This repository contains a Matlab function that calculates federal tax liability, average and marginal tax rates from taxable income for a given year in the US (1913-2013).
The revolutionary Stata command taxsim9 calculates federal and state income tax liability from 21 input variables. The command is very sophisticated, determining, for instance, whether a tax filer is better off itemizing or claiming the standard deduction. However, taxsim9 is less useful for more basic calculations such as estimating tax liability and marginal tax rates given taxable income (rather than gross or adjusted gross income with deductions). Another Stata command, taxcalc (available from the NBER by request), calculates tax liability from more basic inputs, but is still sophisticated in terms of the inputs it receives and, more importantly, is limited to PUF and INSOLE data. You can read the NBER comparison of the commands here. Because taxsim9, taxcalc or — to mine or Kyle Rozema's knowledge—any other Stata command cannot perform analyses on the aggregated IRS SOI data on taxable income, we wrote the taxliab module for Stata. We produced the Matlab version of taxliab in order to perform the same calculations in Matlab.
Install taxliab by opening Stata and running the command:
net install taxliab, from(http://kylerozema.com/stata/taxliab)
The documentation for taxliab is included with the installation. To read the documentation, run the command:
help taxliab
To use the Matlab function contained in this repository, place the taxliab.m file along with its accompanying dataset UStaxschedule.mat into your Matlab working directory and call it in the standard way.
It requires three input variables: nominal taxable income, year, and filing status.
Filing status codes:
- Married filing together
- Married filing separately
- Single
- Head of Household
Run the taxliab_example.m script to demonstrate examples of how the taxliab.m function can be used. An example of the type of data that can be generated by taxliab is in the following graph.
<script data-plotly="hautahi:30" src="https://plot.ly/embed.js" async></script>