Facts don't figure

Part of my job involves benchmarking of manhour numbers to determine what is called a People Index, which is used to compare the workforce population and demographics to other simlar facilities. Great concept, jsut like any other benchmarking effort, but I wonder really how accurate the benchmarking is. It seems that so many assumptions are made, that the essence of local factors that can and do have an influence on various parameters are lost in a sea of data. This is especially so when the comparing across different regions, where the dynamics of the economy are just not the same. I wonder how "normal" normalised data is, especially if the parties doing the normalising have no clue about the backrgound of a particular site? Hmmm..... I wonder.