A reform moves every price.
We model the whole economy's answer.

Microsimulation tells you who pays what the morning after a reform. But people respond: they work, save, and invest differently, firms adjust capital, wages and interest rates move, and the revenue estimate moves with them. Most published policy scores stop before that second act — or bolt it on with a single hand-picked elasticity.

macromod makes the second act a model, not a fudge factor. Each model in the suite is a structural macroeconomic engine that takes a policy change and solves for the economy's response — changes in macroeconomic aggregates (GDP, investment, consumption), wages, interest rates, and tax revenues over time. Different models bring different machinery: a general-equilibrium overlapping-generations engine for the long run, an emulator of the OBR's own equations for near-term fiscal multipliers, and a Bayesian structural VAR that reads the current state of the economy in structural-shock terms.

A reform is a policy change, not a stylized wedge. How it enters depends on the model. The overlapping-generations engine takes a PolicyEngine policy object — actual tax-and-benefit parameters — runs it through PolicyEngine's UK microsimulation, and fits tax functions to the resulting effective and marginal rates, so it sees the reform's true shape across the income and age distribution. The OBR emulator takes a shock to one of its exogenous variables — a spending change, a rate change.

Both then solve for the response and report the deviation from a baseline anchored to official data — ONS, OBR, and Bank of England national accounts for the OLG model; the OBR's Economic and Fiscal Outlook for the emulator — so a result is always relative to current forecast, never a raw level to misread. The structural VAR plays a different role: it doesn't score reforms (yet) but reads the state the economy is in — which structural shocks are driving GDP and inflation right now — before you simulate changing it.