Abstract
Political instability can have important consequences for environmental outcomes, particularly in economies that depend heavily on natural resource extraction. This study investigates whether the 2011 Arab Spring is associated with a detectable change in carbon dioxide (CO₂) emissions in Libya using an interrupted time series (ITS) analysis of annual CO₂ emissions per capita, 1980–2022 (n = 43), with 2011 as the intervention point. A baseline segmented regression finds no immediate level change in emissions in 2011 and only a marginally significant change in trend, one that is sensitive to the choice of standard errors. An extensive battery of robustness and falsification checks—a multibreak specification, exclusion of the 2020 outlier year, alternative breakpoints, placebo tests, and a candidate-break sensitivity scan—shows that this result is not robust: it loses significance once 2020 is excluded, it is smaller than the trend change implied by an alternative 2014 breakpoint, and it is statistically unremarkable relative to breaks placed elsewhere in the series. A separate auxiliary regression using oil production as the outcome shows a large, robust level shock at 2011, confirming that the 2011 revolution produced an abrupt collapse in oil output, but this does not translate into an equally robust, uniquely 2011-dated change in per capita CO₂ emissions. The multibreak model instead identifies 2020 as the most robust structural break in the emissions series, consistent with the COVID-19 demand shock and the 2020 oil-terminal blockade. A supplementary environmental Kuznets curve specification finds no significant relationship between gross domestic product (GDP) per capita and CO₂ emissions over the same period, indicating that this specification of income-based explanations does not account for Libya’s emissions trajectory. The article’s central conclusion is accordingly narrower and more precisely scoped than a single-break design would suggest: Libya’s oil sector shows a clear, well-identified disruption in 2011, but the emissions series itself does not show a robust, specifically 2011-dated structural break once standard falsification checks are applied—the most defensible structural break in per-capita emissions is 2020. We discuss the implications of this for using single-country, single-break ITS designs to study politically volatile, resource-dependent economies and outline extensions (disaggregated conflict episodes, comparative difference-in-differences designs, and direct energy-mix covariates) that could sharpen causal identification in future work.