policyengine-us

What it is and what it does

PolicyEngine US is a microsimulation engine that encodes the rules of the US federal and state tax and benefit system, allowing you to compute tax liabilities, benefit eligibility, and household-level financial outcomes under different policy scenarios. It is built on policyengine-core and integrates with pandas, microdf-python, spm-calculator, and tables to handle data processing and tabular storage.

The package is actively maintained and positioned as the authoritative source for US tax-benefit rules and household calculations. However, the description explicitly notes that direct microsimulation via the Microsimulation() class is deprecated; the maintainers recommend using the policyengine.py bundle for population-wide modeling, which pins certified and calibration-gated datasets. Microsimulation results are estimates and should be validated against the calibration diagnostics.

Use it for:

• Compute household tax liability and benefit eligibility under current or proposed tax-benefit rules.
• Analyze distributional or budgetary impacts of policy changes on specific household types.
• Validate policy proposals by running microsimulation scenarios against calibrated US demographic and income data.
• Integrate US tax-benefit calculations into a larger policy analysis or research workflow.
• Inspect and audit the encoded rules of the federal and state tax-benefit system.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

PolicyEngine US models the US federal and state tax and benefit system for household-level microsimulation calculations, computing tax liabilities, benefit eligibility, and distributional impacts.

Yes, if you need to model US household-level tax and benefit outcomes and are comfortable with AGPL v3 licensing. The package is actively maintained, has low install friction, and is the canonical source for US tax-benefit rules. However, note that direct microsimulation is deprecated in favor of the policyengine.py bundle for population-wide analysis. Verify that your Python version (3.11–3.14) and pandas 3+ are available before installing.

Install

policyengine-us on PyPI

pip install policyengine-us

uv add policyengine-us

poetry add policyengine-us

Installing policyengine-us

Before you install

Low install friction with a pure-Python wheel. Actively maintained with a release on 2026-08-14. Requires Python 3.11–3.14 and pandas 3 or later; verify your environment matches before installing.

License in practice

Licensed under AGPL v3, which requires that any modifications or derivative works be shared under the same license and that source code be made available to users. This is a strong copyleft obligation suitable for open research but may constrain commercial or proprietary use.

Quickstart

pip install policyengine-us

from policyengine_us import Microsimulation

sim = Microsimulation()
# Household-level tax and benefit calculations follow

Requires Python 3.11–3.14 and pandas 3 or later. Direct microsimulation via Microsimulation() is deprecated; the description recommends using the policyengine.py bundle for population-wide modeling instead.

Verify before relying

• Exact calibration quality and validation metrics for microsimulation estimates beyond the reference to calibration-diagnostics.vercel.app.
• Performance characteristics (runtime, memory) for typical household-scale and population-scale simulations.
• Whether the package's tax-benefit rules are current with the latest federal and state legislation as of the fact sheet date.

Package facts

Evidence: policyengine_us-1.808.0-py3-none-any.whl

Tags

More Information Analysis packages

Further reading