As he says, it’s just a first cut, based on two things: household income and travel time to work.
His reasoning is that as price of oil increases, those hit hardest will be those who have the furthest to commute, and the lowest disposable incomes.
Seems reasonable as a first approximation.
The picture he comes up with is shown on the right. (Click to enlarge.)
There are more explanations and more pictures, here: http://earlywarn.blogspot.com/2010/06/peak-oil-stress-map.html