The Relationship Between Energy Intensity and Sectoral Transformation in Türkiye: A Long-Run Analysis Based on the Services Sector Share

Abstract

This study investigates the long-run dynamics of energy intensity in Türkiye over the period 2000–2022, focusing on per capita income and the share of the services sector in total value added as a proxy for the sectoral composition of the economy. Energy intensity, defined as the amount of energy consumed per unit of economic output, is widely used as an indicator of energy efficiency and sustainable growth. Given that economies exposed to crises and macroeconomic shocks may not follow a single stable regime, the study tests for potential structural breaks in the energy intensity series. The Bai–Perron multiple structural break test identifies breakpoints in 2001, 2008, 2018, and 2020, corresponding to major macroeconomic disruptions including the post-2001 restructuring period, the global financial crisis, exchange rate shocks, and the COVID-19 pandemic. The stationarity properties of the variables are examined using the Zivot–Andrews unit root test. The results show that energy intensity and the services sector share are stationary at levels under structural breaks, whereas per capita income is not stationary in levels, making the ARDL bounds testing approach appropriate. The ARDL bounds test confirms a long-run cointegration relationship between energy intensity, per capita income, and the services sector share. Although the long-run coefficients carry the expected negative signs, they are not statistically significant. However, the error correction model indicates a strong adjustment toward long-run equilibrium, and CUSUM and CUSUMSQ tests confirm parameter stability. The findings suggest that energy efficiency policies should consider crisis episodes and cost shocks and be supported by efficiency-oriented structural transformation.