Dimensionality reduction analysis of the renewable energy sector in Azerbaijan: nonparametric analyses of large datasets

Abstract

Although the number of econometric analyses related to the renewable energy sector in Azerbaijan is increasing, studies on nonparametric dimensionality reduction are rather sparse. Principal component analysis (PCA) and multiple correspondence analysis (MCA) were chosen to fill this apparent research gap. As a result, a large dataset including the renewable energy sector and selected key macroeconomic indicators was evaluated. The PCA procedure yielded four distinct principle components reflecting the main macroeconomic variables, renewable energy production, industry-energy relations and natural resource revenues. The PCA method offers the possibility to examine the precise correlations and the underlying patterns between the displayed clusters of variables. Meanwhile, the MCA-based cross-country assessment of Azerbaijan’s wind, solar and hydropower has struck somewhat pessimistic notes, as the country lags behind neighbouring and other post-Soviet countries (e.g. Estonia, Iran, Latvia, Russia) in developing its green energy sector. These findings are of great interest to policymakers, businesses and academics who wish to gain deep insight into the Azerbaijani economy in terms of renewable energy production. The practical value of the present work also lies in the fact that it analyses a multidimensional and relatively longitudinal dataset (1990–2022), which is an example of a methodological application of two nonparametric approaches.