Useful post, thank you.

A long time ago I did some figures on actual movement of people inside the USA, and as I recall, it doesn't add up to anything like the numbers required by the OCA model.

To me this has two implications:

1. The other stuff (fiscal transfers, etc.) is even more important than people realise.

2. This matches up with what we see in the USA, which is that while nice graph looks nice (unemployment between 3% and 8% ? across different US states) there remain significant and persistent wealth/poverty disparities between states.

As usual, this concerns me much more than it seems to bother economists, because it indicates to me that the free market and free movement isn't proving much good at solving this problem...