The development of effective policies for agricultural adaptation to climate change requires an understanding of how impacts are related to exposures and vulnerability, the specific dimensions of the climate system that will undergo the most significant changes, where human impacts will be most severe, and the institutions best suited to respond. A study conducted through a partnership between the University of California, San Diego (UCSD), and Universidade Estadual de Campinas (UNICAMP) evaluated how the agricultural sector in Brazil (production, prices, and loans) is affected by variations in the more predictable components of temperature and rainfall, including trends, seasonality, and unexpected shocks. The study found increased variation in yields and revenues and higher agricultural loan defaults. The authors discuss how resilience strategies should focus on institutions such as water storage, financial services, and reinsurance.

The paper was published at the PNAS - Proceedings of the National Academy of Sciences (click here for full access) by Jennifer Burney (UCSD), Craig McIntosha (UCSD), Bruno Lopez-Videlac (UCSD), Krislert Samphantharaka (UCSD), and Alexandre Gori Maia (UNICAMP). The study used an innovative strategy to decompose the historical variation in temperature and precipitation into five components: i) the original local average of temperature and precipitation; ii) the time trend in each location (municipality); iii) the seasonal deviation from these trends in each location, iv) the national average of the total shock in each period; v) the idiosyncratic shock, i.e., shocks that are specific to each location and period. The authors initially conducted a retrospective analysis to measure productivity (yields and revenues) and economic vulnerability (repayment performance of agricultural loans) in relation to the different aspects of weather variation. They then used the coefficients obtained from the retrospective analysis to forecast future impacts by using climate model projections for Brazil.

Findings show that climate change in Brazil is expected to have negative impacts that ripple across the agricultural sector—from yields to revenues to loan repayment performance of agricultural borrowers. Local idiosyncratic precipitation shocks are detrimental to all outcomes, but the impacts also depend on the nature of the shock, and how widespread it is. Importantly, as the shocks become more correlated across regions, existing endogenous price effects (i.e., higher prices when production drops) may only go so far, and we may also expect more detrimental financial impacts as the climate warms. The authors also found evidence that investments in resilience infrastructure like cisterns do weaken the link between seasonal weather fluctuations and agricultural loan repayment, and are overall associated with better performance.

The final discussion highlight how intentional design around resilience not only has the capacity to help individual producers but also to build resilience into the financial sector. The study also shows how a concerted effort to link longer-term climate, agricultural, and loan data would greatly improve understanding and prediction of climate impacts. This effort could also enhance the evaluation of adaptation strategies and better align academic research with the needs of stakeholders and practitioners.