This PR performs the refactoring as described in issue #37 for C corporations. It now starts by extrapolating income and calculating individual components, and then computes tax revenue for the baseline.
The previous version started from the CBO corporate tax revenue estimate and worked backward, but this version starts with earnings and works forward to tax revenue. Accordingly, the results no longer line up perfectly with CBO (as expected).
In their forecast from June 2017 (which I'm still using here until I update to TCJA law), CBO projected that corporate profits (the growth factor I'm using for earnings) would increase from $2.088 trillion in 2015 to $2.725 trillion in 2027, an increase of 30.5%. They forecast corporate tax revenue to increase from $344 billion in 2015 to $443 billion in 2027, an increase of 28.8%.
My new version forecasts EBITDA to increase from $2.415 trillion in 2015 to $3.152 billion in 2027, an increase of 30.5% (by definition, as I use now CBO's profits forecast to increase EBITDA). However, the new forecast only has corporate tax revenue increasing from $321.2 billion in 2015 to $329.8 billion in 2027, an increase of 2.7%. Upon further investigation, this occurs because two components of the corporate income tax, specifically the net interest deduction (NID) and the foreign tax credit, grow much more quickly over this period than other components (122.4% growth for the NID and 59.0% growth for the FTC). The very large growth rate for the NID occurs because the capital stock increases but primarily because interest rates are forecast to almost double during this period. The FTC grows so quickly because CBO forecasts corporate foreign profits to grow much more quickly than domestic or overall profits.
Subsequent work should endeavor to adjust the growth factor for EBITDA to accommodate the role of NID growth (as EBITDA precedes NID but the CBO forecast for corporate profits comes after subtracting off the NID).
This PR performs the refactoring as described in issue #37 for C corporations. It now starts by extrapolating income and calculating individual components, and then computes tax revenue for the baseline.
The previous version started from the CBO corporate tax revenue estimate and worked backward, but this version starts with earnings and works forward to tax revenue. Accordingly, the results no longer line up perfectly with CBO (as expected).
In their forecast from June 2017 (which I'm still using here until I update to TCJA law), CBO projected that corporate profits (the growth factor I'm using for earnings) would increase from $2.088 trillion in 2015 to $2.725 trillion in 2027, an increase of 30.5%. They forecast corporate tax revenue to increase from $344 billion in 2015 to $443 billion in 2027, an increase of 28.8%.
My new version forecasts EBITDA to increase from $2.415 trillion in 2015 to $3.152 billion in 2027, an increase of 30.5% (by definition, as I use now CBO's profits forecast to increase EBITDA). However, the new forecast only has corporate tax revenue increasing from $321.2 billion in 2015 to $329.8 billion in 2027, an increase of 2.7%. Upon further investigation, this occurs because two components of the corporate income tax, specifically the net interest deduction (NID) and the foreign tax credit, grow much more quickly over this period than other components (122.4% growth for the NID and 59.0% growth for the FTC). The very large growth rate for the NID occurs because the capital stock increases but primarily because interest rates are forecast to almost double during this period. The FTC grows so quickly because CBO forecasts corporate foreign profits to grow much more quickly than domestic or overall profits.
Subsequent work should endeavor to adjust the growth factor for EBITDA to accommodate the role of NID growth (as EBITDA precedes NID but the CBO forecast for corporate profits comes after subtracting off the NID).
@martinholmer @andersonfrailey @hdoupe @Abraham-Leventhal @lucassz
Because this has more substantive changes, I will hold off from merging for a few hours at least.