Expert validation of the simulator "Russia: economy, defence industry and governability"
Dear expert,
The simulator computes whether Russia can keep paying for the war: the budget constraint, the physical defence-industrial loop, the ammunition balance, the reserve loop, a banking sub-model, a regional cut and governability. Above these loops runs a module of endogenous Kremlin response: as pressure grows, the authorities themselves switch on taxes, spending cuts, debt, reductions in regional transfers, devaluation, hidden monetisation and social payments — with a lag of two to three quarters.
That module is the model’s weakest point. The thresholds at which instruments switch on, and the speed at which their use ramps up, are expert assumptions rather than estimates from data. The same applies to the weights of the four cost dimensions for the Ukrainian side, to the split of measures into «levers / Kremlin responses / exogenous conditions», and to the structure of the Urals discount. The uncertainty band (2,000 runs, 13 multipliers at ±30%) describes sensitivity to parameters, but not to the choice of specification itself.
This survey exists to convert those quantities from one team’s estimate into a median and a spread across a group of experts, and to surface structural errors: missing channels of influence, wishes misclassified as levers, double counting between loops.
Completion takes about 16 minutes. Answer only where you are competent: nearly every question offers «not my domain». Skipping is better than guessing — an averaged random answer narrows the spread and creates a false impression of consensus.
What we do with the answers. We publish only anonymised and only in aggregate: the median and the interquartile range as the interval of expert disagreement. A bimodal distribution is not averaged; both positions are described as competing hypotheses. Structural comments are not aggregated numerically — they go into the model’s list of limitations.
The most valuable answer for us is not agreement, but specific disagreement with an explanation of the mechanism.
With respect,
The Advanter Group team