Incremental and transformational climate change adaptation factors in agriculture worldwide: A comparative analysis using natural language processing.

Climate change is projected to adversely affect agriculture worldwide. This requires farmers to adapt incrementally already early in the twenty-first century, and to pursue transformational adaptation to endure future climate-induced damages. Many articles discuss the underlying mechanisms of farmers' adaptation to climate change using quantitative, qualitative, and mixed methods. However, only the former is typically included in quantitative metanalysis of empirical evidence on adaptation. This omits the vast body of knowledge from qualitative research. We address this gap by performing a comparative analysis of factors associated with farmers' climate change adaptation in both quantitative and qualitative literature using Natural Language Processing and generalized linear models. By retrieving publications from Scopus, we derive a database with metadata and associations from both quantitative and qualitative findings, focusing on climate change adaptation of farmers. We use the derived data as input for generalized linear models to analyze whether reported factors behind farmers' decisions differ by type of adaptation (incremental vs. transformational) and across different global regions. Our results show that factors related to adaptive capacity and access to information and technology are more likely to be associated with transformational adaptation than with incremental adaptation. Regarding world regions, access to finance/income and infrastructure are uneven, with farmers in high-income countries having an advantage, whereas farmers in low- and middle-income countries require these the most for effective adaptation to climate change.