About XiNi.app
Each analysis takes one UK policy and asks three questions: what it does to growth, what it does to the public finances, and who gains and who pays.
How an analysis is made
A detailed policy. Each idea is first written up as a costable policy: the measures and their parameters, start dates and phasing, the cost each year and how it is paid for. The full specification is on every analysis page.
Traced through the OBR’s model. An AI model (Claude) reads the Office for Budget Responsibility’s published macroeconomic model, its equations and variables, and traces the policy through them: which levers it moves, and how the effects pass to output, jobs, prices, interest rates and borrowing.
Costed. The change in borrowing is built up line by line, from the direct cost through to the tax receipts from extra activity and the cost of higher debt interest.
Who gains and who pays. The direct effect on households by income tenth and on the groups most affected, plus who ultimately pays for any extra borrowing.
Wider checks. The analysis also looks beyond the model’s equations: changes in land and property values, how quickly the policy could realistically be delivered, Barnett consequentials for spending in England, and anything else material, such as pressure on the fiscal rules or legal risks. Points that do not fit the tables appear under “Other things to weigh”.
Checked. A second review pass checks the figures for consistency (signs, units and totals) before anything is published.
Reading the figures
- Every figure is a change from the OBR’s March 2025 baseline forecast, the economy as it would be without the policy, measured in the final quarter of each year.
- Real GDP is the change in the size of the economy, in % and in £ a year at baseline prices. Public debt is the change in public sector net debt as a share of GDP (percentage points, “pp”) and in £.
- Winners and losers are £ per household a year unless marked one-off. They are the direct effect of the policy, before wider effects such as jobs and prices. Some are changes in wealth, such as homes worth a little less, and some are tangible gains that are not cash, such as homes that would not otherwise be built; each is labelled.
- Future taxpayers is every household’s share, across about 29 million households, of paying for any extra borrowing (or the saving from less): a yearly amount for a policy that continues, or the extra debt left behind by a temporary one, if the government’s fiscal rules are to hold.
Limits
The figures are reasoned estimates, not official forecasts. The policy is traced through the model’s published equations by reasoning, not by solving the model numerically, so small effects in particular are uncertain.
Every analysis runs to the end of 2032. The OBR’s own forecast ends in early 2030, so the last years extend its baseline at its final-year trend. Policies that pay off slowly, such as infrastructure, skills or housing, may show much of their benefit after 2032, so their costs can still look large next to their gains.
XiNi.app is independent. It is built on the OBR’s published model code but is not produced or endorsed by the Office for Budget Responsibility.
Follow new analyses
A new policy is analysed and published regularly. Follow along in any feed reader with the RSS feed.