Impact of current and historical climate shocks on crop diversification in Zambia: Insights from household- and district-level observations

Crop diversification is a farming practice for risk management prevalent in smallholder agriculture, offering adaptive benefits against challenges like climate change, price fluctuations, and crop disease. Despite its importance, there is a lack of comprehensive understanding of the relationship of crop diversification and current and historical climate shock. Our study seeks to bridge this gap through statistical analysis of household- and district-level data in Zambia. Specifically, we use the Pooled Fractional Probit (PFP) estimator to develop regression models for crop diversification, analyzing 6625 households for 3 years and 74 districts for 9 years, using Rural Agriculture Living Survey (RALS) and Crop Forecast Survey (CFS) datasets, respectively. Simpson's Diversity Index (SDI) of crops serves as the dependent variable and is consistently higher at the district level than at the household level, suggesting that aggregation at larger scales may mask localized monoculture vulnerabilities. Our findings reveal that both current and historical climate shocks significantly influence crop diversification decisions at both the household and district levels in Zambia. Heat stress and rainfall deficits during the planting season promote crop diversification, but their effects vary due to the diverse agroecological conditions and crop characteristics in different areas. Historical climate shocks prompt farmers to diversify as a long-term resilience strategy. This study emphasizes the complex, scale-dependent drivers of crop diversification in response to climate shocks, providing valuable insights for policy development in climate-resilient agricultural strategies.