The counterfactual question

“What would Romania's GDP be today if it had done everything Poland did?” The honest answer is: no one knows. But a framework can make the question sharp enough to examine. The engine separates two distinct things: Romania keeps its own 1990 starting conditions and population path, while inheriting a fraction of Poland's annual growth trajectory. That framing prevents the shortcut of simply copying Poland's aggregate GDP onto Romania — which conflates population size with per-person growth.

Confidence bands

Counterfactuals should not be reported as single-point estimates. Uncertainty compounds over time — a projection for 2024 based on 1990 assumptions has far more degrees of freedom than a projection for 1992. The engine models this with square-root uncertainty growth: $\sigma_t = \sigma_0 \sqrt{t/T}$. The shaded band on the GDP chart shows the range.

Policy gap decomposition

The policy basket captures seven dimensions (institutions, investment, education, export complexity, macro stability, state capacity, EU absorption). The gap decomposition shows which dimensions contribute most to the overall difference. This is a cleaner answer than “Poland grew more” — it lets you say “the biggest gap was in export complexity and institutions.”

Reverse framing

The framework is symmetric. The same machinery that asks “what if Romania had been like Poland?” can ask “what if Poland had been like Romania?”. Toggle reverse framing to test whether the model country's outperformance is treated as normative because of evidence or because of ex post availability bias.

What this is not

This is not a causal claim. Poland's outperformance reflects a bundle of decisions, institutional trajectories, EU integration, demographic and geographic factors, and luck. The counterfactual gap is an outcome-proxy, not a courtroom damages figure. The value of the framework is that it makes the comparison precise enough to examine, debate, and improve — not that it delivers a single true number.

Model changelog