Now those who have read me in the past know that I am skeptical about league tables - they are notoriously sensitive to ex ante model specification problems (in other words, they are easily rigged for the outcome one wants). That being said, this is the first serious, refereed journal article on the topic that I've seen and it seems to be very credible.
It also supports my preconceived opinions, which doesn't hurt.
No comments:
Post a Comment