Conference Agenda

Climate variability, marked by anomalies in temperature and precipitation, remains a “threat multiplier” and shapes the conflict and household-level food security situation in developing nations. But the economic mechanisms linking these variables remain insufficiently quantified. We used high-resolution climate and georeferenced conflict events data matched with nationally representative household panel surveys to establish this nexus. First, the study examines the impact of anomalous indicators of climate variability on localized violent conflict using the random-effects negative binomial (RENB) regression. Second, the two-stage least squares (2SLS) framework was used to establish the impact of violent conflict on household-level food security outcomes in the presence of climate variability. The findings reveal that the maximum temperature and maize price anomalies decrease conflict exposure, while the minimum temperature and precipitation anomalies significantly increase violent conflict. Furthermore, conflict and maximum temperature anomaly significantly reduce per capita food expenditure, positioning Nigeria at a serious threat to achieving food security. In contrast, minimum temperature and precipitation anomalies significantly increase food security outcomes. Economic shock mitigates the relationship between climate variability and violent conflict, and conflict and food security outcomes, suggesting the critical role of economic mechanisms in this nexus. We also found heterogeneity in conflict events and household-level food security outcomes across northern and southern geopolitical zones. These findings suggest the role of economic mechanisms in mitigating environmental shocks and provide evidence based on pathways for designing climate-resilient social safety nets in conflict-prone agricultural economies.

The impact of climatic shocks on the livestock sector is under-investigated despite the sector's significance in global food security and the potential for poverty reduction. This paper examines how heat stress affects smallholder farmers' cattle holding and milk production. Using large-scale panel data, containing about 262,000 rural households, from Ethiopia matched with high-resolution weather data; we show that heat stress above a certain threshold reduces cattle holding, daily milk production per cow, and total annual milk production. A one-day increase in humidity-adjusted temperature reduces daily milk production per cow, annual milk yield, and cattle holding by about 13%, 21%, and 14%, respectively. The main mechanisms that explain the reduction in cattle holding are birth reduction and cattle death following heat stress. Heterogenous impact assessment suggests that participation in livestock extension services is vital in reducing the impact of heat stress on milk production and cattle holding. We estimate the total economic loss from heat stress in the smallholder livestock sector to be USD $358.3 million/year. We argue that policymakers should consider cattle breeds, production systems (nomadic and non-nomadic), herd size, and lactation periods in developing adaptation strategies in the sector.

We study the impact of same-day heat on productivity and labor responses among smallholder farmers in South India, who have been contracted to grow cucumbers for an agricultural season. Harvesting is labor-intensive and conducted outdoors, leaving workers exposed to heat stress under high temperatures. Using high-frequency administrative data matched to hourly weather records across dozens of villages, we estimate the relationship between same-day temperatures and daily yields as well as earnings, and examine the mediating role of labor. We document a drop-off pattern in the contemporaneous temperature–yield relationship: yields are flat up to 33°C but decline beyond this threshold. On average, on days hotter than 33°C, farmers have 16% lower yields and earnings. We find supportive evidence that labor constraints may underlie this heat penalty.

Access to water is a critical determinant of agricultural development, yet its redistribution via irrigation often reshapes local production structures in unequal ways. We exploit the staggered expansion of water-withdrawal infrastructure in France over the last five decades to estimate the long-run effects of irrigation access on agricultural productivity and farm structure at the municipal level (about 4 km x 4 km on average). Our main specification instruments local irrigation adoption with regional diffusion shocks in irrigation infrastructure (``adoption waves''), isolating plausibly exogenous variation in the timing of technology availability. Our IV estimates show that irrigation induces persistent structural transformation: one additional irrigation point in the municipality increases average agricultural productivity annually by 836 euros per worker and shifts production toward larger, more crop-oriented farms. Complementary difference-in-differences estimates reveal substantial consolidation dynamics: twenty years after first adoption, the share of large farms rises by 5 percentage points, while the share of small farms falls by the same extent. Overall, the findings emphasize that reallocating access to a scarce natural resource can permanently reshape local agricultural productivity and the organization of production.

