AAG Annual Meeting 2011


Association of American Geographers (AAG) Annual Meeting 2011
Seattle, USA
12-16 April 2011

"Space and Social Networks"
Preliminary Programme

Sponsored by the Political Geography Specialty Group

Dimitris Christopoulos, University of the West of England, UK
Christophe Sohn and Olivier Walther, Centre for Population, Poverty and Public Policy Studies, Luxembourg

Concept and objectives
Until recently, geographers and social networks analysts followed parallel paths. The former documented how different kinds of nodes and edges were organized into physical networks while the latter studied the way social actors are related, rarely paying much attention to their spatiality. In the last decade however, social network analysis has gained significant attention from different scholars, who have progressively investigated how the location of actors in space, the spatial practices or imaginaries, and the geographic arrangements of networks are influencing social ties.

This session is an attempt to consider both networks and space by integrating social theories in geography and spatial network analysis. The main objective is to investigate how an actor's position in geographic space can be analysed simultaneously with their position in social networks.

In the session we wish to explore the following topics:

  • Theories of spatiality in social networks
  • Spatial proximity and homophily
  • Governance networks and geography
  • Crime and terrorist networks
  • Innovation cluster topography and social networks
  • Small worlds and regional economics
  • Social movements and local embeddedness
  • Intra & Inter-organisational networks
  • Spatiality and relational space in workplace environments
  • Social network and GIS mapping


Space and Social Networks 1

Christophe Sohn, Centre for Population, Poverty and Public Policy Studies, Luxembourg,

The spatiality of social networks
The question on how space influences social network ties and structure has recently attracted increasing attention in a renewed way. At first, and from a theoretical point of view, this paper seeks to address three preliminary concerns: (1) what is the spatiality of the network? ; (2) What types of space do we mobilize? ; (3) What spatial variables are to be considered? The first question examines how the network is related to space and allows us to distinguish different types of spatiality. The two following questions determine the spatial effect to be tested by crossing different types of spaces with spatial variables. Following Massey, the spatiality of networks can be clarified through the distinction between a "sense of place" and a "sense of space". The former relies on the location of the actors and, more substantially, on their spatial practices and their symbolic attachment as the latter refers to the spatial assemblage of the network, that is its reach (distance between the nodes) and its scale, the spatial attributes of the embedded space (its substance), and the spatial evolution of the network (its dynamic). In a second stage, these theoretical considerations are confronted to more contextualized case studies (e.g. cross-border metropolitan regions in Europe) where cross-border policy networks operate. The aim is to scrutinize and discuss specific spatial effects over policy networks such as spatial distance, border effects (barrier or bridge) or territorial embeddedness.

Ivy Lingqian Hu, University of Wisconsin-Milwaukee, USA,

Spatial Characteristics of Social Networks
This paper reviews literature on the spatial characteristics of social networks, specifically if space affects the formation and maintenance of social networks. Social networks have been examined and applied in a range of fields, including geography, sociology, and planning. An in-depth study on the spatial nature of social networks becomes imperative as an increasing amount of research has found the significance of social networks, particularly in minorities' labor market outcomes. However, a gap remains between the theoretical concepts and the empirical application of social networks. Furthermore, no consensus has been reached if social networks can exist out of spatial context, particularly with the prevalent information technology. This paper fills in the gap by reviewing the theoretical literature in various fields, and explores if physical propinquity fosters the formation and usage of social networks. Furthermore, this paper reviews empirical literature which attempts to quantify social networks and confirms the strong spatial characteristics of social networks. This paper assists future research by providing both theoretical and empirical clarifications, and for efforts to develop communities which cultivate social networks and encourage social interactions.
Download the presentation

