The engine
How microsimulation works
PolicyEngine's microsimulation model combines three ingredients to estimate the impact of any tax or benefit reform on every household in the country.
Policy rules
Tax codes, benefit formulas, and eligibility rules encoded as executable logic. Over 55 federal, state, and local programs.
Household data
A representative microdata sample of the population, with income, demographics, and program participation for each household.
Behavioural dynamics
Elasticities capturing how people adjust their labour supply and income in response to tax and benefit changes.
A representative sample of households, each with their own characteristics.
| Household | Earnings | Children |
|---|---|---|
| Household A | £30,000 | 2 |
| Household B | £75,000 | 1 |
| Household C | £120,000 | 0 |
| Household D | £50,000 | 3 |
| Household E | £200,000 | 1 |
Current versions
PolicyEngine's country models are open-source Python packages, published to PyPI with every release.