Anthrax guy was a special case since his terrorist attack was done to generate more money for his anthrax vaccation program and keep his job. I thought the anthrax scare of 2001 was very famous, so I didn't see any need to attach a 'motive'. If I was to assign a motive, I would call it "False Flag".
Of course, if you really want to get interesting, my opinion is that you ought to try deeper analysis than this. Another set of graphs, Method of Attacks by Motive: Bombings by Motive, Arson by motive, Shooting by motive, etc. At least that way, the issue of conflation of relatively minor seeming attacks with relatively major attacks is avoided.
I wouldn't recommend that. The data set is just too small. We have about 45 data points and several different sub categories. Any real statistical analysis trying to deal with the sub categories laid out here would find it's conclusions all fail to overcome even a very moderate statistical significance threshold.
What if I instead look at all terrorist attacks that occured throughout the United States starting from the earliest possible date of when the Global Terrorist Database started collecting data? If I classify it on the basis of attacks based on methods
and motives, we could determine if "terrorism" in the US is indeed very serious (sniper attacks), or not so much (spraypainting). In this graph, I'll also classify "buytic acid" as a terrorist attack, if only because the GTD view it as such, as well as to avoid charges of "artibariness".
The conflation between minor attacks and major attacks seem to be the biggest flaw with my study, and looking at methods would help to avoid that.
EDIT: I'll also see if I can post a new post-9/11 graph with all the pie sections labeled, like the pre-9/11 graph.