metadata
license: cc-by-4.0
language:
- en
tags:
- insurance
- agriculture
- crop
- climate
- sub-saharan-africa
- synthetic
- agritech
- index-insurance
- drought
- food-security
pretty_name: African Crop Insurance Payouts
size_categories:
- 10K<n<100K
task_categories:
- tabular-regression
- tabular-classification
data_type: synthetic
⚠️ Synthetic dataset — Parameterized from published SSA literature, not real observations. Not suitable for empirical analysis or policy inference.
African Crop Insurance Payouts
Synthetic dataset of agricultural index-insurance payouts across 12 Sub-Saharan African countries (30,000 records). Parameters are calibrated from real-world programs including IBLI (Kenya/Ethiopia), WFP R4 Rural Resilience Initiative, and ACRE Africa.
Dataset Description
Scenarios
| Scenario | Description |
|---|---|
| baseline | Current-state parameters reflecting existing IBLI/R4/ACRE enrollment and payout patterns |
| expanded_index_insurance | 2.5× enrollment scaling via premium subsidy and mobile distribution (ACRE model) |
| climate_shock | +15 percentage point drought probability shift reflecting IPCC AR6 projections for SSA |
Each scenario contains ~10,000 records.
Countries
Kenya, Ethiopia, Tanzania, Malawi, Zambia, Zimbabwe, Mozambique, Senegal, Ghana, Nigeria, Rwanda, Uganda
Variables
| Variable | Type | Description |
|---|---|---|
| record_id | int | Unique identifier |
| country | str | Country name |
| year | int | Calendar year (2015–2024) |
| season | str | Growing season (long_rains / short_rains) |
| crop_type | str | maize / wheat / sorghum / rice / cassava |
| insurance_type | str | index_based / area_yield / weather_derivative |
| scenario | str | Policy scenario |
| enrolled_hectares | float | Hectares covered (0.2–20) |
| premium_paid_usd | float | Premium in USD |
| ndvi_anomaly | float | NDVI z-score (negative = vegetation deficit) |
| rainfall_deficit_pct | float | Percentage below normal rainfall |
| yield_loss_pct | float | Estimated yield loss (0–100%) |
| payout_triggered | bool | Whether index triggered a payout |
| payout_amount_usd | float | Payout disbursement in USD |
| payout_ratio | float | Payout as fraction of insured value |
| farmer_count | int | Farmers in the micro-cohort |
| enrollment_rate | float | Regional enrollment rate (0–1) |
| lapse_rate | float | Policy non-renewal rate (0–1) |
| basis_risk_score | float | Residual unhedged risk (0 = perfect, 1 = none) |
Parameter Calibration
Key parameters are drawn from published literature and program reports:
- Premium rates: IBLI charges 5.5%/3.25% of insured value for upper/lower Marsabit; R4 farmers contribute 10–15% of premium in cash or labor (WFP 2019)
- Payout trigger: IBLI indemnifies when cumulative NDVI deviation falls below the 15th percentile of historical conditions (ILRI 2013); ACRE triggers when area yield falls below 80% of historical average
- Basis risk: Jensen et al. (2016, AJAE) find IBLI reduces covariate risk by 63% but leaves policyholders with 69% of original risk from high-loss events
- Enrollment: R4 reached ~180,000 insured households in 2020 across 10 countries (WFP 2020); ACRE reached ~400,000 farmers by 2015 (GIIF 2016)
- Climate projections: IPCC AR6 WGII projects increased drought frequency across the Sahel and East Africa through 2050
Sources
- Jensen, N.D. & Barrett, C.B. (2016). "Index Insurance Quality and Basis Risk." American Journal of Agricultural Economics, 98(5), 1450–1469.
- WFP (2023). R4 Rural Resilience Initiative Annual Report.
- ACRE Africa / GIIF (2016). Scaling agricultural insurance in East Africa.
- Ntukamazina, N. et al. (2017). "Index-based agricultural insurance products in SSA." J. Agr. Rural Develop. Trop. Subtrop., 118(2), 171–185.
- IPCC (2022). AR6 WGII, Chapter 9: Africa.
Usage
from datasets import load_dataset
ds = load_dataset("electricsheepafrica/african-crop-insurance-payouts")
License
CC-BY-4.0