Programmatic Similarity and Cluster Analysis of Political Parties (Costa Rica, elections 2026–2030) Using Jaccard Distance and UPGMA

Abstract

Introduction: The comparative analysis of government programs makes it possible to identify patterns of thematic convergence and divergence among political parties, providing relevant evidence to understand their programmatic orientation within the Costa Rican electoral context. Objective: To analyze programmatic similarity among the twenty political parties participating in the 2026–2030 elections by applying distance metrics and hierarchical clustering techniques to their government programs. Method: A review of all available government programs for 2026–2030 was conducted, and a binary matrix of 58 themes was constructed. Similarity between parties was assessed using the Jaccard distance and the Unweighted Pair Group Method with Arithmetic Mean. Statistical software was used to generate dendrograms and identify consistent programmatic clusters. Results: The analysis identified four main clusters: one with minimal agendas/low programmatic coverage; a second with a reduced agenda/limited thematic emphasis; a third generalist cluster subdivided into three subgroups—one labeled Social State/Progressive Approach, another focused on State Reform/Institutional Modernization, and a third characterized by Order–Development/Pragmatic–Productive Approach; and a fourth containing parties with focused thematic coverage (niche parties). Conclusion: The results reveal that although parties converge on certain priority issues, their programmatic architecture displays substantial variation. The study contributes to political analysis by offering a reproducible and transparent methodology. It is recommended that future studies incorporate historical versions of party platforms, automated semantic analyses, and longitudinal comparisons to identify trends in programmatic evolution within the Costa Rican political system.