Abstract
This study investigates the dynamic interplay among agricultural growth, total energy consumption, renewable energy penetration, and CO₂ emissions in Kazakhstan, using annual data for 2001-2024. Augmented Dickey-Fuller diagnostics classify all series as I(1), and information criteria favor a parsimonious VAR(1). Variance decompositions indicate that agricultural growth is primarily driven by own shocks, with a growing medium-horizon contribution from CO₂. Total energy use remains largely self-propelled, accompanied by modest spillovers from agriculture and a gradual feedback from emissions. The renewable share increasingly reflects movements in total energy demand at longer horizons, while agriculture contributes a smaller but persistent portion; the direct role of CO₂ in renewable dynamics is negligible under orthogonalized innovations. Emissions variability is governed chiefly by shocks to aggregate energy demand and by own persistence. Pairwise Granger tests uncover a single predictive channel - total energy use Granger-causing the renewable share - whereas other pairs do not exhibit short-run directionality at conventional thresholds. Taken together, the evidence portrays a demand-led nexus in which aggregate energy conditions steer near-term adjustments, agricultural performance reflects internal dynamics and environmental pressure, and emissions co-move with energy demand. The findings underscore policy relevance for demand management, grid integration with bankable procurement of renewables, and climate-smart upgrades in irrigation and inputs.