electricsheepafrica/african-court-case-backlogs

---
license: cc-by-4.0
task_categories:
- tabular-classification
- tabular-regression
language:
- en
tags:
- governance
- judicial-systems
- court-backlogs
- rule-of-law
- sub-saharan-africa
- synthetic
- public-administration
- justice
- lmic
pretty_name: African Court Case Backlogs
size_categories:
- 10K<n<100K
configs:
- config_name: baseline
data_files: data/baseline.csv
default: true
- config_name: reform_progress
data_files: data/reform_progress.csv
- config_name: crisis
data_files: data/crisis.csv
---

# African Court Case Backlogs

## Abstract

A synthetic dataset modeling court case backlog dynamics across 12 sub-Saharan African countries (2018–2025), parameterized from national judiciary annual reports, the World Justice Project Rule of Law Index, and World Bank assessments. The dataset contains 10,000 records per scenario across three judicial burden scenarios (baseline, reform_progress, crisis), with 16 variables covering case volumes, clearance rates, disposition times, judge density, and backlog severity classifications. Designed for ML classification, regression, and policy simulation tasks in the judicial governance domain.

## 1. Introduction

Court case backlogs represent one of the most critical governance challenges across sub-Saharan Africa. With no equivalent ML-ready dataset on HuggingFace, this resource fills a significant gap for researchers, development finance institutions (DFIs), GovTech companies, and policy organizations working on judicial reform, access to justice, and rule-of-law programming.

National judiciary reports reveal dramatic variation across the continent: Kenya achieves 99–104% clearance rates through technology-enabled reforms, while Nigeria's Supreme Court clears only 16% of matters considered. Tanzania reduced its backlog rate from 11% to 3% through systematic reform, while Rwanda's backlog grew 72% in four years despite strong rule-of-law scores. These dynamics are captured in the scenario-based generation approach.

## 2. Methodology

### 2.1 Target Population

Court-level annual records for 12 sub-Saharan African countries spanning 2018–2025, across three court hierarchy levels (supreme/appellate, high court, magistrate/lower courts).

**Countries included:** Nigeria, DRC, Kenya, Ghana, Tanzania, Uganda, Rwanda, Botswana, Mauritius, South Africa, Senegal, Namibia.

### 2.2 Variable Selection

Variables were selected based on availability in national judiciary reports and international indices, following the CEPEJ (European Commission for the Efficiency of Justice) framework adapted for SSA contexts.

### 2.3 Epidemiological Parameterization

All parameters are grounded in peer-reviewed literature and official reports. The source hierarchy follows:

| Priority | Source Type | Examples Used |
|----------|-----------|---------------|
| 1 | National judiciary annual reports | Kenya SOJAR, Uganda APR, Ghana Judicial Service, Rwanda Strategic Plan |
| 2 | International composite indices | WJP Rule of Law Index 2024/2025 |
| 3 | Development institution assessments | World Bank, UNDP Rule of Law Reports |
| 4 | Academic studies | Voigt & El-Bialy (2024), Kondylis & Stein (2023) |
| 5 | Surveys | Afrobarometer Round 6-9, Gallup |

