electricsheepafrica/african-textile-garment-production

---
license: cc-by-4.0
tags:
- manufacturing
- textile
- garment
- agoa
- sub-saharan-africa
- synthetic
- fashion
pretty_name: African Textile & Garment Production
size_categories:
- 10K<n<100K
task_categories:
- tabular-regression
- tabular-classification
---

# African Textile & Garment Production Dataset

Synthetic panel dataset of textile and garment manufacturing across eight Sub-Saharan African countries.

## Overview

| Property | Value |
|---|---|
| Countries | Ethiopia, Kenya, Madagascar, Mauritius, South Africa, Nigeria, Lesotho, Tanzania |
| Years | 2015–2025 |
| Scenarios | baseline, agip_grow, global_downturn |
| Records per scenario | ~80,000 (10,000 per country) |
| Total records | ~240,000 |
| License | CC-BY-4.0 |

## Variables

| Variable | Type | Description |
|---|---|---|
| `record_id` | string | Unique UUID |
| `country` | string | Country name |
| `year` | int | Production year (2015–2025) |
| `factory_id` | string | Factory identifier |
| `product_type` | string | woven / knit / denim / garment |
| `production_volume_units` | int | Units produced |
| `capacity_utilization_pct` | float | Capacity utilization (0–100) |
| `export_orders_pct` | float | Share of orders for export (0–100) |
| `workforce_size` | int | Number of workers |
| `average_wage_usd` | float | Average monthly wage (USD) |
| `compliance_audit_score` | float | Social/compliance audit score (0–100) |
| `lead_time_days` | int | Order-to-delivery lead time |
| `defect_rate_pct` | float | Product defect rate |
| `export_destination` | string | US / EU / Regional |
| `investment_usd` | float | Annual capital investment (USD) |
| `local_content_pct` | float | Local raw-material content (0–100) |
| `scenario` | string | Scenario label |

## Scenarios

- **baseline** — Historical trend continuation.
- **agip_grow** — Africa Growth & Integration Programme: 8–20% capacity and export uplift per country, reflecting AGOA and AfCFTA policy expansion.
- **global_downturn** — External demand shock: 20–30% contraction in capacity utilization and exports.

## Key Design Notes

- Ethiopia Hawassa Industrial Park factories are modelled at ~60% mean capacity utilization, reflecting real-world ramp-up constraints.
- AGOA beneficiary country export shares to the US are elevated (40–60%) vs non-beneficiaries.
- Lesotho is weighted toward denim production.
- Mauritius has the highest wages and compliance scores, reflecting its mature EPZ ecosystem.
- Nigeria has the lowest capacity utilization and export share, reflecting infrastructure challenges.

## Files

```
data/
baseline.csv
agip_grow.csv
global_downturn.csv
all_scenarios.csv
generate_dataset.py
validate_dataset.py
requirements.txt
```

## Usage

```python
from datasets import load_dataset
ds = load_dataset("electricsheepafrica/african-textile-garment-production")
```

## Reproduction

```bash
pip install -r requirements.txt
python generate_dataset.py
python validate_dataset.py
```

## Citation

```bibtex
@dataset{african_textile_garment_production_2026,
title={African Textile & Garment Production Dataset},
author={Electric Sheep Africa},
year={2026},
license={cc-by-4.0}
}
```