One of the best testing grounds for the ‘tax migration’ theory is among individual states in the United States, which each levy variable state taxes (on top of nationwide Federal taxes,) and where cross-state migration is far easier than moving, say, to a different country with a different language and different laws and different people.
And here the evidence is unambiguous. Take this Stanford paper, for instance, which finds ‘negligible’ effects from a large state tax hike in New Jersey. Or this ITEP paper entitled “Where Have All of Maryland’s Millionaires Gone? (Nowhere – They’re Probably Just Not Millionaires Anymore.) Or this, on New York, or this, on Oregon. (From those links, get a load of that repeated Wall Street Journal hyper-ventilation, in the face of all the evidence). More generally, take a look at Citizens for Tax Justice’s Evidence Continues to Mount: State Taxes Don’t Cause Rich to Flee, which begins:
“There’s been a lot of good research these past few years debunking claims that state taxes – particularly income taxes on the rich – send wealthy taxpayers fleeing from “unfriendly” states.”
What they’re showing is that a change of a percentage point or two doesn’t change behaviour very much. For the real point is in the line “on top of nationwide Federal taxes”. that the State tax rate moves from, imagine, 5 % to 7% on top of the Federal rate of 35% or 39% isn’t, I think we’ll all agree, going to produce much effect on anyone.
This is rather different from say, the French move from 50% to 75%. Which is just one of the general problems with the TJN. They don’t seem to have caught up with the neo-classical revolution. You know, that 1870s, 1880s stuff, the marginalist revolution? Even if you don’t want to do that (and Ritchie has most amusingly, and repeatedly, claimed that all neo-classical economics is wrong) surely the idea that the size of the change in behaviour will be linked to the size of the change in incentives makes simple and common sense?