Reply #35: I haven't reasoned it all the way through, [View All]

[Igel]

but I think that the methodology isn't too bad, *if* you assume a random distribution of incomes *and* there's nothing skewing the curves. Given those assumptions, the average income and the average deductible donations should increase in lockstep. But I can't evaluate the assumptions' validity. I assume the researchers would have considered this--if you don't know your assumptions and can't defend them, you don't have any business venturing into statistics.

In states with very large, poor rural populations and wealthier urban centers, it won't work. Nor will it work if a significant portion of high-earners ("old money") finance their giving primarily through family-controlled foundations.

edited to add: With the caveats the researchers stated implied, of course.