Tax-benefit microsimulation

PolicyEngine microsimulation

Computes what a reform costs and who gains or loses, for one household or the whole population.

Exact for rules; costings static, no interval. Evidence Household arithmetic is checked rule by rule. Population estimates add survey error.

Good for

  • Taxes, benefits and net income for one household
  • Static costings and decile incidence, UK and US
  • The static costing the OBR emulator and OG-UK start from

Not for

  • Behavioural or macro feedback: OBR emulator, OG-UK
  • Rules not yet implemented in the model (coverage)
  • Intervals on population estimates: none computed (limits)

What it returns

1p on the UK basic rate, costed on the enhanced FRS.

Static revenue, £bn a year

PolicyEngine microsimulation: 1p on the UK basic rate, our static costing vs the HMRC ready reckoner (£bn/yr) Grouped bar chart in billions of pounds per year. PolicyEngine's static costing of a 1 percentage point rise in the UK basic rate of income tax, against HMRC's Direct effects of illustrative tax changes ready reckoner, June 2025 vintage. For the basic rate +1pp, 2026–27 group, ours is 6.55 against HMRC’s 6.90, a deviation of -5.1%. For the basic rate +1pp, 2028–29 group, ours is 7.01 against HMRC’s 8.20, a deviation of -14.5%. Both microsimulation figures are scored runs on the hosted server (policyengine 6.11.1): £6.55bn in 2026 and £7.01bn in 2028.
  • ours
  • HMRC
Static costing of a 1pp basic-rate rise against HMRC’s Direct effects of illustrative tax changes (June 2025 vintage). Ours are scored runs on the hosted server (policyengine 6.11.1; run record).

The same call also returns incidence by decile. Macro feedback on this yield comes from the OBR emulator.

A benchmark, not a validation. HMRC’s figure is post-behavioural and ours is static, so the gap is not an error measure (why).

With the macro models

Every bridge in the suite starts or ends here.

Reform PolicyEngine microsimulationstatic costing OBR emulator · OG-UKmacro feedback Dynamic score

score_reform accepts microsim, obr, og and og+microsim (score a reform). The BoE SVAR, FRB/US, US HANK and DEFINE-UK connect the other way, as incidence on households (every join in the suite).

Model cardinputs, outputs, horizon, data, access
Answers
What a reform costs and who gains or loses, for one household or the whole population
Inputs
A household you describe, or survey microdata; a parameter reform
Outputs
Taxes, benefits, net income, revenue, distribution
Horizon
Single policy year
Data
UK: enhanced FRS (enhanced_frs_2023_24) · US: CPS · recorded per run
Uncertainty
None for household arithmetic; survey and calibration error for population estimates, no interval computed
Access
Hosted MCP, CLI, Python
Runtime
Sub-second per household, minutes per population run
Status
Exact for implemented rules (household arithmetic); population estimates add survey error
Cannot answer
GDP, inflation, interest rates or macro feedback: costings are static (limits)