Céline Rozenblat, University of Lausanne, Switzerland,

Dynamic of World cities' system through multinational firms' networks: a multi-level and multi-dimensional approach based on agent and cities networks
The classical approaches privilege an object like the cities or the MNF asking what are the effects of the resources of cities on the development of MNF, or at the opposite side, what are the effects of MNF on the development of cities. Here one object is central and the other one is taken as exogenous variable. Better is to start from the point of view that MNF are a part of the cities' dynamics and vice-versa. Then, in order to build a "multidimensional approach", we can transfer and adapt concepts and approaches coming from other fields. The shift is to pass from a mono-dimensional network of cities linked by multinational firms developing some activity sectors to a multi-dimensional network linking MNF, cities and activities. Starting from the 600,000 subsidiaries of the 3,000 first world groups for 2007 and 700,000 for 2010, we'll take some samples in order to precise at the three levels the emergence of networks, the role of path dependence and the consequences for cities positions and dynamic in such patterns. We'll propose the development of Agent based simulation models adapted at each different level: agents will represent sometimes firms, cities or group of cities, groups of firms or activities. Then, we'll compare the results and we'll discuss their possible interactions.
Download the presentation

Edward Pultar, University of California, Santa Barbara, USA,

Social networks, travel behavior, social capital
Social networks are common in contemporary times and the use of digital social networks allows for new behaviors. This research examines how social networks change a member's travel behavior. Specifically, the CouchSurfing network with over 2 million members is used here as a case study. In this network, individuals on every continent offer each other a couch to sleep on while traveling. After members find each other virtually they meet physically face-to-face. This paper presents results from a survey showing what influences people to be active in this network and barriers that limit their activity. Additionally, social capital within this social network is examined. The methods presented utilize an explicit influence of geographic or spatial distribution of an individual's connections within the social network. GIS methods are used to visualize spatial behavior in these social networks.
Download the presentation

Petr Matous*,, Yasuyuki Todo, The University of Tokyo, Japan, Dagne Mojo, Ethiopian Institute of Agricultural Research, Ethiopia

Boots are made for walking: The spatiality of social networks in a pedestrian, phone-free society
The relationship between the spatiality of social networks and newly available communication technologies has recently attracted great popular and academic attention. In contrast, sufficient research has not yet been dedicated to the geography of social interactions in vast infrastructure-poor rural areas of the developing countries where walking is the dominant form of transport and talking face-to-face is the sole mode of communication. In three randomly selected wards in Tiyo Woreda, Arsi Zone, Ethiopia, we have gathered data from 297 randomly selected farmers on their 4078 relations to randomly selected associates and on 1246 relations to individuals that could provide them with tangible support. Half of the respondents' personal contacts lived within 15-minute walking distance and 87% of those questioned had no way to get in touch with their associates other than by walking to meet face-to-face. We analyze the farmers' frequency of meeting with others in relation to the estimated walking distance between their homes, the farmers' socio-economic characteristics, their personal network composition, and the type of their relationship. Preliminary findings of an ordered logistic regression suggest that a short physical distance, being of the same ethnicity, and kinship relation are the main determinants of frequent social interactions. These findings have implications for the planning of network-based interventions (such as information dissemination) in areas lacking communication and transportation infrastructure.
Download the presentation

Space and Social Networks 2

Olivier Walther, Centre for Population, Poverty and Public Policy Studies, Luxembourg,

Understanding the spatiality of cross-border governance networks
The aim of this paper is to investigate the simultaneous embeddedness of political actors in both geographical and network space. The case of the Basel cross-border metropolitan region located between Switzerland, France and Germany offers interesting perspectives. Using spatial and network data from an ongoing study of governance networks we test for distortions in network structure by looking whether spatial proximity, center-periphery patterns, and the presence of national borders affect the intensity of cross-border cooperation. In our work we hypothesize that there is a strong relation between geographical topology and network. We expect to find core network actors located in central geographical locations and periphery network actors located in peripheral geographical locations. We also assume that national borders will strongly limit exchange between actors, even though these actors are engaged in cross-border cooperation. The contribution of this study is to provide a better understanding of governance networks, by combining political geography and network analysis.
Download the presentation

