I think there are interesting things to say about brain drain, and I'm certainly under the impression that there is a tendency for people to move away from the middle of the country and towards the coastal cities (and apparently Texas, Colorado, and Illinois).
But I find the choices of units for these statistics to be very odd. For example, I look at Massachusetts Net difference in Absolute terms in 2017 (where stars indicate an option on the graph) and I see -16.41.
The article doesn't seem to say directly what -16.41 is, other than that it's the "excess of highly-educated share of movers over highly-educated share of entrants". So what is it?
Looking at the data they provide, I believe that they are subtracting two different percentages of two separate numbers. It appears they examine all the people who left and take the percentage of highly educated people in that group (call this X); and they examine all the people who entered the state and take the percentage of highly educated people (call this Y); and they return X - Y.
This appears to me to be a terrible statistic in practice, especially if the population of a state is either growing or shrinking.
For an example of an unbelievably misleading scenario, consider the fictional state Hackernewsyork, with a population of 100 people (30 of whom are highly educated). Say Hackernewsyork had exactly 1 person move away and she was highly educated (so X = 100) and 10 people move in but only 5 were highly educated (so Y = 50). Then they would report that X - Y = 50, a sign of massive brain drain per the article --- when in fact both the number of highly educated people in the state and the percentage of people in the state who are highly educated went up.