This PR adds some guardrails to the calculation of the PYMTC. The calculation depends heavily on transition rates onto and off the AMT. I used transition rates based on Lyon (1997), but I made them sensitive to changes in the differential between the corporate tax rate and AMT rate. While exploring some results, I found that they were too sensitive, so this PR adds guardrails on changes to the transition rates, including reducing the sensitivity of the transition rates and ensuring that they are bounded between 0 and 1.
This does not change the test results, and it's mostly just relevant for unusual policy changes, like reducing the main corporate tax rate to close to the AMT rate.
I also modified the example.py file to round output to 4 decimals, in order to avoid floating point errors.
This PR adds some guardrails to the calculation of the PYMTC. The calculation depends heavily on transition rates onto and off the AMT. I used transition rates based on Lyon (1997), but I made them sensitive to changes in the differential between the corporate tax rate and AMT rate. While exploring some results, I found that they were too sensitive, so this PR adds guardrails on changes to the transition rates, including reducing the sensitivity of the transition rates and ensuring that they are bounded between 0 and 1.
This does not change the test results, and it's mostly just relevant for unusual policy changes, like reducing the main corporate tax rate to close to the AMT rate.
I also modified the
example.pyfile to round output to 4 decimals, in order to avoid floating point errors.