I like the intent of this chart, but the Y-axis is a mess. Combining percent change from 2007 (as units for the y-axis) with roll-overs of actual numbers that c

I like the intent of this chart, but the Y-axis is a mess. Combining percent change from 2007 (as units for the y-axis) with roll-overs of actual numbers that contain data pre-2007 is very unintuitive.