Calculating the Cost of Emigrating

Sometimes, economics just seems to break down. Here's George Borjas, on the economics of emigration, specifically from Puerto Rico to mainland US:

Between 30 to 40 percent of the Puerto Rican population chose to move out. But that means that about two-thirds did not. Why? If economic theory is right, it must be the case that there are huge costs associated with the migration. Since these costs are unlikely to be monetary in nature, they represent the fact that most people, if given the choice, would much rather stay where they are at. It is not hard to calculate the migration cost for the "marginal" migrant. And they are substantial: probably around $400,000 to $500,000.
(For those interested in technical details, the discounted gain from moving for the marginal migrant must equal the cost of migrating. The half-million dollar range comes about if one assumes that a worker can, say, double his salary by migrating to the United States and the rate of discount is 5 percent.)

I fail to see how this "migration cost" is in any sense a useful number, or that it reflects anything in the real world. Let's say that Peter Kann is right, and that financial journalists would make much more money if they became bankers. Using Borjas's technique, then, we can work out the cost, to the "marginal" journalist, of becoming a banker; my guess is that it's probably many millions of dollars.

What's more, that cost of becoming a banker has been soaring over the past few years. Back in the wake of the dot-com bust, when bankers' bonuses were much lower, the cost of becoming a banker was also relatively low. But now that bonuses look set to hit new record highs, the cost of becoming a banker is surely enormous.

This is silly. There aren't costs to becoming a banker, there are reasons why one might not want to be a banker. If you're happy as a journalist and you wouldn't be happy as a banker, then the amount of money that bankers make is not going to change anything.

Similarly, there aren't costs of half a million dollars or so associated with moving to mainland US from Puerto Rico. It's just something which a lot of Puerto Ricans have no interest in doing -- and given how nice their beaches are, you can see why.

What's more, by Borjas's calculations, if Puerto Rico's incomes caught up with those on the mainland, then the cost of moving would drop to zero. Which also makes no sense. Or am I missing something, here?