The discussion section is an analysis of the data; it should form the bulk of your essay. Case studies, provided they are developed in sufficient depth, are invariably useful. There is often merit in selecting a few, particularly apt case studies, rather than trying to include as many as possible. It is good (usually essential) to present quantitative data to support your argument.
Don’t be afraid of diagrams and tables if you feel they:
a) would save you from a great deal of writing, and
b) are something you are confident you could produce in an exam.
You should cast a critical eye over all the data you present: for example, what are the limitations of the data? Is it up-to-date? Are there competing theories?