A cross-country analysis of the role of service sector in the relationship between CO<SUB align="right">2 emissions and economic growth using machine learning techniques

Abstract

The study aims to explore the relationship between CO2 emissions per capita, service sector share in GDP and GDP per capita using decision tree, and multiple ridge and lasso regression techniques on cross-sectional data of 175 countries. GDP per capita is a better determinant of the CO2 emissions of a country than the share of services in GDP. The fit between emissions and income improves on account of service sector share in GDP. The study finds that an increase in service sector share in high income countries leads to decrease in emissions while in low income countries it leads to an increase in emissions. An N-shaped relationship is found between CO2 emissions and income across the countries. Service sector share acts as a moderator in this relationship.