Conference Agenda

Agriculture contributes in several ways to the achievement of sustainability objectives. Its role in that context is highly controversial. There is no agreement as to what exactly constitutes “sustainable agriculture”. This is reflected in public discourses, which are driven by the perspectives and interests of various societal groups, including farmers, consumers and scientists. These discourses affect how the problem is perceived and which potential solutions are considered and, ultimately, implemented by policy makers. Also, they may affect individual actions of those involved in a given discourse. However, it is currently unclear to what extent the different discourses interact and overlap. In this paper, we investigate the sustainable agriculture discourses using Germany as a case study. We apply topic modelling to extract topics discussed in three large-scale text corpora: newspapers, agriculture magazines and scientific journals. We run separate topic models for each corpus and use the identified topics to qualitatively compare the patterns. Results show a continuing separation between the three discourses, both overall and across time. We find very limited overlaps between the corpora, and these were restricted to topics such as agricultural policy discussions. A particularly surprising result is the absence of climate change topics from the agricultural magazine corpus. These disparities across discourses may create challenges for formulating and implementing democratically legitimized policies to promote sustainable agriculture.

Based on the position of organic farmland in France in 2019, the objective of this paper was to identify local conditions for the development of organic farming. We began by identifying four characteristics that explain the heterogeneous development of organic farming practices in France using a Spatial Durbin Model to control the spatial autocorrelation of organic practices. This study shows that the protected designation of origin label has some ambiguous impacts on practices (i.e., positive for wine labels and not signifcant for

others labels). The proximity of a high demand for organic products, proxy by jobs and level of education of the population, influences organic development. Also, municipalities with low quality land and a high share of forest participate in the development of organic farming.

The Ricardian approach is a popular cross-sectional method regressing average farmland rents on climate across regions to assess the value of climate for agriculture. This paper investigates whether the Ricardian approach is able to recover such value when agriculture competes for land access with other climate-sensitive sectors. We theoretically show that, as long as sectors differently sort depending on climate, Ricardian estimates are actually biased. Instead, they could reflect spillovers from non-agricultural sectors which, by adjusting to climate shocks (either by expanding or shrinking), affect regional farmland area and subsequently contaminate average farmland rents (even when climate does not theoretically affect agricultural productivity). Such situations would be characterized by Ricardian estimates for non-agricultural sectors that are of similar signs than those for agriculture. A simple correction allows us to recover the standard Ricardian interpretation: to regress the product of average farmland rents by regional farmland area - that is, total farmland rents - on climate. We empirically test our approach on original data detailing the land rents for agriculture, forest and urban sectors in France from 2000 to 2018. In top of documenting climate-induced land-use changes between agriculture and the other sectors, we find that our standard Ricardian estimates do present evidence of related biases. Compared to our land-adjusted Ricardian estimates, they significantly underestimate the projected climate change costs for agriculture and the other sectors.

Land consolidation is a standard policy instrument to reduce the fragmentation of farmland by spatially redistributing land ownership. While its primary goal is to improve agricultural productivity, evaluation should integrate its potential impact on the landscape, as it may threaten its sustainability. The French consolidation program in the second half of the 20th century is a case in point. Often blamed for the drastic decline in hedgerows observed in the countryside, researchers debate its responsibility. Our study proposes the first causal estimation of its impact on hedgerows by applying a staggered difference-in-difference setting to a longitudinal survey in Lower Normandy, France (1972-2010). The results indicate that consolidated municipalities experienced an additional loss in hedgerow density of 13.55m/ha (standard error: 2.07). Importantly, this loss accounts for only 17% of the total decline, challenging the common narrative that consolidation is the leading cause of the hedgerow decline. Our analyses also reveal heterogeneous impacts over time since consolidation and consolidation period. Our findings call for explicit accountability of land consolidation for landscape damage while placing this policy instrument in a broader context of political, social, and market drivers of landscape dynamics.