Assessing the effects of Land Registration on Climate-Smart Agriculture and Soil Investment Decisions: Machine learning informed lessons from Malawi’s Chikwawa and Nkhotakota Districts

Abstract

Secure land tenure is widely recognized as a key driver of sustainable agricultural development. Yet, empirical evidence on how tenure security, particularly land registration, shapes farmers’ soil investment decisions remain limited. This study applies Multivariate Bernoulli Mixture Models and machine learning approaches to investigate whether formal land registration influences patterns of climate-smart agriculture adoption and soil investment behaviour, while also identifying the socioeconomi