A persistent concern expressed over the years has been about the effect of migration upon the economy or development of the countries that the migrants leave. We see it now in the comments about how the NHS hiring doctors or nurses from Africa or India deprives those countries of doctors and nurses. A similar idea was behind Tony Benn's decades old suggestion that no graduate of an English university should be allowed to emigrate until they had "paid back" the cost of their education.
Aggregating these effects up, the effects of remittances, trade, investment, and repatriated savings typically more than offset the fiscal cost, and on top of this, the migrants themselves have seen enormous benefits from migration. In order for this high-skilled migration to have a net negative effect on development, one would therefore require very large externalities from the presence of these individuals. Based on current economic evidence of externalities, we instead estimate the externalities to be at most $800 per year – or only 2% of the private gains.
It seems that the interactions between those who have left and the place they have left are of more benefit than what they would have done had they stayed.
However, in another manner, it should never have been necessary to conduct such research in the first place. For by calculating such numbers we are implicity accepting that the needs of a society for a certain person or skill over ride the personal wishes and liberties of that person or possessor of that skill: as if we should act like some Chinese Emperor forbidding the porcelein makers to travel. That's an acceptance that we really shouldn't allow ourselves to be gulled into.
Either we believe in liberty, which clearly includes the liberty to leave the society of your birth, or we do not believe in liberty and thus society can be allowed to, well, in essence, hold you hostage for your skills. And as liberals that's absolutely not something which we should be willing to accept.