My sense in teaching over the past four years is that I have been assessing higher grades. (I shy from the term "giving grades" as it sounds like a gift based on character or my fondness for the student.) Beyond an anecdotal report on what the department median grade is (for which I appear to be one half letter grade above), I have no other information for the grading distributions in other classes in my department or at NYU, including other sections of the classes I teach. So, my philosophy is to tell students that if everyone performed excellently, that would be accordingly reflected. I then remind students frequently of how I evaluate their work, based on the departmental criteria, and at the beginning of the course provide exemplars of what I consider to be excellent work.
If there is an improvement over the initial semester, this doesn't surprise me in that I feel like my classes are now more honed, with exemplars students have a better sense of my expectations, and I've debugged assignment specifications. I also feel that while the material and assignments in the Media, Technology, and Society (MTS) class are more difficult than Conflict Management (CM), the students are more consistent. So, I performed a five number summary and generated the following box plots (with outliers below 70 truncated):
My conclusion is that while I assessed lower grades in my first semester of teaching each course, there is otherwise no consistent trend. Also, my sense of the MTS students being more consistent in performance is confirmed.