of the politics of political parties, policy fields and – especially in Paris – different government levels. This suggests that governance involves a reorganization of government functions and increased policy specialization, but this is not done by reducing the role of politics or the state itself. Lastly, in all three cities we find a greater centrality of the same kind of agencies more strongly associated with urban infrastructure and planning.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the authors.
Notes
1 Another analytic difficulty was caused by changes in administrative structures and policy responsibilities over time. In São Paulo, for example, social assistance policies were delivered by municipal secretaries for family and welfare, social welfare, social assistance and social development. For our purposes, these were all considered the same agency, reclassifying them to maintain as much consistency as possible about who delivered which policy, regardless of their names and apparent functions.
2 The data came, respectively, from the ‘Relação de nomes, cargos e endereços do governo do Estado de Sao Paulo’, published every four months by the Fundação Seade in São Paulo, the yearly ‘Bottin administratif’ and the ‘Bottin des communes’ for Paris, and documents published by the Chamber of Commerce for Milan.
3 We also analysed the circulation of individuals to other municipalities and to state and federal posts (elected and not elected), but just to control for their careers in descriptive statistics. These posts were not included in the network.
5 For a detailed analysis of the São Paulo case, see Marques (Citation2017), covering a broader period (1985–2012), as well as including education, health and social assistance agencies, totalling 2032 individuals, or 5042 if we also consider the top three tiers of government.
6 Throughout, initials after acronyms of the political parties indicate their relative ideological position: thus R, CR, C, CL and L refer, respectively, to right, centre-right, centre, centre-left and left administrations. These classifications follow common practice in the local literatures.
7 The Parisian case study builds upon results published earlier (Barwick & Gross, Citation2019). This article also contains a complete and detailed account of the methodology and the details per policy sector.
8 The mayors of the arrondissements, for example, correspond to submayors in São Paulo, two administrative levels below the mayor, although they have the same name.
9 Currently, she is still the incumbent, but the study includes data just for 2015.
10 Companies include the management of public transport, construction and management of metro lines, supply and management of water pipes, management of roads, municipal public housing, planning, management of sports infrastructures, management of pharmacies, provision of foods, and general markets for food. Housing and metro lines are managed by the same company and are grouped here under the label ‘mobility and infrastructure’.
11 Information came from the Ministry of the Interior (http://amministratori.interno.it/AmmIndex6.htm), while data on education and careers came from CVs available on the internet. Most of this information has not been used in the present article. For a broader account, see Andreotti (Citation2018, Citation2019).
12 Technically, we started with tables of individualsXposts in agencies under each administration (one affiliation matrix per city). These were transformed into a table of individualsXagencies–administration (another affiliation matrix), and later into tables of agencies–administrationXagencies–administration (connectivity matrix). Finally, was built by collapsing agencies of the same administration in both rows and columns. The latter connectivity matrices (one per city) were used to build the networks to be analysed (as sociograms and measures) in the next section. All three cities followed this same procedure.
13 The cells above the diagonal have been left empty since the matrix is mirrored around the diagonal, meaning that the upper cells are identical to the lower ones.
14 This dimension is strongly influenced by the institutions that regulate appointments and varies considerably between our three cases. If top positions are associated with fixed terms that match mayoral terms, for example (as in Milan`s public companies), individuals clearly cannot migrate during the administration in question. The Parisian case presents intermediate numbers due to the highly consolidated administrative structure but with a lack of fixed terms for many posts.
15 Considering the three upper positions of each agency (and not just the two upper tiers of the hierarchy considered in CitationTable 2 and throughout this article), the total number of posts jumped astronomically from 1316 to 4278 during the same period.
16 Degree corresponds to the number of connections between one node and other nodes. Density is the proportion of existing connections divided by those theoretically possible. Centralization compares the network around the most central modes with a completely centralized network (like a star in which all the connections radiate from the centre). All network measures were calculated at Ucinet 6.691 or Gephi 0.09.2 and all sociograms were generated in the latter programme.
17 To standardize the analyses, the positions of the nodes in all sociograms were determined by the Gephi procedure ‘Multigraf Force Atlas 2’, subsequently edited to enhance legibility.
18 These are: Urban: transport, housing, infrastructure and works, sub-municipalities, services, urban development, urban control and environment; Political articulation: cabinet, government, government relations; Administration: finance, legal affairs, budget, audit office; Social: culture, sports, employment and economic development; and Others: food supply, security, communication and several others of very short duration, such as support for the organization of the FIFA 2014 World Cup.
19 The agencies in these three policy fields have been merged from the following policy sectors: Administration: city government, finance, information, legal, security, service; Social policies: culture, education, health, sports, welfare; and Urban policies: environment, habitation, infrastructure, transport, urban development.
20 To obtain the different sociograms, we began with the connectivity matrices of Agency–administrationXAgency–administration, as described above in note 10, to create undirected graphs where the nodes are specific organizations, territorial levels or policy areas, and the weighted edges (or links) are the number of persons connecting them.
21 Pharmaceutics refers to the management of pharmacies. The original idea was to ensure that pharmacies were equally distributed within the municipal boundaries.