PhD Defence: Julie Bompas

Julie Bompas
Cerdi, Université Clermont Auvergne

Examiners

Douadia Bougherara, Research Director INRAE, CEEM.
Philippe Delacote, Research Director à INRAE, BETA.
Vianney Dequiedt, Professor, Université Clermont Auvergne.
Catherine Araujo Bonjean, Researcher, CNRS.
Arona Diedhiou, Emeritus Research Director, Géoscience, IRD.
Catherine Simonet, Economist, AFD. 
Philippe Roudier, Economist, AFD.

Abstract

This CIFRE PhD thesis, funded by the French Development Agency (AFD), is structured in three chapters and focuses on climate services (CS) from the perspective of demand, use, and impact among farmers in northern Côte d’Ivoire. Climate services refer to the provision of weather and/or climate information to support informed decision-making among users. Two of the three chapters are based on survey data collected in cooperation with the private operator Ivoire Coton and in co-supervision with local consultants (CIRES and Hervé Kakou). 

Chapter 1 examines the relative value of integrating local forecasting knowledge (LFK) into climate services through a discrete choice experiment involving 285 farmers. The chapter begins by documenting LFK present in northern Côte d’Ivoire and assessing farmers’  perceptions of its reliability. Although farmers tend to place greater trust in LFK compared to available scientific forecasts, I show that the inclusion of LFK is not a decisive factor in choosing one climate service over another. Instead, the primary criteria influencing farmers’ preferences are the cost of the service and access to oral information (via radio or a cooperative focal point). Only older farmers show a marked preference for climate services that incorporate LFK.  These findings suggest that public policy should prioritize low-cost, orally delivered scientific forecasts in this region. Nonetheless, additional research is needed to explore the complementarities between CS and LFK, especially in light of farmers’ continued reliance on, and trust in, local knowledge in this region. 

Chapter 2 employs an experimental approach involving 313 farmers to assess: (i) how farmers’  maize sowing decisions respond to dry spell forecasts; (ii) the effect of such forecasts on potential harvests; and (iii) the individual factors that influence farmers’ decisions. The experiment asks each participant to divide their maize sowing between an “early and risky”  period—subject to a probability of dry spell occurrence—and a “late” period, which guarantees lower but certain yields. Forecasts of dry spells are randomly assigned across five hypothetical agricultural seasons and compared to a no-forecast scenario. I find that farmers adjust their strategies in accordance with the forecasts, leading to higher expected harvests relative to a no-forecast baseline. Sowing decisions are also shaped by previous decisions and past exposure to dry spell shocks. In addition, individual characteristics—such as locus of control, adaptability, and past experiences of crop loss—significantly influence how farmers interpret and respond to forecast probabilities. From an operational perspective, these findings underscore the value of probabilistic formats in communicating weather information to farmers with limited literacy. 

Chapter 3 presents a systematic review of the methodologies used to evaluate the ex post impacts of climate services in agriculture. The extent to which evaluators can control farmers’  access to climate services introduces different challenges: (i) spillover effects in control groups, and (ii) selection bias in both access to and use of information. Instrumental variable approaches can help mitigate contamination effects, though further research is needed to quantify their magnitude. To avoid selection bias, distinct measurement strategies should be employed depending on whether the evaluation concerns access or actual use. Most quantified impacts to date have focused on yields and farm management, while cash crops, livestock, and social and environmental outcomes remain underexplored.  This thesis concludes that individual heterogeneity plays a central role in shaping the demand for, use of, and impact of climate services. Overall, the findings support the need to consider farmers’ individual characteristics to maximize the potential benefits of such services 

Keywords

Climate services, adaptation, climate change, Côte d'Ivoire, climate, risk. 

theses.fr/s363272