Decarbonisation pathways and traps in central Asian agriculture: typology, drivers and evidence from Kazakhstan

Purpose. This study aims to identify the presence and key characteristics of decarbonisation trajectories in the agricultural sectors of Central Asian countries. Using Kazakhstan as a case study, it further examines the main drivers and constraints shaping these trajectories within the broader context of the green transition. Methodology. The paper develops and applies an original three-stage quantitative framework. First, Tapio decoupling analysis is conducted for four Central Asian countries. Second, additive LMDI decomposition is employed to disentangle the contributions of scale, structural and intensity effects. Third, ARDL modelling is applied to the case of Kazakhstan to estimate long-run elasticities with respect to key drivers. The analysis draws on harmonised data from FAOSTAT, the World Bank and national statistical sources covering the period 2003–2023. This approach enables the identification of decoupling states, the measurement of factor contributions, and the estimation of long-run relationships. Results. The findings reveal two distinct development patterns across the region. Uzbekistan and Tajikistan exhibit an emissions-intensive growth trap, characterised by persistent coupling between agricultural value added and greenhouse gas emissions. In contrast, Kazakhstan and Kyrgyzstan display features of a volatility trap, marked by unstable decoupling states in which episodes of strong decoupling largely coincide with periods of economic contraction. For Kazakhstan, the LMDI and ARDL results indicate that scale effects constitute the primary source of emissions growth. Moreover, higher investment is statistically associated with higher, rather than lower, emissions over the sample period, pointing to an apparent “investment paradox” that should be interpreted cautiously given the aggregated nature of the investment data and the limited econometric design. Originality. The study contributes to the literature in two principal ways: (1) empirically, it addresses Central Asia, a region for which systematic evidence on agricultural decarbonisation and decoupling remains limited; (2) methodologically, it proposes an integrated framework combining regional decoupling analysis with LMDI decomposition and ARDL-based estimation of long-run elasticities for Kazakhstan. In contrast to much of the existing literature, agricultural emissions are modelled not solely as a function of output levels, but also as outcomes influenced by investment, price and climate-related drivers, allowing their respective contributions to be quantified. The paper further advances a conceptual typology of agricultural decarbonisation traps in transition economies, distinguishing between emissions-intensive growth traps and volatility traps and linking these categories to the observed trajectories in Central Asia. Practical implications. The results underscore the importance of country-specific decarbonisation strategies that extend beyond uniform mitigation targets. In the case of emissions-intensive growth traps, policy should prioritise improvements in resource use and technological efficiency. For volatility traps, emphasis should be placed on stabilising growth dynamics, redirecting investment towards genuinely low-carbon technologies and aligning agricultural support mechanisms with national climate objectives. The proposed framework offers policymakers a diagnostic tool for assessing and monitoring agricultural decarbonisation pathways in other transition economies.