🇿🇦 South African Economic Reform Agenda (SA-ERA)

A data-driven analysis of 10 years of South African economic policy recommendations (2015-2025).

📖 Project Overview

This project digitizes and analyzes a decade of Budgetary Review and Recommendation Reports (BRRR) to track government performance, fiscal compliance, and policy implementation. By processing 5,256 specific recommendations across 50 parliamentary reports, we identify actionable "Quick Wins" and structural reforms needed to unlock economic growth.

• Analysis Date: November 24, 2025
• Scope: 6 Priority Sectors (Energy, Labour, Finance, SciTech, Infrastructure, Trade)
• Dataset: 50 Full-Text Parliamentary Reports

---

⚡ Executive Summary: Key Findings

We identified 5,256 recommendations and scored them using an algorithmic framework based on Impact, Feasibility, and Cost.

Category	Count	Description
🚀 Quick Wins	441	High-impact, low-cost actions implementable immediately (<6 months).
🔥 High Priority	2,060	Critical reforms scoring highly across all dimensions.
🏛️ Institutional	1,473	Deep systemic changes requiring legislative or major structural reform.

🔍 Top Recurring Themes (10-Year Trend)

1. Compliance & Reporting (857)
2. Budget Execution (616)
3. Service Delivery (559)
4. Unemployment (400)
5. Procurement/Supply Chain (248)

---

📂 Deliverables & Repository Structure

1. Policy Documents

• 📄 Executive Summary – Start here for top 20 priorities.
• 📘 Full Policy Memo – Comprehensive 37-page analysis including implementation roadmaps.

2. Data Files

• analysis/recommendations_prioritized.xlsx – Master Dataset. Includes ROI scores and filtering flags (is_quick_win).
• analysis/recommendations.json – Machine-readable format for developers.
• analysis/report_summaries.xlsx – Metadata for all 50 analyzed reports.

3. Source Reports

• brrr_reports/ – PDF archive organized by sector (Energy, Finance, Labour, etc.).

4. Streamlit app

• Live

---

📊 Methodology

Our analysis pipeline uses Python (PyMuPDF) to extract text and a custom scoring algorithm to prioritize recommendations.

The ROI Scoring Framework

Each recommendation is scored on three dimensions to calculate a final ROI Score (1-10):

1. Feasibility (1-5): From "Major Investment" (1) to "Administrative Action" (5).
2. Impact (1-5): Based on sector importance and economic significance (e.g., Energy > Admin).
3. Cost (1-5): From ">R1bn" (1) to "<R1m" (5).

Formula: $ROI = \frac{Impact \times Feasibility}{Cost}$ (Normalized)

---

🚀 Usage Guide

For Researchers & Developers

Reproduce the analysis by running the pipeline scripts in order:

# 1. Install Dependencies
pip install -r requirements.txt

# 2. Download latest BRRR reports from PMG
python scripts/download_brrr_direct.py

# 3. Extract text and identify recommendations
python scripts/analyze_brrr_reports.py

# 4. Run the scoring and prioritization algorithm
python scripts/prioritize_recommendations.py

# 5. Generate the Policy Memo Markdown files
python scripts/generate_policy_memo.py