Models and their Effects on Development paths: an Ethnographic and comparative Approach to knowledge transmission and livelihood strategies

MEDEA is a collaborative project funded through the European Commission’s FP7. The project is led by the Anthropology Department at Goldsmiths and 5 international partners: the University of Barcelona, the University of Bologna, the Comenius University of Bratislva, the Instituto de Desarrollo Económico y Social of Buenos Aires and the University of Brasilia.

The project will use fine-grained ethnographic data in conjunction with qualitative and quantitative models to examine and predict patterns of adaptation, networks of socio-economic expansion and potential opportunities for innovation in the contemporary global context.



Taking a comparative and inter-generational perspective, the project addresses the impact of development paths on livelihoods, transmissions of knowledge and innovative responses to changing local and global circumstances. Focusing on four detailed case studies (Argentina, Brazil, Slovakia and Spain) the project traces patterns of connections between different forms of employment and unemployment, relations of solidarity and cooperation, and the elaboration of alternative or complementary entrepreneurial projects across the social spectrum. Starting from the premise that the analysis of any (dominant or alternative) development path must be situated within the complexities of historically unfolding links and relationships, the project explores how these emerge in specific socio-economic environments.

Drawing on a theoretical background that engages critically with development models, the project draws on the work on global trajectories carried out by the research team in Brasilia. The interdisciplinary approach combines qualitative research and comparative methodologies with modelling to explore the dynamic effects of development models and their context-bound implementations, and to trace and outline the interconnections of micro and macro level phenomena. By focusing on heavy industry, primarily the steel industry, the research will identify critical points in the transmission of knowledge and skills across generations, regions and economic spheres. An ethnographic approach enables the detailed recording of social networks (including those of solidarity and support), both within work places and formal economic activities and beyond them, in informal economies and strategic friendship, kinship and neighbourhood relations. These are envisaged as potential vehicles for the reconfiguration of work practices and new forms of economic activity. The project will thus contribute to the comparative analysis of models for both policy and entrepreneurial action and will elucidate the complexity of development paths as they unfold within and across social terrains.

Collaboration of anthropologists with sociologists and economists is central to the project, and in particular to its aim of exploring the relationship between models and qualitative research. Using System Dynamics simulation techniques, the research team at Bologna University will explore the implications of the range of variables emerging from ethnographic fieldwork in the four sites. At the same time, this exercise will provide critical purchase on the limitations of SD, particularly its assumptions about causality and its reliance on notions of individual behaviour.



From the widest perspective, MEDEA aims to make a substantial contribution to debates and interventions regarding sustainable development in Europe and beyond. We anticipate that the data generated through the research will provide significant insights into the complex interrelationships between regional development paths, in particular concerning Europe and Latin America, and more general proposals and recommendations regarding the design of more effective and appropriate models, responsive to the specificities of socio-economic contexts and their particular, historical trajectories.

Ethnographic research in selected sites will also enable identification of specific configurations of factors that may be detrimental to sustainable strategies, whether at the level of individuals or societies. The research will, however, also explore and record, where appropriate, those sets of circumstances which have generated innovative responses from individual or institutional actors. In this respect, the research will be particularly attentive to the effects of different global flows and of potential discontinuities and inequalities in the distribution of strategic resources and opportunities across regions, classes and genders.

The potential impacts of this research have been highlighted (explicitly and implicitly) throughout this proposal. Here we wish to stress that the project will:

  1. lead to a better and more comprehensive understanding of processes by which knowledge, innovation and entrepreneurial skills can be encouraged to flourish, and
  2. help to anticipate, identify and hence counter or even overcome the unforeseen consequences of economic change.

The driving force of this proposal is to facilitate and put into action research which has a direct bearing on economic exclusion, inequality and poverty.

The purpose of the final conference is to disseminate the work of the project and to facilitate dialogue across disciplines and between academics, civil society organizations and policy makers.


The main goal of the Global Flows work package carried out by the Brasilia Team was to provide theoretical perspectives and frameworks in order to analyse the global flows of development models. In order to achieve its goal the Brasilia team, consisting of 17 members, was subdivided into three research groups.

