Agricultural planning in Argentina

Guillermo Podestá, Poonam Arora, Federico Bert

The Argentina field project focuses on agricultural production in the Argentine Pampas. Modern agricultural ecosystems are a convenient domain to study real-world behavior: highly consequential decisions are made by individuals embedded in social and spatial (neighborhood) networks, and influenced by multiple uncertain drivers. The main goals of this field project are (a) to develop a model of agricultural production that incorporates recent advances in understanding of decision-making processes and social interactions; and (b) to implement decision experiments in a real-world context to enhance the interaction between the laboratory, field, and modeling components of CRED. The following paragraphs describe recent activities.

To deal with ambiguity in future climate and socio-economic conditions, we are exploring robust decision-making (RDM) approaches that examine the performance of multiple strategies over a wide range of scenarios. RDM approaches identify strategies that perform sufficiently well across most conditions (unlike maximization approaches that focus on a single, “best” solution). A robust strategy will perform almost as well as an optimal strategy found by classic decision analysis if future conditions turn out as expected, but it also will perform relatively well if unexpected future conditions arise.

We have conducted decision experiments to examine the importance and role of 13 distinct goals vis-à-vis farmers’ main production decisions. The goals considered are economic and social, as well as long- and short-term. A spinoff of interactions during CRED annual meetings was a proposal funded by the NSF focusing on the connections between climate, hydrology and land use. The complex two-way dynamics of shallow groundwater in flat plains set up an interesting social science problem: interdependence arises among neighboring farmers − that is, risks (e.g., floods) faced by an agent are influenced not only by her own decisions, but also by those of others. To explore the emergence of coordination, another decision experiment focused on use of a common good. We used a hypothetical example involving cooperation or defection of a common resource (fisheries).


Major Findings

  • Despite the importance of agriculture for Argentina’s economy, modern agent-based approaches had not previously been used to provide a “bottom-up” simulation of patterns emerging from the intertwined, cross-scale effects of human behavior and external drivers.
  • Our model reproduced observed land use and land tenure patterns: the growing dominance of soybean, the disappearance of smaller farmers, and rapid changes in land ownership regime – an increase in the proportion of land cropped by tenant farmers.
  • Given a realistic spectrum of strategies and profits for robust land allocation strategies there is very little difference among the top strategies identified by RDM approaches (defined by their “regret” with respect to optimal actions that would be taken with perfect foreknowledge). That is, most land allocations that “made practical sense” also were fairly robust to varying conditions.
  • Ownership of farmed land is a significant predictor of the relative importance assigned to different goals, and is not limited to any type of goal.  Thus, ownership predicts the relative importance of both economic and social goals.  Ownership also appears to lead to a greater emphasis on longer-term goals.
  • Land ownership status also introduces subtle nuances in the intent underlying farmers’ land decisions.  The motivational intent underlying owned land was to maintain the value of the land and lifestyle associated with the land, while the motivational intent underlying rented land was to maximize accrual of value from the land for the farmer and agribusiness.
  • Real-world and lab advisors in social dilemmas are not merely “rational economic maximizers,” but act in congruence with the psychological perspective afforded by the advising role. They focus on the positive potential and advise cooperation to others while defecting themselves: They would have advisees do as they say, and not as they do.


Broader Impacts

  • We are improving our understanding of how real-world decision-makers (e.g., farmers in the Argentine Pampas) cope with risks to their livelihood introduced by fluctuations in climatic and economic contexts. These insights have been incorporated into a realistic model of agricultural production that serves as a “laboratory” to explore the implications of plausible future climate scenarios and to understand the possible aggregate outcomes of individual responses to policies and incentives.
  • We held five workshops on “the fundamentals of decision making.” The workshops targeted Argentine farmers (12-15 in each workshop). The workshops were conducted by F. Ruiz Toranzo, an Argentine outreach specialist who was trained and assisted by CRED investigators (mainly P. Arora) in developing the contents of the workshop.  The workshops will be continued in 2013, with 5-6 more workshops to be conducted by the end of 2013.
  • A workshop on the dynamics of shallow groundwater in the Pampas in July 2012 brought together several farmers, technical experts and political authorities from the semi-arid Pampas. During this workshop we conducted an exercise on use of a common good. The results from the exercise were communicated back to participants via the following web site:


CRED2 Award (2010-2015): Funding was provided under the cooperative agreement NSF SES-0951516 awarded to the Center for Research on Environmental Decisions.