De-industrialization in Azerbaijan's Textile Subsectors: Canonical Correlation Analysis and the Dutch Disease Hypothesis

Abstract

Since 2014, academic studies have increasingly underscored the potential adverse consequences associated with Azerbaijan's imbalanced and dependent economic structure. It is widely recognized that countries relying heavily on the export of primary commodities are ill-prepared for situations characterized by sharp declines in international commodity prices. Hence, the objective of this paper is to examine the Azerbaijani economy amidst two parallel developments: the growth of oil-related macroeconomic indicators and the contraction of non-oil subsectoral industrial production. To achieve this, the analytical framework of the Dutch disease, a widely preferred theory to study commodity exporters, and canonical correlation analysis (CCA) were employed in the period 1995 to 2021. The findings reveal statistically significant canonical correlations between certain subsectors of the textile industry (such as ginned cotton, cotton fabric, silk fabric, bed linen of cotton, and cotton yarn) and the Dutch disease variables (e.g., oil rent, real effective exchange rate), while other subsectors (including carpets, crocheted legwear and garments, outerwear, underwear, and footwear) do not exhibit similar patterns. These results show that non-consumer subsectors of the textile subsectors, especially the cotton sector, are more vulnerable to the effects of Dutch disease than consumer subsectors of the textile subsectors. In addition, the oil rent variable is a persistent channel that shows a negative correlation with the latent variables of the textile subsectors. These results prompt policymakers and researchers to reassess the role of large extractive industries in a small, open economy like Azerbaijan and to formulate economic policies that safeguard and foster specific subsectors.