#### Parameterization Evidence Table

| Parameter | Value Used | Source | DOI/URL | Year | Note |
|-----------|-----------|--------|---------|------|------|
| Kenya clearance rate | 99–104% | Kenya Judiciary SOJAR | judiciary.go.ke | 2023/24 | Cases resolved / cases filed |
| Nigeria SC clearance | ~16% | CJN Annual Address | nairametrics.com | 2024/25 | 369 judgments / 2,280 matters |
| Ghana resolution rate | 49.8% | Ghana Judicial Service | judicial.gov.gh | 2023 | Cases tried / cases filed |
| Tanzania backlog rate | 3% (down from 11%) | Tanzania Judiciary / World Bank | worldbank.org | 2024 | Post-reform achievement |
| Uganda pending cases | 198,554 (48,326 backlog) | Uganda Judiciary APR | judiciary.go.ug | 2025 | Backlog = >2 years unresolved |
| Rwanda backlog rate | 59% | Rwanda Judiciary Strategic Plan | rcsprwanda.org | 2023/24 | Up from 25% in 2017/18 |
| DRC judge density | 6.9 per 100K | UNDP DRC Report | rolhr.undp.org | 2023 | Post-2,500 magistrate recruitment |
| Kenya judge density | 1.3 per 100K (incl. magistrates) | Kenya Judiciary | judiciary.go.ke | 2019 | 153 judges + 546 magistrates |
| Criminal case share | 57% | Kenya SOJAR | judiciary.go.ke | 2023/24 | Declining 3% YoY |
| WJP scores (12 countries) | 0.34–0.63 | WJP Rule of Law Index | worldjusticeproject.org | 2025 | Rwanda highest in SSA (0.63) |
| International judge benchmark | 10 per 100K | UNDP | undp.org | — | SSA countries at 0.3–7.0 |
| SSA contract enforcement | +158 days vs Europe | World Bank / ScienceDirect | doi.org/10.1016/j.jpubeco.2023.104953 | 2023 | Additional delay over ECA average |
| Public trust in courts | 53% | Afrobarometer | afrobarometer.org | Round 6+ | Declining from 62% in 2005/06 |

### 2.4 Scenario Design

| Scenario | Description | Clearance Mult | Disposition Mult | Backlog Rate (mean) |
|----------|-------------|---------------|-----------------|-------------------|
| **baseline** | Current SSA judicial landscape (2018–2025) | 1.0× | 1.0× | ~0.61 |
| **reform_progress** | Active judicial reforms (e.g., Tanzania, Kenya post-2020) | 1.3× | 0.7× | ~0.51 |
| **crisis** | Overburdened judiciary under fiscal/political stress | 0.6× | 1.5× | ~0.74 |

### 2.5 Generation Process

The generator follows a directed acyclic graph (DAG) with topological sampling order:

1. **Root nodes** (sampled independently): country (weighted by population), year (uniform 2018–2025), court_level
2. **Intermediate nodes** (sampled conditionally): population, judge_density, cases_filed (with annual growth), case_type, clearance_rate, cases_resolved, avg_disposition_months
3. **Leaf nodes** (derived): pending_cases (with independently noised carryover), backlog_rate, backlog_severity classification, wjp_score

Key technique: Backlog accumulation uses log-normally distributed carryover and prior-year backlog stock to decouple backlog_rate from same-year clearance_rate (observed r ≈ −0.89 vs mechanical r ≈ −0.99 without noise injection).

## 3. Dataset Description

### 3.1 Schema

| Column | Type | Units | Range | Description |
|--------|------|-------|-------|-------------|
| record_id | int | — | 1–10,000 | Unique record identifier |
| country | categorical | — | 12 countries | Sub-Saharan African country |
| year | int | year | 2018–2025 | Observation year |
| court_level | categorical | — | 3 levels | Court hierarchy: supreme_appellate, high_court, magistrate_lower |
| population | int | persons | varies | Estimated national population for that year |
| judge_density_per_100k | float | judges/100K | 0.3–8.0 | Judges + magistrates per 100,000 population |
| total_judges | int | persons | ≥1 | Estimated total judges in country |
| cases_filed | int | cases | ≥1 | Cases filed at this court level in this year |
| case_type | categorical | — | 7 types | Dominant case type: criminal, civil_general, family_matrimonial, land_property, commercial, labour_employment, constitutional_admin |
| clearance_rate | float | ratio | 0.05–1.20 | Cases resolved / cases filed (>1.0 = clearing backlog) |
| cases_resolved | int | cases | ≥0 | Cases resolved in the year |
| avg_disposition_months | float | months | 1–72 | Average time from filing to disposition |
| pending_cases | int | cases | ≥0 | Total pending cases (current + carryover) |
| backlog_rate | float | ratio | 0.0–1.0 | Pending cases / total caseload |
| backlog_severity | categorical | — | 4 levels | low (<30%), moderate (30–50%), high (50–70%), critical (>70%) |
| wjp_rule_of_law_score | float | score | 0.15–0.90 | World Justice Project Rule of Law Index score |

