Abstract
<title>Abstract</title> <p>Energy commodities sit at the centre of the transmission of global supply shocks to domestic price levels, yet the relative contribution of crude oil versus natural gas to short-run consumer price index (CPI) inflation across a broad and heterogeneous set of economies is still only partially mapped. This paper examines that question for eighteen G20 economies — Argentina, Brazil, Canada, China, France, Germany, India, Indonesia, Italy, Japan, Mexico, Russia, South Africa, South Korea, Turkey, the United Kingdom, the United States, and the European Union — using 72 months of data from June 2018 to May 2024. Country-by-country bivariate OLS regressions, estimated for the fourteen economies with complete monthly series, are complemented by Pearson correlation analysis across all eighteen units. Oil prices are a statistically significant, positive predictor of CPI inflation in twelve of the fourteen regression economies; natural gas prices are significant in a smaller and structurally distinct subset — Italy, Mexico, Argentina, and Turkey — with the Italian coefficient both the largest and most precisely estimated of the sample, consistent with Italy's heavy reliance on gas for heating and power generation. Correlation coefficients between oil prices and CPI inflation are uniformly positive, ranging from near zero in China and South Korea to moderately strong in the United States, Italy, and the United Kingdom; natural gas correlations are more dispersed and, in several emerging markets, negligible. Because the specification is bivariate and estimated in levels, without panel structure or macroeconomic controls, the results are best read as a systematic descriptive mapping of where oil-versus-gas co-movement with inflation is strongest, rather than as causally identified pass-through elasticities. The paper situates these patterns within the cost-push and expectations-augmented Phillips curve literature and closes with a concrete agenda — panel estimation, unit-root and cointegration testing, and asymmetric NARDL specifications — for the causal work that should follow. JEL Classification: E31, Q41, C13, C22, F43</p>