During the 2016 campaign, I explored a new way of aggregating – or maybe the right word is disaggregating – horse race polls for the presidential race. One of the rationales was that I wanted to keep an eye on the fieldwork dates (when the interviews for a poll are conducted) for the new polls when major campaign events were used to confirm or deny a major swing in the race.
I thought a similar strategy might actually be useful for looking at Gallup’s tracking data for Donald Trump’s prediential job approval. They do a 3-day rolling average with around 500 interviews conducted daily, and they represent it in the traditional way with a new datapoint for every average.
That’s fine, but in order to capture the actual days included in each average I made the graphic below. This format is also fairly easy to annotate with major events. As you can see, today’s is the first daily average to include only interviews conducted after the immigration executive orders (aka the #MuslimBan) announced last Friday.
His net approval is at -8, or 43% approve to 51% disapprove. Tired of winning yet?