OK, I understand it is your opinion. I'd say thought the often quoted Pareto principle is "80% of the problems come from 20% of the causes", so that would imply counts.
If we go to the source, allegedly Wilfredo Pareto's idea was that in 19th century Italy, 80% of the wealth was controlled by 20% of the population. The idea was co-opted by Juran to 80% of money lost is from 20% of the problems; the "significant few." Even in your characterization of the Pareto Principle as "80% of the
problems...20% of the
causes," it isn't useful information unless we deal with it in terms of cost.
Even if you are limiting the use to dollars, I'd still check that the costs were stable prior to Pareto charting.
A Pareto chart is a "snapshot" of a period of time when things happen, bit it stands to reason to have as much information as possible before making decisions. If you construct two Pareto charts, one today and another next week, and there are significant differences, you know that you shouldn't be taking action until things settle down (when some measure of stability is present). That doesn't mean that you shouldn't do the charts, though. There's a difference between making a Pareto chart and acting on what it's telling you. This idea is fundamentally different from the discussion on histograms, where there should be underlying evidence of stability (Shewhart charts, e.g.) before a histogram is cast.
2000;3000;4000)