Team 1

Team 1 elaborated a framework for the analysis of global flows of development models based on a discursive matrix. This matrix is transmitted through two principal means: a diffuse a concentrated mode of dissemination. The team explored macro, meso and micro models, considering how resistance to global flows takes place. The team also focused on how capital flows, fusions and mergers influence global flows of models and prefigure social, economic and political realities. We prioritized the history of steel making in the Steel Valley of Minas Gerais, the oldest area of metallurgy and steelmaking in Brazil. The election of the Belgium Mineira Company as the main concrete focus of the research proved highly productive in tracing changes in model making and transmission since the early 19th century. We trace the trajectory up until the 21st century when this major steel plant is incorporated by Arcelor and later by Mittal.


Team 2

Team 2 prioritized the revision of the literature concerning the intersections of global flows, state configurations and development models. The goal was to articulate the latter with the colonial and transnational flows of people, models and technologies of governance while considering, at the same time, internal political and social pressures. From the socio-historical literature, the team was able to identify some variables that play a conditioning or enabling role in the making of state and development models. The team concluded that such phenomena are conditioned by colonial legacies, by ethnic and regional conflicts, by the position each country had during the Cold War, by different trajectories, projects and interests of the elites and by the constraints imposed by the international cooperation field.

Team 3

Team 3 focused on the dynamics and structures of the relationships between multilateral agencies of international cooperation and civil society actors, especially concerning the promotion of models and knowledge transmission for development in several parts of the world. The team considered that when different multilateral agencies of international cooperation promote the worldwide globalization of development models they also promote the globalization of models of civil society’s involvement in developmental practices. This involvement, in turn, is backed by conceptual models that include sets of relations between civil society, the state and the market. The research showed that these models may contradict but also complement each other. In fact, they increasingly converge in relation to their orientation towards the market. The team worked with several multilateral agencies (UNDP, OECD, European Union, Organization of the American States, IMF and the World Bank) and a range of civil society organizations (NGOs, social movements, networks, etc.). The relationships amongst these actors occur with greater or lesser degrees of involvement of other actors and institutions (state actors, private sector actors, other international organizations). From a historical and sociological perspective, the team identified several scenarios structured by “developmental” initiatives and by “poverty alleviation” interventions, involving local, regional, national and international practices.

The Global Flows workpackage identified processes, mechanisms, structures and dynamics of global flows and connections related to development models in conditions where capitalist firms, state organizations, international cooperation agencies and civil society’s organizations play a major role.  The analysis is supported by reference to a number of representative cases that illustrate the diversity of the actors involved. We highlighted some of the interactions’ main impacts and trends.


Civil Society

Discussions of civil society in parts of social sciences and politics have been narrowed to Western models of liberal individualism. The anthropological task -- and our major aim dealing with civil society- requires investigating the ‘actually existing civil society’: informal, interpersonal practices, kinship, friendship and community ties, church associations, people’s everyday actions and so on. Our objective is to describe locally/regionally grounded civil society alternatives. We pay particular attention to the steel sector, and trace civil (and uncivil) movements among workers. We are interested in learning about local/regional alternatives of social emancipation as well as in policies of development and how/whose people are involved in them.

Our approach

Under the work package titled "State and civil society" we interview key representatives of the elites in Spain, Slovakia, Brasil and Argentina. The representatives include the political and NGO leaders, academics, trade union leaders, representatives of employers and think/tank analysts. The analysis of interviews serves as the entry point for long-term ethnographic fieldwork in actual locations.

Our approach combines the perspective of our interviewees (analysis of the interviews) and wider analytical perspectives on the development of state - civil society relations in each particular case, to map key societal transformations that influenced formation of state and civil society relations. In all cases industrialisation, nation-building, and neoliberal transformations (including privatization) represented major forces of transformation.

There are important similarities in the legacy of autocratic regimes in the development of civil society. We identify several discourses with reference to past legitimation, present meaning and future projection of the civil movement and activism in each case. Among the key factors of civil society/state relations in all four locations appear global restructuring the heavy industry and consequent de-industrialisation. The effect of these processes has been fragmentation of class interests, less organised labour movements and formation of identities reflecting ethnic, racial and other “post-modern” forms of identification, including conservative mobilisations.  In this context, populism appears to be specific form of mobilization among workers.



