income-inequality-colombia

A comprehensive analysis of income inequality in Colombia using the 2024 GEIH dataset. This repository includes data processing, descriptive statistics, visualizations, and policy-oriented insights into socioeconomic disparities across regions and social classes.

---

Overview

This project explores the distribution of household incomes in Colombia, identifies the main socioeconomic variables distinguishing high- and low-income households, and uncovers regional patterns of inequality. We leverage the GEIH (Gran Encuesta Integrada de Hogares) 2024 microdata to:

• Compute central-tendency and dispersion metrics
• Visualize skewness, outliers, and class-based differences
• Map geographic disparities
• Assess labor, education, housing, and demographic factors
• Formulate data-driven policy recommendations

---

Data Description

• Source: DANE’s GEIH microdata (2024)
• Coverage: Non-institutional civilian households across all Colombian departments (except Providencia & San Andrés)
• Key Variables:
  • Ingresos_finales: Total monthly household income (COP)
  • Demographics: household size, education level, number of children
  • Labor: contract type, formal employment status, tenure, pension contributions
  • Housing: ownership status, rental or mortgage payments
  • Geography: department codes, region classifications

File Structure

.
├── GEIH Analisis.do   # Main Python script with all the project and conclutions
├── DATA               # The necesary DataFrames to run the project local
└── README.md          # Project documentation

Requirements

• Python: 3.8 or higher
• Environment: Jupyter Notebook or JupyterLab
• Dependencies:
  • pandas
  • numpy
  • matplotlib
  • seaborn
  • scipy
  • geopandas
  • folium

Analysis Workflow

1. Descriptive Statistics: Compute mean, median, mode, SD, range, skewness, kurtosis; plot histograms and boxplots (linear & log scales).
2. Outlier Detection: Identify extremes via IQR criterion and z-score thresholding.
3. Class Definitions: Map DANE class proportions (60% low, 35% middle, 5% high) to household income ranges.
4. Socioeconomic Profiling: Compare education, contract type, formalization, tenure, and job demands across classes.
5. Geographic Patterns: Generate department-level heatmaps and regional choropleths for income and education.
6. Policy Insights: Synthesize findings into targeted recommendations.

Key Findings

• Highly right-skewed distribution: Mean (2.38 M COP) ≫ median (1.30 M COP); 75% earn < 2.3 M COP.
• Housing burden: Rent/mortgage consumes 30–37% of income for most households.
• Educational divide: Upper class median at university level vs. basic secondary for low-income.
• Labor formalization gap: 94% of upper class under formal contracts vs. 21–51% in low-income groups.
• Regional inequality: Andean region leads in income; Pacific and Caribbean lag despite high education levels.
• Structural poverty: Majority live below living‐wage thresholds, creating a persistent poverty cycle.

Policy Recommendations

1. Invest in education: Expand technical and tertiary programs to boost access to formal, higher-paying jobs.
2. Promote labor formalization: Offer incentives and reforms for stable contracts and pension contributions, especially in lagging regions.
3. Support regional development: Improve infrastructure to lower costs of housing, healthcare, and food.
4. Combat corruption & strengthen institutions: Ensure equitable opportunities nationwide to break the cycle of structural poverty.

Contributing

Contributions are welcome! Please open issues or submit pull requests at
https://github.com/pablo-reyes8

License

This project is licensed under the Apache License 2.0.