Does participation in innovative farmer groups allow reducing pesticide use? Evidence from French vineyards of the Dephy network.

Margaux Lapierre, Philippe Le Coënt, Alexandre Sauquet, Julie Subervie

In 2008, France has set up the national action plan Ecophyto with the objective of halving the pesticide use by 2018. One of the main challenge is to do so while maintaining high crop productivity. The Dephy network was created in 2010 to demonstrate the feasibility of this objective. It is constituted of groups of a dozen farmers which are designing and testing various approaches to reduce their use of pesticides, with some technical assistance from an engineer (Pillet, 2014). This individual and collective support combined with the sharing of knowledge and experience must contribute to deploy efficient agricultural systems and practices. In order to establish whether this initiative should be maintained and expanded, it is crucial to assess whether the Dephy network has achieved its goal. In this study, we aim to provide evidence on the impact of the Dephy network on the use of pesticides and productivity in wine production, using microeconometric methods of impact evaluation.

Recent empirical studies assess the potential for reducing pesticide use by half in the French arable sector. Some lead to the conclusion that the use of pesticides could be significantly reduced in French field crops without affecting farmers’ incomes (Jacquet et al., 2011). Others quantify significant yield losses resulting from a 50% reduction in pesticides use (Hossard et al., 2014). First results based on data collected on Dephy farms suggest that pesticide use could be reduced by about 40% in more than half of farms, all crops combined, without lowering both productivity and profitability (Lechenet et al., 2017). Because the participation in the Dephy network is voluntary, an analysis supported only by data on Dephy farms suffers from selection bias. The type of crop considered also seems to be a crucial parameter in such an analysis. We contribute to this literature by proposing an analysis that explicitly takes into account the self-selection of the participants in the network. We focus on winegrowing, which is the cropping system with the highest levels of pesticide use.

Our main parameter of interest is the additionality of this program: how much pesticide reduction has been achieved thanks to the program. To address the issue of selection bias, data on Dephy farms’ practices are compared to data collected on non-Dephy farms for the period 2010-2016. We use various identification strategies to estimate the impact of the program including matching procedures and difference-in-differences methods.

Our project contributes to the third theme of the symposium: Innovation in techniques, organisations, products, and communication.