Multi-kink quantile regression based analysis for industrial structure of non-high-income economies

This paper examines the non-linear relationship between industrial structure and GDP per capita in non-high-income countries using a Multi-Kink Quantile Regression (MKQR) framework. Building on classical growth theory and applying the empirical methodology proposed by Zhong et al. (2022), we use data from 125 countries for the years 2002 and 2023 to identify income thresholds at which structural shifts occur in the contribution of industry to economic output. The findings reveal a consistent three-phase pattern across quantiles: an initial stage of rapid industrial expansion, followed by a period of slower growth, and ultimately a phase of deindustrialization as higher income levels are reached. Notably, the estimated turning points vary according to the level of industrial development, with less industrialized countries reaching the peak of industrialization at significantly lower levels of GDP per capita. These results underscore both the heterogeneity of the industrialization process and the increasing challenges faced by developing economies in sustaining industrial growth.