Bettina Lelong*, Research Institute for Regional and Urban Development, Germany,, Gernot Grabher, HCU Hamburg, Germany

Policy networks in urban development processes: conflict versus change
The main aim of the paper is to highlight conflicts and changes in urban development policy from the perspective of network theory. It investigates the effects of specific local conditions and global trends on problem solving processes in the context of land-use conflicts about urban waterfront locations. The main focus of the research is to discuss the process of problem solving without a superior authority, which can determine the decision-making process in urban development policy hierarchically. The applied methodological approach combines the policy network approach with structuralist-constructionist concepts of Social Network Analysis (SNA) and elements of the network governance approach. The analysis focuses on mechanisms which create new possibilities for action in complex and constraining conditions. It analyses data from urban development processes in Hamburg, Cologne (both Germany) and Rotterdam (The Netherlands), which is collected in the context of a dissertation and a research project at the ILS Research Institute for Regional and Urban Development. The case-studies are analysed with regards to the structural features of the overall network, their specific network positions as well as the governance mechanisms between the involved actors. The findings show that change in urban development policy occurs in multiple phases. These are affected by different actors and network configurations, which are responsible for the generation of ideas, mobilisation of alliances and political consolidation. Amongst other mechanisms, 'publics' as special moments of social opening seem to improve the ability of actors to unite for 'fresh action'.

Jeremy Mennis*, Temple University, USA, Michael J. Mason, Villanova University, USA

The Spatial and Social Network Contexts of Adolescent Substance Use
This research investigates the influence of place and social network characteristics on substance use among a sample of 215 urban, primarily African American, adolescents in Philadelphia, Pennsylvania. We utilize survey data that captures the places that adolescents perceive to be risky and safe, as well as their home location. The survey also captures the egocentric social network characteristics that are associated with each type of place for each adolescent. Place-characterization data included indicators of a variety of physical features we theorize may enhance or mitigate the likelihood of substance use, such as the locations of alcohol outlets, recreation centers, and violent crime. We employ multivariate Tobit regression to investigate the relationships of place and social characteristics with substance use behavior. Substance use was greater for males and older adolescents, but was associated with few physical features of adolescents' home or perceived safe places. Substance use was enhanced by several characteristics of adolescents' perceived risky places, such as commercial activity in distressed neighborhoods. In addition, the presence of substance users at an adolescent's perceived risky place was associated with an increase in substance use. Substantial differences concerning gender and age were also observed. Girls and older adolescents tend to be more sensitive to social and place characteristics as compared to boys. This study highlights the importance of capturing activity space characteristics, perceptions of places, and social network data for investigating substance use, as well as for studies of other health and crime behaviors.
Download the presentation

Karen Haandrikman, Stockholm University, Sweden,

Do geographical roots matter on the partner market? An analysis of Swedish marriage fields The spatial dimension of the partner market is a dimension that has received little attention in recent decades. In a time when people have the possibility to find a life partner from all of the world, given increases in educational participation, affluence, social and spatial mobility, and internet access, this paper examines the role of geographical distance in the current Swedish partner market. The geographical origins of couples are explored in order to examine whether the place where one grew up is an attraction factor on the partner market. This is based on the idea that spatial homogamy, the similarity concerning geographic origin, enhances mutual attractiveness, as shared mentality, dialect and values lead to similar values and opinions. The level of spatial homogamy of Swedes is analyzed by using the PLACE dataset, a rich individual longitudinal dataset for the whole Swedish population for the period 1990-2008. New couples, defined as those married or cohabiting with a common child, are traced in the data, after which their former residential addresses and birth places are examined. Then, the distance between partners in a couple is measured at the time before cohabitation and at birth. The spatial variation in spatial homogamy will be the centre of analysis, with separate analyses for those with a migration history. Spatial regression techniques are employed to model which factors explain the spatial variation in distances to the chosen partner, taking into account spatial effects caused by the spatial mismatch between residential areas and partner search areas.
Download the presentation

<< back