I find it quite surprising.
Sure being a pot head is going to have a detrimental effect on your grades.
But given my experiences with university in a place where marijuana was not legal I can't believe there are enough students who would not smoke when it is illegal but would when it is legal to swing the overall grade by 5%.
So in the article the authors stated the following:
“The effects we find are large, consistent and statistically very significant,” Marie told the Observer. “For example, we estimate that students who were no longer able to buy cannabis legally were 5% more likely to pass courses. The grade improvement this represents is about the same as having a qualified teacher and, more relevantly, similar to decreases in grades observed from reaching legal drinking age in the US.”
For low performers, there was a larger effect on grades. They had a 7.6% better chance of passing their courses.
Note that they don't seem to be saying that pot smokers in specific are affected, but rather that any student who was legally barred from buying pot got a 5% better chance of passing. I can't imagine the total number of students who were buying legal pot before and stopped buying after was very large.
Note they got their study populations by looking at a city where they banned most foreign students from buying pot for several years:
Economists Olivier Marie of Maastricht University and Ulf Zölitz of IZA Bonn examined what happened in Maastricht in 2011 when the Dutch city allowed only Dutch, German and Belgian passport-holders access to the 13 coffee shops where cannabis was sold. The temporary restrictions were introduced because of fears that nationals from other countries, chiefly France and Luxembourg, were visiting the city simply to smoke drugs, which would tarnish its genteel image.
After studying data on more than 54,000 course grades achieved by students from around the world who were enrolled at Maastricht University before and after the restrictions were introduced
So I looked around and I think I found the paper here. I haven't digested it but there's a table on page 33 that shows the average grade change over the study period. There's two things that strike me about this table.
First there's not a lot of data to convince me that big divergences in grades by nationality are abnormal. For all we know secondary factors cause quirks like this all the time.
Second, half way through the prohibition period the grades of the DGB students who can buy pot starts shooting up. By the end of the prohibition they're actually doing better relative to the non-DGB students then they were before prohibition!
Maybe they're onto something, but of the stuff I've seen I'm still really skeptical.