Governance Project

Builds a country-year panel combining World Bank governance indicators and GDP growth data, then supports SQL-based analysis of how institutional quality and economic performance evolve over time.

---

Overview

This project processes raw World Bank datasets into cleaned, analysis-ready outputs and constructs a panel dataset for downstream SQL queries. The workflow integrates governance indicators with GDP growth at the country-year level, enabling structured analysis of relationships between institutional quality and economic outcomes.

---

Data Pipeline

Raw Data (World Bank)
    │
    ├─ GovernanceData.xlsx
    └─ GDPGrowth.csv
    │
    ▼
Python Processing (src/)
    ├─ process_governance.py
    └─ process_gdp.py
    │
    ▼
Cleaned Outputs (data/processed/)
    ├─ GovernanceDataCleaned.csv
    └─ GDPDataCleaned.csv
    │
    ▼
SQL Transformation (sql/)
    ├─ Countries table
    ├─ Governance wide table
    └─ Final panel (country-year)
    │
    ▼
Analytical Queries (Q1–Q12)

---

Data

This project uses publicly available World Bank datasets:

• Governance Indicators (WGI): https://www.worldbank.org/content/dam/sites/govindicators/doc/wgidataset_with_sourcedata-2025.xlsx (2025 edition, accessed April 22, 2026)
• GDP Growth Data (World Bank): https://api.worldbank.org/v2/en/indicator/NY.GDP.PCAP.KD.ZG?downloadformat=excel (April 8, 2026 edition, accessed April 22, 2026)

Governance includes six dimensions:

• Voice & Accountability (VA)
• Political Stability (PV)
• Government Effectiveness (GE)
• Regulatory Quality (RQ)
• Rule of Law (RL)
• Control of Corruption (CC)

Raw data is not included in the repository.

Setup

Download the datasets and place them in:

data/raw/

Rename files:

• wgidataset_with_sourcedata-2025.xlsx → GovernanceData.xlsx
• World Bank GDP file → GDPGrowth.csv (CSV format required)

---

Installation

Create and activate a virtual environment, then install dependencies:

python -m pip install -e .

or

python -m pip install -r requirements.txt

---

Running the Pipeline

Process raw data into cleaned outputs:

python -m governance_project.main

Outputs are written to:

data/processed/

---

Database Setup (PostgreSQL)

Create a database and connect:

CREATE DATABASE governance_db;

Run SQL files in order:

1. 01_load_data.sql
2. 02_create_countries_table.sql
3. 03_create_gov_wide.sql
4. 04_create_final_panel.sql
5. 05_checks.sql

To run SQL files, use command: psql -d governance_db -f sql/<file_name>.sql

This produces a country-year panel combining governance indicators and GDP growth.

---

Query Descriptions

• Q1: Calculates 10-year average governance scores by country to identify long-term institutional levels.
• Q2: Measures 5- and 10-year changes in governance to track medium- and long-term institutional shifts.
• Q3: Identifies countries with the largest year-over-year improvements in governance.
• Q4: Identifies countries with the largest year-over-year declines in governance.
• Q5: Computes average GDP growth over 10-year periods to capture long-run economic performance.
• Q6: Calculates 5- and 10-year average GDP growth to compare short- and medium-term trends.
• Q7: Identifies countries with the largest year-over-year increases in GDP growth.
• Q8: Identifies countries with the largest year-over-year declines in GDP growth.
• Q9: Measures the correlation between governance indicators and GDP growth across countries and years.
• Q10: Examines the relationship between governance quality and GDP growth volatility.
• Q11: Compares GDP growth before and after the 2008 financial crisis to assess overall impact.
• Q12: Classifies countries based on whether governance and GDP growth move together or diverge over five-year periods.

---

Key Findings

• The relationship between governance and GDP growth is weak in aggregate.
• Higher governance quality is associated with lower growth volatility.
• Average GDP growth declined after the 2008 financial crisis (approximately -0.19 percentage points).
• Governance and growth do not consistently move together over medium-term horizons.

---

Data Validation

Basic validation is applied during processing:

• ISO3 country codes are standardized and checked for format
• Years are constrained to valid ranges
• Governance estimates are validated for numeric bounds

Invalid or malformed rows are flagged during processing.

---

Project Structure

repo-root/
├─ README.md
├─ pyproject.toml
├─ requirements.txt
├─ .gitignore
├─ src/
│  └─ governance_project/
│     ├─ __init__.py
│     ├─ main.py
│     ├─ process_gdp.py
│     └─ process_governance.py
├─ tests/
│  └─ test.py
├─ sql/
│  ├─ 01_load_data.sql
│  ├─ 02_create_countries_table.sql
│  ├─ 03_create_gov_wide.sql
│  ├─ 04_create_final_panel.sql
│  ├─ 05_checks.sql
│  ├─ Q1_10_year_gov_avgs.sql
│  ├─ Q2_5_and_10_year_gov_changes.sql
│  ├─ Q3_most_improved_by_year.sql
│  ├─ Q4_least_improved_by_year.sql
│  ├─ Q5_10_year_gdp_growth_avg.sql
│  ├─ Q6_5_and_10_year_gdp_avgs.sql
│  ├─ Q7_most_improved_gdp_by_year.sql
│  ├─ Q8_least_improved_gdp_by_year.sql
│  └─ additional SQL analysis files
└─ data/
   ├─ raw/
   └─ processed/

---

Limitations

• Missing data varies by country and year, leading to an incomplete panel.
• Results are based on correlations and descriptive comparisons, not causal inference.
• Governance effects may operate with time lags not captured in simple comparisons.

---

Testing

Run the included test script:

python tests/test.py

This verifies governance data processing using repository-relative paths.