Conservation practices are central to sustainable agriculture, yet adoption remains limited because farmers face pervasive transition costs and uncertainties that often require policy support to overcome. Beyond standard socioeconomic constraints, behavioral traits may be pivotal in shaping adoption decisions especially when conservation practices generate uncertain short-run yield effects but delayed benefits through improved yield stability. However, evidence is scarce on how loss aversion, risk preferences, and time preferences jointly relate to conservation practice adoption and its yield dynamics. We develop a behavioral adoption framework that links these traits to the mean-variance tradeoffs farmers face when switching from conventional to conservation systems, and we test its implications using detailed survey measures of farmer preferences combined with plot-level production data from China’s major maize regions. The results show that behavioral traits on the demand side are systematically associated with adoption, and their importance depends on heterogeneous agronomic conditions and socioeconomic characteristics. Risk and time preferences are the most consequential dimensions, while loss aversion matters primarily through the short-run yield losses. These findings imply that public interventions mitigating behavioral barriers related to risk and intertemporal tradeoffs should be tailored to local yield dynamics and farmer heterogeneity.

In 2015, the California Department of Pesticide Regulation implemented restrictions on certain pesticide products to reduce volatile organic compound (VOC) emissions in the San Joaquin Valley (SJV). Specifically, the new regulation prohibited the use of high-VOC products containing abamectin, chlorpyrifos, gibberellins, and oxyfluorfen during the peak ozone season, May to October, for seven major crops: alfalfa, almond, citrus, cotton, grape, pistachio, and walnut. This study evaluates the regulation’s effectiveness in reducing peak-season VOC emissions and its impacts on other environmental and health hazards, including air, water, soil, pollinator health, and human health. Using monthly and annual data on pesticide use from 2010 to 2018 and a regression discontinuity design, we analyze the crop-specific effects of the regulation. We find that total VOC emissions from all targeted crops decreased by 25.8% during the peak season, with notable variations across crops. While the regulation reduced VOC emissions during the peak season and year-round, we observe year-round increases in almost all other hazards post-implementation, suggesting that the regulation reduced VOC emissions at the expense of increasing other hazards for all crops. Additionally, the regulation proved costly, leading to a $296.83 million increase (in 2018 US dollars) in annual total pest management costs, equal to 1.7% of the annual average of total revenues from the targeted crops post-implementation.

Climate change and the resulting increase in crop failures pose significant production risks, particularly for smallholder farmers dependent on agriculture as their primary income source. Growing multiple crops is regarded as an important resilience strategy, enabling risk diversification through both inter-seasonal practices (e.g., double or triple cropping) and intra-seasonal approaches like mixed cropping. This study examines the role of mixed cropping - the simultaneous cultivation of multiple crops on a single plot - in production risk management among smallholder farms in semiarid and tropical regions in India. From a farm management perspective mixed cropping is expected to support higher farm incomes, improved dietary diversity, and lower production costs. However existing research exploring its farm productivity and production risk impacts is limited, inconclusive and predominantly based on agronomic experimental data. Here we investigate the effects of a cotton-pigeonpea traditional mixed cropping system on farm productivity and production risk. We employ beta regressions in conjunction with a non-linear instrumental variable framework to estimate input-conditional yield densities using plot-level primary survey data (Arora et al. 2021). The welfare implications of mixed cropping are measured using certainty equivalent - the expected income minus the cost of risk exposure. To estimate risk exposure, we use higher-order moments of the yield distributions, adapting the framework developed by Di Falco and Chavas (2009) to our multiple cropping context. We find that mixed cropping reduces the variability of returns and increases skewness, leading to an overall reduction in downside risk exposure despite its slight negative effect on mean returns. To our best knowledge, we provide the first piece of evidence on the role of mixed cropping cotton-legumes in managing farm-level production risk in India.

JEL Codes: Q12, O13, C14, C36

Keywords: Yield Density Estimation; Risk Influencing Inputs; Beta Regression; Irrigation and Nitrogen Application; Instrumental Variables