Before delving into Azari’s conclusions on mandate rhetoric, her methodology is interesting. She analyzes a dataset of news conferences, television and radio addresses, minor addresses made during travel and at meetings of groups such as the Chamber of Commerce, messages to Congress, and various remarks and memoranda addressed to members of the executive branch, totaling 1,467 communications from 1933 to 2009. The important thing to note is that she confines her dataset to the first 70 days of each term of a president. She argues this controls for “exogenous events and facilitates comparison across presidential administrations.” The problem with this small size is that it limits the dataset to roughly 5 percent of a single presidential term.