Introduction: Research Problem
Tradycyjne podziały geopolityczne, takie jak „Zachód” czy „BRICS”, są często używane jako uproszczone ramy analizy gospodarczej. Rodzi to pytanie, czy te historyczne kategorie adekwatnie odzwierciedlają współczesne, wielowymiarowe profile rozwojowe państw.
The aim of this analysis was to empirically verify the homogeneity of these blocs. To this end, I conducted a multivariate cluster analysis of 31 key economies (EU-27 plus the US, China, India, and Brazil) to identify their actual similarity structure.
Research Methodology
To ensure objectivity, the analysis was based on a quantitative procedure.
Selection of Diagnostic Variables
Three key variables were selected, representing three different aspects of the economic profile:
- GDP per capita (Nominal, USD): An indicator of the current level of wealth.
- Global Innovation Index (Ranking position): An indicator of potential, human capital and the ability to generate value in the future.
- CO2 emissions (tons per capita): An indicator of the structure of the economy and its environmental cost.
Taxonomic Method Used
Used Wrocław Taxonomy (also known as the Wrocław dendrite method or minimum spanning tree), which is one of the agglomerative cluster analysis methods.
The analytical procedure included three stages:
- Data Standardization: In order to ensure comparability of variables expressed in different units (USD, item, tons), the data were subjected to unit standardization (z-score).
- Distance Matrix Calculation: Based on standardized data, the following was calculated: Euclidean distance matrix (D) for all 465 country pairs. This distance represents a synthetic measure of the "dissimilarity" of an object in three-dimensional feature space.
- Dendrite Structure: A minimum spanning tree was constructed, which graphically represents the similarity structure of objects, combining them into one coherent graph while minimizing the sum of distances (edges).
Analysis Results and Discussion
The taxonomic analysis revealed a cluster structure that significantly differs from traditional geopolitical divisions.
Identifying the High-Carbon Innovators Cluster: Estonia and China
The most striking result is the identification of an extremely strong similarity (d=0.71) between Estonia and China. Despite fundamental differences in scale and nominal GDP, the model identified them as "statistical twins."
This similarity is due to almost identical profiles in dimensions. innovation (rankings 11th and 16th) and CO2 emissions per capita (8.0 t vs 8.4 t). This model can be defined as the "High-Carbon Innovator Archetype", which is structurally different from the rest of Europe.
Relative proximity of the Chinese and European (industrial) models
The analysis showed that Germany's economic profile is more than twice as similar to that of China (d=1.47) as to that of the US (d=3.23).
The US model (Extreme GDP, High Innovation, Extreme Emissions) is unique. However, the Chinese model shares key structural features with EU industrial economies (Germany, the Czech Republic, Belgium) (high innovation with high emissions), making them statistically closer to each other than to the US model.
Heterogeneity of the BRICS bloc
The study questions the economic homogeneity of the BRICS bloc. China (in a cluster with Estonia) represents a different development model than Brazil and India. The latter form a common, strong cluster D(India, Brazil) = 0.68), which is in turn closely linked to Southeastern European countries (e.g., Romania, D(Romania, Brazil) = 0.66).
The data indicates that China is an innovation-driven economy, while Brazil and India exhibit a Low/Medium GDP, Low Innovation, Low Emissions profile, similar to the “Global South” model.
Outlier Identification: USA and Luxembourg
The largest distances in the matrix are between the USA and Luxembourg, classifying them as extreme outliers. Both countries form a separate, two-element cluster (D(USA,Luxembourg) = 2.81) based on the Extremely High GDP profile. This cluster connects to the main "trunk" of the dendrite last (by the longest edge), demonstrating its extreme statistical dissimilarity from the other 29 economies studied.
Conclusions
Multidimensional taxonomic analysis provides a picture of global economic interdependence that significantly departs from traditional geopolitical divisions. Diagnostic variables such as innovation and emissions structure have been shown to create new, non-obvious axes of similarity.
The identified clusters (e.g. “High-Emission Innovators”, “Industrial Heart of the EU”, “Global South” or “Extreme Wealth”) seem to better describe actual development patterns than the “East-West” categories.
#Taxonomy #DataAnalysis #DataScience #Economics #Innovations #FocusAnalysis #MultidimensionalMethods #WorldEconomy