Through ethnographies of generations we explore the trajectories of workers, families and companies. We pay particular attention to the transmission of crucial resources such as knowledge, social relations, networks, access to employment and/or subsistence. The study of generations is located within the historical trajectories of each site, noting the effects of critical events on the process of intergenerational transmission.

Our approach

We reconstructed actors’ perspectives through in-depth interviews and personal histories, which were contextualised in relation to key events in local histories.  To address the issue of change and continuity across generations, we take into account two temporal dimensions: changes in working conditions and intergenerational changes in the experience of work and access to skills and employment.  What we find in the steel industry are significant differences in the level of training and education required to access jobs in the industry as well as marked differences in recruitment practices across the different sites. Another important feature of work that distinguishes the younger generations from older workers is the growing precarity of employment, which are reflected in different experiences and attitudes to work.




Ethnography of livelihood focuses on the ethnographic account of the multiple connections and relationships that underpin work practices, livelihood pursuits and other survival and/or advancement strategies in specific contexts. We will focus on the formal and informal connections between historically constituted ‘habitus’ of work, family, and knowledge transmission processes. Our objective is to identify significant continuities or ruptures within specific spaces framing the circulation of goods, knowledge, skills and labour, and to asses their impact on age, gender and class attitudes and capabilities regarding work and entrepreneurship opportunities.

Our approach

The work package titled "Ethnography of livelihood" uses conventional ethnographic techniques such as ethnographic interviews and participant observation. Snowballing techniques and tracing networks enables the researcher to straddle the boundaries of work and sociality, of family life and livelihood activities, of formal and informal work.  These rich, descriptive data will be supplemented through structured interviews and questionnaires to generate a data base on households and work units.

Our approach aims at addressing how people cope with change in the work opportunities  for self and future generations. We want to address questions such as: What resources circulate among livelihood spaces: Work, knowledge, skills, information, goods?  What new connections emerge between differently defined spaces (formal / informal; home/ firm)? How are social relations re-configured spatially in a new industrial conjuncture (subcontracting chains, homework, precarious work, care networks, migration etc.)? How do the livelihood projects of individual subjects, households, firms intersect?

Livelihood spaces as defined by different actors in particular locations, create habitus, a framework for expectations both of stability / continuity across generations and of mobility (both physical and social) between generations. Opportunities are assessed and socio-economic projects defined in particular spaces endowed with material prospects and meaning.

Ruptures in socio-economic expectations re-configure spaces both public and private (home, firm, and their boundaries) and produce tensions in the habitus. This leads to adaptations and innovations, but also to resistances and increased uncertainty.


Pattern of Privatisation

Consumption / Production + Privatisation


The logic and the aim of modeling

In our study, computer modelling and simulation, provided a bridge between such diverse objects asquantitative time series and verbal explanation of a longitudinal phenomenon. More specifically, weused a computer model as a theoretical laboratory to test whether and how different way of modelling acausal model underpinning an observed phenomenon generates longitudinal patterns that are comparablewith those empirically observed. In particular, the phenomenon observed is the proportion of state versus private capital on the steel industry in the time period 1945-2010.

In this respect, we adopted a logic of inference that we might call abduction. That is, given an observed phenomenon, we used computer modelling and simulation to derive hypotheses that, if true, wouldsuggest a plausible and coherent explanation of the observed phenomenon. Thus, in our work, computer modelling and simulation are used to put together an environment for theory development.


The product of modelling

The product of our work, thus, is a computer model that can be used to generate hypotheses on how social, economic and political pressures produce patterns of privatisation. This kind of method is a form of computational “thought experiment:” in which we ask “what if” questions in an artificial world. However, the ultimate aim is to allow us to develop hypotheses and theories that can then applied to real world phenomena and data. Our ultimate aim is to understand the real world. We use the computer model at this stage to help us to generate and test, in a rigorous and deductive way, candidate ideas.


