Methodology and data

HIXI risk-factor index

• A. Labour-market formality: employment rate (↓ risk) and derived unemployment = (activity − employment) / activity (↑ risk).
• B. Income capacity: GRP per capita relative to the national level, logged (↓ risk). The ratio removes inflation.
• C. Economic structure: small-business share of GRP (↑ risk). Many small units are harder to observe; this does not mean small business is shadow.
• Normalisation: min-max over all region-years 2017–2025 with fixed bounds, so years are comparable.
• Aggregation: equal within components, user weights between them; arithmetic or geometric mean.
• Sensitivity: 2,000 Monte Carlo draws (Dirichlet weights × method). Validation: Spearman ρ against the official non-observed share, which is not an input.

Security zones

Non-observed share < 20%: within norm; 20–40%: above threshold; ≥ 40%: critical. Unemployment threshold: 8%. Thresholds come from the literature on Russian regions; Uzbekistan has no officially approved thresholds.

Simulator

• Layer 1: pure arithmetic. Extra revenue = (legalised shadow + registered informal value added) × effective fiscal burden. No causal claim.
• Layer 2: panel regression, 14 regions × 9 years, region and year fixed effects, CR1 clustered errors, t(13) critical value. The result is an association, not an effect.

District level

The map drills down to 199 districts and cities. The non-observed economy and index inputs are published by region only, so districts are coloured with their region’s value and labelled as such. Tashkent City boundaries are outdated (7 districts).

Limitations

• HIXI is not an official shadow-economy estimate and does not replace one.
• 14 regions × 9 years = 126 observations: low statistical power, unstable regression.
• The methodological reason for the sharp 2024→2025 drop in the informal share is not officially explained.
• Regional series for cashless payments and tax revenue are not yet included.
• Some 2024–2025 values are marked preliminary in SIAT.

How it differs from the “Shadow Economy Map”

The Prosecutor General’s platform is a closed, enterprise-level enforcement tool. SHADOW.UZ uses only open aggregate data, publishes its full methodology, and is built for independent assessment of policy effectiveness. No personal data is used.

Open API

FastAPI service (api/main.py). All responses are JSON with a source link.