### 3.2 Classification Criteria

| Class | Criteria | Source |
|-------|----------|--------|
| **low** backlog | backlog_rate < 0.30 | Based on Tanzania post-reform target (~3%) and CEPEJ benchmarks |
| **moderate** backlog | 0.30 ≤ backlog_rate < 0.50 | Ghana-level clearance (~50%) |
| **high** backlog | 0.50 ≤ backlog_rate < 0.70 | Rwanda current level (~59%) |
| **critical** backlog | backlog_rate ≥ 0.70 | Nigeria/DRC-level systemic crisis |

### 3.3 Summary Statistics (baseline scenario)

| Variable | Mean | SD | Min | Max |
|----------|------|-----|-----|-----|
| clearance_rate | 0.584 | 0.208 | 0.062 | 1.200 |
| backlog_rate | 0.614 | 0.161 | 0.002 | 0.962 |
| avg_disposition_months | 17.2 | 9.8 | 1.0 | 71.2 |
| judge_density_per_100k | 2.06 | 1.06 | 0.30 | 7.63 |
| wjp_rule_of_law_score | 0.437 | 0.081 | 0.214 | 0.746 |

## 4. Validation

### 4.1 Prevalence Fidelity

| Outcome | Target Range | Observed (baseline) | Status |
|---------|-------------|-------------------|--------|
| Criminal cases | 40–60% | 52.5% | PASS |
| Civil cases | 12–30% | 20.1% | PASS |
| Land/property | 3–18% | 9.3% | PASS |
| Backlog severity: critical | 2–80% | 32.3% | PASS |

### 4.2 Distribution Quality

All continuous variables pass moment checks against literature benchmarks across all three scenarios.

### 4.3 Correlation Structure

| Pair | Target r | Observed r | Status |
|------|----------|-----------|--------|
| clearance_rate ↔ backlog_rate | −0.85 | −0.894 | PASS |
| clearance_rate ↔ disposition_months | −0.20 | −0.052 | PASS |
| judge_density ↔ clearance_rate | 0.10 | −0.007 | PASS |
| backlog_rate ↔ disposition_months | 0.15 | 0.108 | PASS |

### 4.4 Cross-Scenario Monotonicity

| Metric | Reform | Baseline | Crisis | Monotonic? |
|--------|--------|----------|--------|-----------|
| backlog_rate (mean) | 0.510 | 0.614 | 0.736 | Yes |
| avg_disposition_months | 13.3 | 17.2 | 24.9 | Yes |

### 4.5 Diagnostic Plots


## 5. Usage

### 5.1 Loading with HuggingFace datasets

```python
from datasets import load_dataset

# Load baseline scenario (default)
ds = load_dataset("electricsheepafrica/african-court-case-backlogs")

# Load specific scenario
ds = load_dataset("electricsheepafrica/african-court-case-backlogs", "crisis")
```

### 5.2 Loading directly from CSV

```python
import pandas as pd

df = pd.read_csv("data/baseline.csv")
print(df.shape)
print(df.describe())
```

### 5.3 Regenerating with custom parameters

```bash
# Install dependencies
pip install numpy pandas scipy matplotlib

# Generate baseline (10K records)
python generate_dataset.py --scenario baseline --n 10000 --seed 42

# Generate all scenarios
for scenario in baseline reform_progress crisis; do
python generate_dataset.py --scenario $scenario --n 10000 --seed 42
done

# Run validation
python validate_dataset.py
```

## 6. Limitations & Ethical Considerations

1. **Synthetic data**: This dataset is synthetically generated and must not be used as a substitute for real judicial statistics in policy decisions, litigation, or official reporting.

