Challenging the establishment: A computational grounded theory of the emergence of sustainable food companies in Colombia
Keywords:Agent-based model, markets, semi-structured interviews, social complexity, text analysis
Understanding the underlying mechanisms that drive institutional change is a necessary step toward developing paths to sustainability. One possible instance is studying the emergence of new market logics in food systems. For this, we focus on a case study in Colombia. Current industrialized food production systems have been under scrutiny for their negative impacts on human health and the environment. New businesses have been emerging as a social movement and are being formed with a different market logic. Organic food producers are striving for their establishment in a market dominated by industrialized food producers. These organic food producers favor small-scale organic food production over industrialized food chains, and focus on social rather than on monetary profitability. To address this case study, we use a mixed-method approach that combine data from semi-structured interviews carried out to 30 small and incipient organic Colombian food companies, a subsequent computational text analysis, and an agent-based computational model. We focus on the question: How can markets transition to a logic of sustainable food production systems?
Backing up the agent-based model construction with the key dimensions, actors, and firm strategic elements identified by text mining techniques, the model aims to explore actionable strategies that lead to the emergence and establishment of sustainable enterprises.
This case study casts light on the intricacies of sustainable transitions, as well as it serves as an example of the importance of mixed-method research to study social systems / social complexity.
We carried out semi-structured interviews on 30 small food producers in Colombia. Results were analyzed by (i) finding meaningful word associations through bigrams, (ii) inspecting representative interview terms using term frequency / inverse document frequency (tf-idf) techniques, and (iii) interpreting major topics in groups of words (topic modeling). These text mining techniques allowed for identifying key dimensions of interest among producers, relevant actors and interactions in the organic food production, and important element in firm’s business strategy. This information provided input for building an agent-based model to explore plausible theoretical scenarios for institutional change.
Text analysis revealed five dimensions of interest among interviewees: two of them are related to traditional markets (marketing issues, legal legimitation), and the remaining three deal with alternative market approaches (human and nature driven motives, consideration of healthy lifestyles in consumers preferences, and social profitability). We used tf-idf related techniques to identify salient actors (e.g., consumers, producers, community, etc.) Topic modeling was used to identify recurrent themes in the way interviewees approach their businesses.
Nonetheless, text mining techniques are limited in inspecting temporal implications of interdependent choices made by actors. Being this a complex social system, we then built an agent-based model to theoretically explore plausible scenarios for sustainable transitions.
Agent-based model, markets, semi-structured interviews, social complexity, text analysis