Behavioral Drivers of Carbon Emissions: The Roles of Intention, Knowledge, and Capability

Carbon mitigation is central to the global response to climate change, yet the influence of individual behavioral intention remains insufficiently understood. Here we adopt a human agency perspective by developing an indicator framework integrating intention, knowledge, and capability. Using panel data from eight countries during 1995–2022, we apply machine learning and statistical approaches to identify dominant drivers of carbon emissions and assess their relative effects. Nonlinear, probabilistic and machine learning models are further employed to project emission trajectories under contrasting socio-economic scenarios. Four key variables jointly explain about 67% of national emissions. Economic growth and the share of renewable energy are the most influential factors, while environmental awareness and education exert growing importance. Across all countries, both scenarios show declining emissions, with the proactive pathway achieving earlier peaks, lower maxima, and faster reductions, alongside smaller predictive uncertainty. These results highlight the synergy between renewable energy expansion and enhanced public engagement, emphasizing that effective mitigation requires coordinated progress in policy, education, and institutional capacity.