The process of modeling

The development of the simulation model unfolded in three steps:


Qualitative model of steel industry

First, we generate a qualitative model that portrays key concepts and processes that recur in discourse on the evolution of steel industry. The sources of the information that underpin this first attempt are the reports produced by each team, which we refer to as the country reports. These reports narrate history on steel industries in Argentina, Brazil, Slovakia and Spain. Grounding on these reports, we developed the qualitative model of cause-effect relationships that characterize steel industries. In this phase, we tried to balance the need for generalization with the need to maintain idiosyncratic features of each country. As a result, we developed a general model that include general concepts and cause- effect relationships and sub-models that address country-specific issues. More specifically, to extract key concepts and elicit a web of cause-effect relationships among the concepts from country reports, we conducted a content analysis on the texts


Base computer model of steel privatization

The second step was the development of our base model for privatization. Here, we referred to three kind of sources. First, again, we referred to country reports where these latter contributed information concerning privatization processes. Second, we used extensive interviews and workshops with informants. Third, we used secondary data such as official reports from international organizations (e.g. World Bank, OECD), books and newspapers.


Final computer model of steel privatization

The final step of our research process was the development of the final version of the model. This last step refers to the calibration of the model with real empirical data. The empirical data were collected by the modeling team and were verified with members of other research teams. In data collection, modeling team referred to data from the World Steel Association but also was supported by OECD, IRI Historical Archives, Information Centre Techint of Buenos Aires, Assolombarda and Historical Archives of the European Union. In the final step we used the computer model to devise hypotheses to explain the observed patterns of steel industry privatisation. In particular, we compared the cases of Argentina and Spain, for which more data were available. In addition, to increase confidence in our model, we collected data on the process of Italian steel industry privatisation and investigate how the mechanisms depicted in the model could support the understanding of that context.


Key issues and learning points: the MEDEA research approach

Under a methodological point of view, MEDEA has been a challenging project. The challenge materialized in two directions. First, being a multi-disciplinary project, it required scholars from different disciplines, such as anthropology and economics, to build a common language and to iteratively and reciprocally integrate underlying ideas on what contributions the project was aimed at. Second, MEDEA proposed a mixed methodology that combines qualitative research, formal modeling and computer simulation. This choice required involved researchers to make an effort to develop such an original
approach by discussing and adapt entrenched ideas on research methodology. In this light, the project was very useful in developing a new research approach. In the following, we report key learning points the
emerged from the project and that can be useful to build future research projects that intend to replicate the MEDEA approach to formal modeling. In proposing the following items we make four assumptions
(i) we refer to large research projects that include a number of partners; (ii) the project requires at least one deliverable that foresees the formal modeling of a general issue thereby requiring information and involvement of all other partners; (iii) there is one research team that coordinate and is responsible for the modeling process; (iv) we take the point of view of the partner that is responsible for the modeling process.


Importance of plenary meetings to explain methodology at the beginning of the project. We set two of plenary meetings at the beginning of the project to explain the methodology to the partners.


Importance of plenary meetings to inform on the progresses. They proved useful and we suspect that more plenary meetings should be positioned along the unfolding of the project to present results, and to share problems and issues.


Divide the modeling process in sub-processes and sub-models. We suggest that the modeling
process should be divided into sub-processes leading to different sub-models. For example, in
the MEDEA project, we worked on parallel to different sub-models that, despite their sharing
the same core causal structure, were adjusted and calibrated to reflect specific countries. This
efforts scores two goals. First, it balances tension to generalization with awareness of country-
specific issues. Second, the articulation in sub-models encourages and stimulates involvement of
researchers of each team.


Define model-owners for each sub-model. To foster the involvement of researchers into the
modeling process, each sub-models should refer to a model-owner who is not a modeler. Rather,
the model-owner should be a researcher belonging to one of the partners’ team different from the
partner that is responsible for modeling.


Plasticity of modeling team: the modeling team should integrate the need for involvement and participation of researchers with the flexibility required to achieve results. Thus, we suggest that the modeling team should have plastic boundaries. Following, we describe what different shapes the modeling team can take.


Core Modeling Team (CMT) – Leads the modeling process. In the CMT are the researchers belonging to the partner that is responsible for modeling.


Focused Modelling Teams (FMT) – Each FMT includes the members of the CMT and the members of other partners that collaborate with the CMT on the modeling of a specific issue leading to a sub-model. These members are representatives of the partners to whose teams they belong and are the bridge between a specific partner research team, which owns a specific sub-model, and the core modeling team.


Enlarged Modeling Team (EMT) – includes the members of all the operating FMT. The EMT meets to share issues and problems that emerge in the modeling sub-processes.