I guess it'd depend on how this is counted. If it is simply "total volume of gallons of water consumed divided by permanently resident population" then your suggestion probably explains the bulk of it. I would have hoped that it'd be measured slightly smarter however.