Tax Lab · Macro-fiscal & revenue mobilisation
Uganda Fiscal Deficit Model
A real Monte Carlo stress test. Four shocks, calibrated from Uganda's own 2014-2024 fiscal history, propagate through the prototype's exact formulas to simulate the deficit-to-GDP and debt-to-GDP paths through 2029.
Calibration and IMF baseline are real; shock draws are stochastic by design.
What this simulates
Starting from Uganda's own IMF World Economic Outlook baseline (April 2024 vintage) for 2025-2029, each simulated year draws four possible shocks, an oil price collapse, an aid withdrawal, an exchange rate depreciation, and a GDP growth slowdown, each active with its own calibrated annual probability. When a shock fires, it feeds into revenue or expenditure through a specific elasticity, and the fiscal deficit (revenue minus expenditure) and the public debt path (via the Blanchard debt-dynamics equation) update accordingly. Run thousands of independent trajectories and the spread of outcomes is the fan chart below.
The simulation
The four shocks below start at the prototype's own calibrated defaults. Adjust a probability or a shock size and press run for a fresh set of random trajectories.
Oil price collapse
Aid withdrawal
Exchange rate depreciation
GDP growth slowdown
Press run to simulate. Sliders start at the prototype's own calibrated defaults.
Where the calibration comes from
The noise parameters (GDP growth volatility, expenditure rigidity, oil price log-volatility, exchange rate depreciation mean and spread) are computed directly from the prototype's own uganda_fiscal_data.csv and commodity_prices.csv, over the same 2014-2024 calibration window its R code uses. The debt dynamics follow Blanchard's (1990) standard equation at a flat 3.5% assumed real interest rate. The four shock magnitudes and probabilities (25% chance of an oil shock averaging -40%, etc.) are the prototype's own stated calibration, referencing real Ugandan episodes: the 2014-16 oil collapse, the 2012-16 aid/GDP decline, and the 2011 and 2015 currency depreciations.
Limitations, stated by the prototype itself
- Shocks are drawn independently each year; in reality oil price and exchange rate shocks are correlated.
- The model doesn't account for Uganda's oil production coming online, which would materially change revenue dynamics.
- Bank of Uganda's policy response to depreciation and inflation isn't modelled.
- No cross-country contagion from EAC trade partners.
Sources: MoFPED Budget Speeches and IMF Uganda country pages (historical fiscal data, 2000-2024); IMF World Economic Outlook, April 2024 vintage (2025-2029 baseline); World Bank Pink Sheet (commodity prices). Blanchard, O. (1990), Suggestions for a New Set of Fiscal Indicators, OECD Working Paper 79.