Testing the Carbon Hysteresis Hypothesis in Azerbaijan: Evidence from Nonlinear Unit Root Tests

Abstract

This study investigates the validity of the Carbon Hysteresis Hypothesis (CHH) in Azerbaijan by employing advanced nonlinear time-series econometric techniques. Using annual CO₂ emissions data from 1901 to 2022, sourced from the Global Carbon Project and Our World in Data, the analysis begins with Harvey (2007) and Harvey et al. (2008) linearity pre-tests, which confirm the nonlinear nature of the emissions series. This finding justifies the application of nonlinear unit root tests—Kapetanios, Shin, and Snell (KSS, 2003), Sollis (2009), and Kruse (2011)—over conventional linear approaches such as ADF(1981), PP(1989), and KPSS(1992). The empirical results consistently fail to reject the null hypothesis of a unit root across both symmetric and asymmetric specifications, indicating that Azerbaijan’s CO₂ emissions follow a non-stationary process and that carbon hysteresis is present. This persistence suggests that shocks—whether from economic expansion, external crises, or policy interventions—have permanent effects on emissions levels. The study aligns with existing evidence from other oil-exporting economies and highlights the structural challenges of decarbonization in hydrocarbon-dependent contexts. Based on these findings, the paper proposes a set of policy recommendations aimed at long-term decarbonization, renewable energy integration, industrial modernization, and regulatory strengthening, contributing to both the empirical literature on carbon hysteresis and the design of sustainable climate policies.