TL;DR: We've rebuilt GiveCalc as a standalone web application at givecalc.org. The free calculator now covers US federal and state taxes, UK Gift Aid, and tax years 2024-2026. Use it to find out how much your charitable donations actually cost after tax savings.
A year ago, I introduced GiveCalc as a Streamlit app on PolicyEngine's infrastructure. This Giving Tuesday, we're launching GiveCalc 2.0—rebuilt from the ground up with a faster React interface, expanded to cover UK Gift Aid, and now available at its own domain: givecalc.org.
The new version supports multiple income types (wages, self-employment, capital gains, dividends, interest), works on mobile, and lets you calculate for tax years 2024, 2025, or 2026.
GiveCalc uses PolicyEngine's microsimulation models—the same engines used at the highest levels of US and UK policymaking, as well as policy research organizations like the National Bureau of Economic Research, Niskanen Center, and University of Michigan.
For US taxpayers, enter your income, filing status, state, and itemized deductions. For UK taxpayers, enter your income, region, and Gift Aid amount. GiveCalc computes your taxes with and without the donation, showing your total tax savings, marginal giving discount (tax saved per additional dollar or pound), and net cost.
Take a single California filer earning $200,000 with $20,000 in mortgage interest. A $20,000 charitable donation saves them $5,857 in federal and state taxes, making their net cost $14,143—about 71% of the donation amount. Their marginal giving discount is 33%, meaning each additional dollar donated saves 33 cents in taxes.
![IMAGE: GiveCalc screenshot showing $20k donation calculation - ~/Downloads/GiveCalc_screens/1.png]
US taxpayers who itemize can save money by "bunching"—concentrating multiple years of donations into a single year to exceed the standard deduction threshold.
Consider the same California filer donating $20,000 per year. In 2025, their net cost is $14,143. In 2026, it's $14,621 (the standard deduction rises with inflation). That's $28,764 total over two years.
But if they bunch $40,000 into 2025 and skip 2026, their net cost is $27,483. Bunching saves $1,281—a 4.5% reduction in the cost of giving.
![IMAGE: GiveCalc screenshot showing 2025 $20k donation - ~/Downloads/GiveCalc_screens/2.png]
![IMAGE: GiveCalc screenshot showing 2025 $40k bunched donation - ~/Downloads/GiveCalc_screens/3.png]
GiveCalc now supports UK taxpayers. Gift Aid provides tax relief at two levels: charities claim back the basic rate tax (20%), receiving £1.25 for every £1 you give, and higher rate taxpayers claim back the difference on Self Assessment.
A higher rate taxpayer donating £1,000 pays a net cost of just £750. The charity receives £1,250 (grossed up), and the donor claims £250 in tax relief. Everyone benefits except HMRC.
Scottish taxpayers face different rates (19%-48%), which GiveCalc automatically applies based on your selected region.
![IMAGE: GiveCalc UK screenshot showing Gift Aid calculation - TODO: capture UK example]
PolicyEngine has signed a memorandum of understanding with NBER to develop a TAXSIM emulator, validating our tax calculations against the research community's standard.
PolicyEngine advances evidence-based policy through open-source economic simulation models. We build epistemic infrastructure that improves institutional decisionmaking by making tax and benefit policy transparent, accessible, and rigorously quantifiable.
Our core work falls into three areas:
Policy understanding: PolicyEngine's household calculator lets anyone model how taxes and benefits affect their specific situation—and how proposed reforms would change it. This reduces information asymmetries between policymakers and the public.
Benefit access: Tools like MyFriendBen, Amplifi Benefit Navigator, and Starlight use PolicyEngine's API to screen families for 40+ federal and state programs. Collectively, these tools have identified over $1 billion in unclaimed benefits—direct welfare gains from reducing administrative barriers.
AI and inequality research: We're investigating how public policy mediates the relationship between AI-driven economic transformation and distributional outcomes. Our research framework examines the causal chain from AI economic shocks (rising productivity, capital share growth, labor disruption) through policy interventions to effects on income, consumption, and wealth inequality.
This addresses an underexplored question: as AI potentially drives major economic shifts—unevenly distributed productivity gains, changing capital-labor shares, technological unemployment—how would different policy responses shape who benefits and who loses? We model how current policies versus alternatives (UBI, expanded safety nets, capital taxation) would differentially mediate these AI-driven changes. Microsimulation provides a rigorous framework for quantifying these distributional effects across multiple AI scenarios rather than relying on point estimates.
Donations to PolicyEngine are tax-deductible in the US and eligible for Gift Aid in the UK (charity no. 1210532).
Donate to PolicyEngine and use GiveCalc to calculate your tax savings.
Visit givecalc.org to calculate how your charitable giving affects your taxes this Giving Tuesday. The source code is fully open source.