2. **Data gaps**: Several countries (DRC, Botswana, Mauritius) lack comprehensive published judicial statistics. Parameters for these countries are estimated from WJP scores, population ratios, and peer-country benchmarks rather than direct administrative data.

3. **Definition inconsistency**: "Backlog" is defined differently across jurisdictions (Uganda: >2 years; Rwanda: >6 months; some countries have no formal definition). The dataset uses a unified backlog_rate metric that may not match any single country's definition.

4. **Clearance rate methodology**: Kenya uses cases resolved/cases filed (CEPEJ methodology), Ghana uses cases tried/cases filed. These methodological differences are smoothed in the synthetic data.

5. **Informal justice excluded**: Afrobarometer data shows a majority of sub-Saharan Africans would not use formal courts. Formal court statistics (and this dataset) capture only a fraction of justice-seeking behavior.

6. **Temporal simplification**: The model does not capture COVID-19 disruption effects on 2020–2021 court operations, seasonal filing patterns, or reform discontinuities.

7. **No individual-level data**: Records represent court-level aggregates, not individual cases. No personally identifiable information is modeled.

## 7. References

1. Kenya Judiciary, *State of the Judiciary and Administration of Justice (SOJAR) Annual Report FY 2023/24*. Nairobi: Judiciary of Kenya, 2024.
2. Uganda Judiciary, *Annual Performance Report FY 2022/2023*. Kampala: Judiciary of Uganda, 2023.
3. Ghana Judicial Service, *Analysis of Civil Cases and Criminal Offences 2022/2023*. Accra: Judicial Service of Ghana, 2023.
4. Rwanda Judiciary, *Strategic Plan 2018-2024*. Kigali: Judiciary of Rwanda, 2023.
5. Tanzania Judiciary, *Case Backlog Reduction Strategies* (presented at AJC 2024).
6. World Justice Project, *WJP Rule of Law Index 2024/2025*. Washington, DC: WJP, 2024/2025.
7. UNDP, *Rule of Law and Human Rights Annual Report 2023: DRC*.
8. Mustard Insights, *Ranking: African Countries with the Most Effective Judicial System (2022)*.
9. Nairametrics, *CJN reveals Supreme Court delivered 369 judgments from 2,280 matters*, Sept 2025.
10. World Bank, *Bringing Justice Within the Grasp of Ordinary Men and Women in Tanzania*, July 2024.
11. Afrobarometer, *Access to Justice Is Still Elusive for Many Africans* (Policy Paper No. 39).
12. Kenna Partners, *Civil Litigation Duration in Nigeria*, 2023.
13. Capital FM Kenya, *Courts Post 99pc Clearance Rate*, Nov 2024.
14. Voigt, S. & El-Bialy, N., *Explaining Cross-Country Differences in the Number of Judges per Capita*, IJCA, 2024. DOI: 10.36745/ijca.581.
15. Kondylis, F. & Stein, M., *Courts, Crime and Economic Performance*, Journal of Public Economics, 2023. DOI: 10.1016/j.jpubeco.2023.104953.
16. UNDP Rwanda, *Rwanda Launches Electronic Court System to Deliver Justice for All*.
17. Gallup, *Majority in Sub-Saharan Africa Wouldn't Use Formal Courts*.
18. World Bank, *Doing Business 2020*. Washington, DC: World Bank, 2020.

## Citation

```bibtex
@dataset{esa_court_backlogs_2026,
title={African Court Case Backlogs},
author={{Electric Sheep Africa}},
year={2026},
publisher={HuggingFace},
url={https://huggingface.co/datasets/electricsheepafrica/african-court-case-backlogs},
license={CC-BY-4.0}
}
```

## License

CC-BY-4.0