Comparative Analysis of Energy Poverty and Health Equity Across OECD Countries Using Clustering Analysis (2020–2023)

Creative Commons License

Kalaycı Oflaz N., Yiğit P., Özen O.

7. International Applied Statistics Congress, İstanbul, Türkiye, 11 - 13 Mayıs 2026, ss.143, (Özet Bildiri)

  • Yayın Türü: Bildiri / Özet Bildiri
  • Basıldığı Şehir: İstanbul
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.143
  • İstanbul Medipol Üniversitesi Adresli: Evet

Özet

Energy poverty is defined as a household's inability to access energy services necessary to maintain basic living standards, or having to spend a disproportionate portion of its income on these services. The World Health Organization (WHO) states that energy poverty is not only an economic indicator but also a key predictor of health and an indicator of social justice. This study aims to compare the dynamic relationship between energy poverty and basic health indicators in OECD countries during the 2020-2023 period. Methodologically, the study applying Panel Data Clustering, which can simultaneously evaluate time and cross-sectional dimensions, and Grey Panel Data Clustering, which is effective in managing uncertainties. The findings were analyzed comparatively using artificial neural network-based Kohonen-SOM and traditional K-Means clustering algorithms. The empirical dataset was curated from the World Bank Open Data. All statistical computations, modeling, and visualization procedures were executed using R software (version 4.4.0) within the RStudio environment. The research results revealed that panel data-specific clustering methods performed better than other methods in capturing temporal variations and dynamic interactions between variables, and yielded more consistent results. Panel data clustering provides superior results for investigating health inequalities; its ability to model the data suggest policy makers with a better information than other clustering models.