Classification of Oil Consumption Patterns In 58 Countries by Hierarchical Cluster: Global Trends and Context of Costa Rica (2019–2023)

Abstract

The objective of this research, with a quantitative approach and exploratory scope, was to categorize 58 countries included in the 2024 Annual Statistical Report of the Organization of Petroleum Exporting Countries (OPEC) into homogeneous groups according to the evolution of their oil demand during the period 2019–2023, through the application of the hierarchical clustering model, with the purpose of identifying global patterns of energy consumption and providing reference elements for analysis of Costa Rica's energy policy. The methodology was developed in the statistical environment R, using multivariate analysis techniques based on Euclidean distance and the Ward's linkage method, from an optimal solution of three clusters. The results showed a high classification performance of the model, with an interclass inertia of 94,72 % and an intraclass inertia of 5,28 %, indicating an adequate separation between groups and a high internal cohesion. The analysis identified three differentiated profiles of oil demand: a group of economies with high consumption and relatively stable behavior, a group of countries with intermediate demand and greater sensitivity to economic cycles and a group characterized by low levels of consumption and high relative variability. The findings show that the hierarchical cluster constitutes a robust tool for the classification and characterization of international patterns of oil consumption, allowing the generation of relevant inputs for the formulation of public policies, energy planning and the prospective evaluation of demand scenarios in hydrocarbon-dependent countries such as Costa Rica.