PolicySandbox is a macroeconomic sandbox for Australia. Move a lever and watch inflation, mortgages, household budgets and wellbeing respond - with every assumption visible, every trade-off on one screen, and a named household attached to each number.
Most models give you a spreadsheet. PolicySandbox computes macro, households and politics together, so the second-order effects are impossible to miss - and refuses to hide a trade-off behind an average.
Inflation, jobs, GDP, wages and the federal interest bill, solved in deviation-from-baseline space across 132 variables and 342 relationships. Every coefficient carries an elasticity, a lag and a confidence class you can read.
64 synthetic cohorts and 21 named households - from a first home buyer with a mostly variable loan to a pensioner renting on the Gold Coast. Mortgage repayments use annuity arithmetic; housing stress is a threshold crossing over a distribution, not a scaled mean.
5 platforms expressed as lever settings, each run through the engine. Every household is assigned to whichever leaves it best off - modelled alignment, never an election forecast, and households too close to call are reported undecided rather than allocated.
The transmission mechanism as the engine actually wires it. One lever propagates through credit, housing and household budgets - and every channel here is a real edge in the graph, carrying its own elasticity, lag and confidence class.
Three monetary stances, the same indicators re-settling under each. Every figure is the engine’s, computed at build time - this panel cycles on its own, and a click takes the controls.
Cycling every few seconds. Click to take over.
Party alignment is not shown here on purpose. The engine models it, and it barely moves on the cash rate: the model is linear, so a shock applied equally to every platform shifts every household’s score by the same amount - and no party sets the cash rate anyway. Alignment swings sharply on tax, spending and migration, and it lives in the workspace where those levers are.
Move a cash rate, a tax rate, a migration target - or load a preset from the policy library. Or describe the policy in plain English and confirm what the parser understood before anything runs.
Every variable recomputes with its own lag, uncertainty band and decomposition. Click any number to see exactly which relationships produced it and how much to trust each one.
Winners and losers by household, region and age; the assumptions that mattered most; and what the model cannot see. The parts that are weak say so on the page.
Ten segments already spend serious budget answering “what happens if rates move”. PolicySandbox answers it live. Open your use case.
Pre-brief the hard questions before the press gallery asks them.
See use case →The only model that computes votes, not just GDP.
See use case →Turn every rate decision into a client conversation that lands.
See use case →Macro scenarios your investment committee - and your members - can touch.
See use case →Australia policy risk, on one screen, in one minute.
See use case →White-label the engine. Bill the insight.
See use case →A semester of macro intuition in one lecture.
See use case →Every RBA decision day becomes an interactive franchise.
See use case →Stop asserting the household cost of policy. Demonstrate it.
See use case →Scenario widgets that turn browsers into leads.
See use case →Everything here runs as a demo simulation on a plausible but unofficial baseline. Live simulations connect real ABS, RBA and Treasury series. Tell us who you are and what you would use it for.
The workspace is live now - every variable, every household, every assumption you can argue with. It takes one drag to see why the average was hiding something.