What shapes climate change perceptions in Africa? A random forest approach

TL;DR

This study uses a random forest approach on Afrobarometer data to identify key factors influencing climate change perceptions across Africa, highlighting the roles of agriculture, education, and trust in institutions.

Contribution

It introduces a novel application of random forest methodology to analyze diverse factors shaping climate perceptions in Africa, integrating survey and climatic data.

Findings

Perceived agriculture conditions strongly influence climate change perceptions.

Country-level factors and local weather changes are key predictors.

Education, information access, poverty, and trust significantly shape perceptions.

Abstract

Climate change perceptions are fundamental for adaptation and environmental policy support. Although Africa is one of the most vulnerable regions to climate change, little research has focused on how climate change is perceived in the continent. Using random forest methodology, we analyse Afrobarometer data (N = 45,732), joint with climatic data, to explore what shapes climate change perceptions in Africa. We include 5 different dimensions of climate change perceptions: awareness, belief in its human cause, risk perception, need to stop it and self-efficacy. Results indicate that perceived agriculture conditions are crucial for perceiving climate change. Country-level factors and long-term changes in local weather conditions are among the most important predictors. Moreover, education level, access to information, poverty, authoritarian values, and trust in institutions shape individual climate change perceptions. Demographic effects -- including religion -- seem negligible. These findings suggest policymakers and environmental communicators how to frame climate change in Africa to raise awareness, gather public support and induce adaptation.