Threshold effects of globalization on economic growth: insights from Azerbaijan using regression and Fuzzy C-Means

Abstract

The object of this research is the relationship between globalization and economic growth in Azerbaijan, analyzed using regression analysis and Fuzzy C-Means (FCM) clustering. While globalization is often linked to economic expansion, its effects in resource-dependent economies remain unclear. This study examines whether increasing globalization, measured by the KOF Globalization Index and its sub-indices (economic, social, and political globalization), positively impacts GDP per capita growth.

Regression analysis reveals significant negative impacts of globalization on GDP per capita growth, with social globalization showing the strongest negative effect (coefficient: –10.93, p<0.01), followed by political (–9.55, p<0.01) and economic globalization (–5.96, p<0.05). Conversely, institutional quality, measured by the rule of law, significantly promotes growth (10.22, p<0.01). Fuzzy C-Means clustering further identifies clear nonlinear (threshold) patterns: moderate globalization levels (KOF Globalization Index≈49.36) correspond with the highest average GDP per capita growth (≈14.47), whereas lower (≈32.13) and higher (≈62.52) levels associate with significantly lower or negative growth (–10.91 % and 1.29 %, respectively). These findings indicate that excessive global integration can undermine economic stability in resource-dependent economies at the early stages of industrialization, such as Azerbaijan.

In practice, globalization should be approached strategically. Policymakers must strengthen institutions, enhance long-term investment policies, and prioritize industrial learning before promoting deeper integration. A balanced approach will maximize globalization’s benefits while minimizing risks.