what's the threshold?

from a post i sent to an sociology list...

I have been reading about the 94 incidents of prisoner abuse and how these do not indicate a “systemic” problem. I find this very interesting as I wonder what is the threshold for # of incidents that WOULD trigger characterization as a systemic problem? Needless to say, their current characterization of it NOT being systemic doesn’t make sense to me. Despite the actual # of persons involved in the abuse, it was apparent in the Taguba report (and I believe he said so himself) that at least Abu Ghraib was a systemic problem in that it was a breakdown of the system which allowed the service personnel to commit the abuses.

I also recall what W. Edwards Deming (“father” of TQM) posited as a truism…when you have a failing in an organization, 10% of the time it is due to individual failing, and 90% of the time it is due to the system failing. Ironically, when I first started looking into TQM some years ago, some of the best info on its utilization came from the military.