> Yes, in positivist sciences 20% intending to stay would be very high by historical standards.

i'd be interested in a source for this. i did not find in the article you cite mention of historical trends.