In particular, if you're speaking about his 2015 article that came out, you're exactly right: the differences stem mostly from methodology and a little bit from scope. It's our understanding that we're using a formulary, whereas his 2015 article didn't use a particular formulary. It was even more encompassing.
Beyond that, it was the assumptions around, for example, an increase in volume due to lower costs to the patient. Generic substitution effects in our database were targeted based on what we observed nationally, whereas in Professor Morgan's report he used rates observed in public plans and targeted a specific rate. The prices themselves as well were targeting other nations, as opposed to using a 25% price discount, and they varied by the drug class.
Finally, I believe that we were able to calculate—more or less—the pharmacists' fees and the markups together, whereas Professor Morgan was only able to estimate them and try to guess at what they were.