its just the beginning of a model. It needs work and it needs changes. This is the process that people go through when developing economic theory, what we are doing, and it takes a while to consider the options, consider alternate ways of producing better effects, etc.
By and large, your idea gave me an idea and the two theories just need to be justified or combined in a way that meets all criteria.
When I saw your formula I hesitated for a day to answer because I know the answer isn't just one or two posts... its a very elaborate exchange of ideas and tinkering back and forth.
Can you tinker with my model and get it closer to yours? Show me how to simplify your formula and mine so that they can be compatible... tell me what I need to ditch to get to where you want to be.
First, stop adding things that don't add.
I started this way -- asking what the variables are. The target is to differentiate tax rates between domestic, neighbors/NAFTA, and foreign/father away, so (starting over) designate those by D, N, F. Since what we're looking at is where the workers are, these become %D, %N, %F, for "percentage of workers located in". Then we want different rates for there, so H% (home percent), M% (to go with N), and O% (overseas).
The calculation is simple: what they pay for operations at home plus what they pay for neighboring/NAFTA operations plus what they pay for foreign operations; thus %D*H% + %N*M% + %F*O%.
Now how you set the values is a different question. Since the idea is to reward companies for keeping jobs here, H would be low, and the other two higher. But they don't have to be constants; functions would thoroughly baffle McCain (though as a naval aviator, he had to take calculus). So for the value of O%....
The time zone idea is attractive, but it only works one direction really. So let's standardize it: it's 360 degrees around the world, 24 time zones, so there's 15 degrees in a time zone. Set that as the standard: units of 15 degrees around the globe from the U.S., so it doesn't matter that Peru is in the same time zone as the western U.S., but that it's around 45 degrees from the closest point in the U.S. -- three separations, not zero.
Though at this point, there's no reason not to make it such large units -- just use degrees of separation (
). If I'm guestimating right on the cheap-as map I found online, Lima is 49 degrees from Houston -- using the closest large city rather than merely the border ~~ port of entry would be another point possible, but I don't have a map of those.
The question then is how to use that. Standard procedure with such things when modeling wildlife or geology is to turn the number into a decimal fraction of 1, to use as a direct modifier on whatever your present total is. The obvious move would be to divide by 360, but that's misleading, because 180 degrees is as far as one can get from the U.S. Hang on, though; the U.S. spans four time zones one way and the equivalent of two the other (neglecting Alaska and Hawaii; I'm working with just the contiguous states). Now we have two choices: we can treat the U.S. as a mathematical point, or we can take some degrees out of the equation. I'm for the second, and take off that 30 degrees north-south, so we have a total maximum 'distance' of 150 degrees. So to get the coefficient for company x, we take its distance in degrees over 150, giving a number from zero to one.
Then use that to make whatever tax rate you want.
To put that in the equation:
%D*H% + %N*M% + %F*f(d)%, where f(d) is a function of distance.
That's enough stretching the brain for now.
