Energy consumption and carbon dioxide emissions in Algeria with scale normalization and real oil price asymmetry tests

Algeria’s fossil-intensive energy system makes emissions reduction particularly difficult because domestic energy consumption, hydrocarbon revenues, and economic activity remain closely interconnected. This study investigates the stability of the long-run relationship between primary energy consumption and carbon dioxide emissions, considering aggregate scale, lag uncertainty, structural instability, and nonlinear oil-price transmission. Annual data for 1990–2024 are analyzed using the autoregressive distributed lag bounds-testing approach. Per-capita and intensity models, alternative lag-selection criteria, structural-stability diagnostics, nonlinear ARDL decomposition, dynamic ordinary least squares, and Toda–Yamamoto tests provide complementary robustness evidence. The baseline model supports a long-run level relationship at the 5% significance level. Primary energy consumption has a positive long-run elasticity of approximately 0.80. The estimate declines to about 0.68–0.71 when emissions and energy consumption are normalized by population or real output, but remains positive and highly significant across ARDL, NARDL, and DOLS specifications. By contrast, the independent coefficients on real GDP per capita and real oil prices are more sensitive to the selected outcome, lag structure, and estimator. Wald tests do not reject either short- or long-run oil-price symmetry, while evidence on predictive direction varies with lag choice. The findings identify domestic energy consumption as the most persistent correlate of Algeria’s emissions, whereas the direct oil-price channel is weak and estimator-dependent. The findings are therefore consistent with policies aimed at lowering fossil-energy intensity through energy efficiency, gradual pricing reform, renewable-electricity deployment, grid investment, and the productive use of hydrocarbon revenues. The estimates are interpreted as long-run associations rather than structural causal effects.