electricsheepafrica/african-election-infrastructure

---
license: cc-by-4.0
pretty_name: African Election Infrastructure
tags:
- governance
- elections
- electoral-integrity
- sub-saharan-africa
- synthetic
- lmic
- democracy
size_categories:
- 10K<n<100K
configs:
- config_name: baseline
data_files: data/baseline.csv
- config_name: modernized_elections
data_files: data/modernized_elections.csv
- config_name: contested_elections
data_files: data/contested_elections.csv
---

# African Election Infrastructure Dataset

## Overview

Synthetic dataset modeling electoral infrastructure across **12 Sub-Saharan African (SSA) countries** under **3 scenarios**. Each scenario contains **10,008 records** (834 per country). Parameters are grounded in empirical literature on election technology, observer ratings, cost structures, and dispute resolution in SSA.

## Scenarios

| Scenario | Description |
|---|---|
| `baseline` | Current-state elections reflecting observed patterns in recent cycles (2017–2025) |
| `modernized_elections` | Hypothetical digital investment and institutional reform pathway |
| `contested_elections` | Disputed outcomes with technology failures and elevated disputes |

## Countries

Nigeria, Kenya, Ghana, Rwanda, Uganda, Tanzania, South Africa, Ethiopia, Senegal, DR Congo, Cameroon, Zambia

## Variables (22 columns)

| Variable | Type | Description |
|---|---|---|
| `country` | str | Country name |
| `scenario` | str | Scenario identifier |
| `population_millions` | float | Total population (millions) |
| `registered_voters_millions` | float | Registered voters (millions) |
| `polling_stations_total` | int | Total polling stations |
| `polling_stations_per_100k` | float | Polling stations per 100k registered voters |
| `voter_turnout_rate` | float | Voter turnout rate (%) |
| `biometric_verification_pct` | float | % of polling stations with biometric verification |
| `electronic_voting_pct` | float | % of votes cast via electronic means |
| `observer_rating_score` | float | International observer rating (1–10 scale) |
| `result_transmission_speed_hours` | float | Hours to transmit results from polling station to tally centre |
| `election_disputes_count` | int | Number of formal election disputes filed |
| `dispute_resolution_days` | float | Days to resolve election disputes |
| `election_cost_per_voter_usd` | float | Election administration cost per registered voter (USD) |
| `election_compliance_score` | float | Composite electoral compliance score (0–100) |
| `election_quality_class` | str | Quality classification: well-managed / partially_managable / poorly_managable |
| `n_political_parties` | int | Number of contesting political parties |
| `international_observer_missions` | int | Number of international observer missions deployed |
| `internet_penetration_pct` | float | Internet penetration at time of election (%) |
| `security_incidents` | int | Security incidents at or near polling stations |
| `voter_education_index` | float | Voter education coverage index (0–100) |
| `female_candidate_pct` | float | Female candidate share (%) |

## Literature-Informed Parameter Sources

Country-level priors are drawn from:

- **Biometric systems**: Nigeria BVAS (INEC, 2023), Kenya KIEMS (IEBC), Uganda BVVKs (AU COMESA-IGAD 2026), Rwanda (>98% biometric coverage)
- **Voter turnout**: Nigeria 2023 at ~35% (INEC), Rwanda consistently >90%, SSA average 50–70% (IDEA)
- **Cost per voter**: Ghana $2.13–2.19 (2024), Kenya $25.4 (2017), Rwanda $1.05, Nigeria $8.61, SSA average $11.30 (van der Straaten, 2018)
- **Electronic voting/transmission**: Nigeria IReV/e-transmission, SSA digitalisation trends (GIGA Focus 2025), Ghana BVRV systems
- **Observer ratings**: AU election observer missions, NDI/IRI missions, Commonwealth observers
- **Dispute resolution**: Kenya Supreme Court nullification (2017), Nigeria election tribunals, DR Congo contested outcomes

## Data Generation

```bash
# Install dependencies
pip install -r requirements.txt

# Generate all 3 scenarios (~10K records each)
python generate_dataset.py --n 834 --seed 42

# Run validation + diagnostic plots
python validate_dataset.py
```

## Diagnostic Plots

The validation script produces an 8-panel diagnostic covering:
- A: Mean voter turnout by country and scenario
- B: Biometric verification distribution
- C: Election cost per voter (box plots)
- D: Observer rating vs compliance score
- E: Disputes vs resolution time
- F: Election quality class distribution
- G: Result transmission speed
- H: E-voting vs internet penetration

## Limitations

- This is a **synthetic** dataset; it does not represent actual election results for any specific election cycle.
- Country priors reflect literature estimates and may not capture sub-national variation.
- Scenario modifiers are illustrative — real-world modernization or contestation dynamics involve complex institutional and political factors not fully captured here.
- Quality classifications are based on a simplified composite score and should not be used as authoritative assessments of any country's elections.

## Citation

If you use this dataset, please cite:

```
@dataset{african_election_infrastructure_2026,
title={African Election Infrastructure Dataset},
year={2026},
license={cc-by-4.0},
note={Synthetic dataset of electoral infrastructure across 12 SSA countries}
}
```

## License

CC-BY-4.0