Taking this ‘study’ at face value and assuming its not a complete burning shit-pile of piss-poor methods, I wonder how much this ‘average’ is being affected by the upper-middle class and higher income brackets? Because mean averages can be skewed if there a uneven variations in the data.