Climate change, driven by greenhouse gas emissions, necessitates effective natural resource management. Natural resource rents, the economic gains from resource exploitation, are a significant source of carbon dioxide emissions, often leading to the 'carbon curse'. The COVID-19 pandemic and the Russia-Ukraine conflict have heightened economic and political risks, impacting global energy supplies and carbon emissions. Existing research highlights the influence of various risk factors on carbon emissions, but often examines them in isolation. This study investigates the impact of natural resource rents on carbon emissions across different emission levels (quantiles), considering the moderating roles of economic, financial, and political risks. It also explores threshold effects, where the relationship between natural resource rents and emissions shifts depending on the level of risk. Using data from 66 countries over three decades (1990-2020), this research addresses the gap in understanding the complex interplay between natural resource rents, risk factors, and carbon emissions, providing valuable insights for policy formulation.

The literature extensively documents the link between natural resource rents and increased carbon emissions, often termed the 'carbon curse'. Studies show a positive correlation between resource rents and emissions in various contexts, including Saudi Arabia, the United States, and G7 countries. However, some research suggests potential mitigation through clean energy and sustainable resource use. Regarding risk factors, economic uncertainty has been shown to increase carbon emissions in both the short and long term, potentially hindering green technology innovation. Financial risks also play a role, with stable financial markets supporting investments in green projects. Political risks are linked to carbon emissions, with political stability often correlating with more effective carbon reduction policies. However, the literature often examines these factors separately. This research integrates this prior work by investigating the simultaneous effects of natural resource rents and various risk factors on carbon emissions, providing a comprehensive analysis of their interplay.

This study uses panel data from 66 countries spanning 1990 to 2020. The dependent variable is per capita carbon emissions (InCO2), while the independent variable is natural resource rents (Inrent) as a percentage of GDP. Economic risk (Iner), financial risk (Infr), and political risk (Inpr) from the International Country Risk Guide (ICRG) act as moderating variables. Control variables include economic growth (Ingdp), urbanization (Inur), and industrialization (Inind). Before analysis, logarithmic transformations were applied to address potential heteroscedasticity. Stationarity of variables was checked using first and second-generation unit root tests (Fisher-PP, Fisher-ADF, and CIPS). The Pedroni cointegration test assessed long-term relationships among the variables. Quantile regression, using a nine-quantile approach (0.1 to 0.9), examined the impact of natural resource rents on carbon emissions across various emission levels and the moderating effects of the risk variables. Interaction terms were included in the quantile regression model to capture the moderating effects. Finally, a dynamic threshold regression model, using economic, financial, and political risks as threshold variables, was employed to identify potential shifts in the relationship between natural resource rents and carbon emissions at specific risk levels.

Quantile regression consistently showed a positive relationship between natural resource rents and carbon emissions across all quantiles (0.1 to 0.9), except for the 0.8 quantile. The effect was more pronounced at higher quantiles. Economic risk reduction generally decreased carbon emissions, particularly at higher quantiles. Financial risk reduction also showed a predominantly negative correlation with emissions across most quantiles. Conversely, political risk reduction tended to increase emissions across all quantiles. Economic growth consistently increased emissions. Urbanization showed a positive correlation at lower quantiles and a negative correlation at higher quantiles. The impact of industrialization varied across quantiles. Robustness checks using POLS, 2SLS, and GMM confirmed the positive relationship between natural resource rents and carbon emissions. Dynamic threshold regression revealed significant threshold effects for economic, financial, and political risks. Specifically, when economic and political risks were low (below certain threshold values), increased natural resource rents were associated with decreased carbon emissions.

The findings confirm that natural resource rents are a significant driver of carbon emissions. However, the moderating role of economic, financial, and political risks highlights the importance of effective risk management in mitigating the negative environmental impacts. Reduced economic and financial risks, by promoting innovation and green investments, can lessen the emission-increasing effects of natural resource rents, particularly at higher emission levels. Political stability, by enabling effective environmental policies, can significantly reduce the positive relationship between natural resource rents and emissions. The threshold effects highlight the importance of considering different risk levels when formulating policies. Effective policies would focus on improving economic and political stability and fostering robust financial systems to promote sustainable practices.

This study demonstrates a strong positive relationship between natural resource rents and carbon emissions, emphasizing the crucial role of risk management in mitigating environmental consequences. Effective governance and financial systems are key to reducing the negative impact of natural resource extraction. Future research should investigate the effects of different types of natural resources and their unique environmental impacts, allowing for more targeted policy recommendations.

This study relies on aggregate data, which may obscure variations within countries. Future research could benefit from disaggregated data to provide more nuanced insights. The analysis also assumes a linear relationship between some variables, while non-linear relationships may exist. Finally, the study focuses primarily on the direct effects of risks, and future studies could explore indirect effects and feedback loops.