What is SARII?
The South Africa Risk Intelligence Index (SARII) is a composite score developed by SAFacts that aggregates data from four domains into a single 0–100 risk rating for South Africa. A higher score indicates greater risk; a lower score indicates a more stable environment.
SARII is designed for analysts, journalists, investors, NGOs, and policymakers who need a concise, data-driven summary of South Africa's risk environment — without the opacity of commercial country risk ratings.
Every SARII calculation is fully traceable: the underlying data, source documents, and formula are publicly documented. There are no black-box adjustments.
The Four Pillars
SARII aggregates four data domains, each given a specific weight in the composite score:
1. Crime Pillar — 30% weight
The crime pillar captures violent crime risk using SAPS data combined with Stats SA population estimates to derive per-capita rates.
The primary indicator is the murder rate per 100,000 population, which is the most internationally comparable and consistently reported crime measure. Secondary indicators include robbery and assault rates.
A high crime score reflects elevated violent crime rates relative to historical baselines and regional benchmarks.
2. Fiscal Pillar — 25% weight
The fiscal pillar measures government financial health using National Treasury budget data.
Key indicators include:
- Debt-to-GDP ratio: High debt loads increase fiscal vulnerability
- Debt service as a share of expenditure: Crowding out of productive spending
- Budget deficit: The gap between revenue and expenditure
A high fiscal score reflects a deteriorating fiscal position — rising debt, widening deficit, or accelerating debt service costs.
3. Economic Pillar — 25% weight
The economic pillar captures economic conditions and stability using Stats SA data.
Key indicators include:
- Expanded unemployment rate: The broadest measure of labour market slack
- Real GDP growth: Sustained low or negative growth increases economic vulnerability
- CPI inflation: Above-target inflation erodes purchasing power and monetary stability
A high economic score reflects high unemployment, stagnant growth, or above-target inflation.
4. Narrative Pillar — 20% weight
The narrative pillar captures media sentiment and information environment volatility using article volume data from NewsAPI.org.
The methodology:
- A daily count of South Africa-related articles is maintained
- A rolling 12-month baseline of average article volume is calculated
- Current volume is compared to the baseline using a z-score:
z = (current − mean) / std - Z-score thresholds:
z < 1→ low risk,1 ≤ z < 2→ moderate risk,z ≥ 2→ high risk
A spike in media coverage — particularly negative coverage during political crises, civil unrest, or economic shocks — contributes to a higher narrative score. Importantly, this pillar measures volatility, not sentiment; unusual quiet can also be informative.
The Composite Formula
The four pillar scores are combined using a weighted average:
SARII = (Crime × 0.30) + (Fiscal × 0.25) + (Economic × 0.25) + (Narrative × 0.20)
Each pillar score is normalised to a 0–100 scale before combination. The composite is calculated on a monthly basis (or more frequently when a daily narrative update is triggered).
What the Score Means
| Score Range | Interpretation | |---|---| | 0–25 | Low risk — stable conditions across most pillars | | 25–50 | Moderate risk — some indicators elevated, others stable | | 50–75 | Elevated risk — multiple pillars showing stress | | 75–100 | High risk — broad deterioration across pillars |
It is important to interpret SARII as a relative and directional indicator, not an absolute judgement. A score of 62 is not "bad" in isolation — it is meaningful relative to South Africa's own historical range and relative to trend direction (improving or deteriorating).
What SARII Is Not
SARII is not a prediction of specific events. It does not forecast coups, credit rating downgrades, or specific crime incidents. It is an aggregation of available data — as objective as the underlying sources allow.
SARII is also not comparable across countries in the current version. The pillar weights and normalisation baselines are calibrated to South Africa. A future version may support cross-country comparison once methodology for normalising across different data availability contexts is established.
Transparency and Methodology
The full SARII methodology is documented in the SAFacts Risk Index Methodology page. All source data is available through the SAFacts API. The pillar calculation code is open source.
If you believe a pillar weight should be adjusted, or that an important data source has been excluded, the SAFacts team welcomes feedback via GitHub.