It estimates what U.S. tax and spending proposals do to the federal budget over ten years: what a policy costs, who bears it, and how much weight the number can carry.
It grew out of teaching public economics. Official scores from CBO, JCT and Treasury are the standard, but they arrive as finished verdicts. This tool shows the working. Every number names its data, its method and the nearest published score, and says whether the model was tuned to that score or predicted it blind. It is a companion to the Public Economics textbook, and Build lets you start from your values rather than from a list of policies.
Laurence Wilse-Samson, an economist (PhD, Columbia) who teaches at NYU's Wagner School. This is a personal research project. The views and modelling choices are my own and do not represent any employer.
Treat every figure as an estimate with a stated source, not an official score. Where the model predicts a published score it was never tuned to, it is off by 15.2% on average across 44 cases, and the gap varies a lot by kind of policy. Methods gives the figures for each kind and lists every benchmark.
Email lw3387@nyu.edu. Source code on GitHub.
Code is under the MIT License and written content under CC BY 4.0. Please cite as Wilse-Samson, L. (2026). Fiscal Policy Calculator. https://fiscal.lwilsesamson.com