Productivity study: worrying about what a dollar buys

I'm uneasy about the cost-of-living adjustments in the new productivity study from the Center for American Progress. The study looks important, but before I highlight some interesting results, I want to flag this possible concern.

When the study adjusts for cost-of-living differences, it assumes that a dollar buys a lot more in Breathitt, Knott, Lee, Leslie, Letcher, Perry, Owsley, and Wolfe counties. The study treats each dollar spent there as worth $1.39 elsewhere.

 That's the biggest upward adjustment, and then there's a sliding scale of smaller adjustments showing most of the state being able to get more for a dollar than the national norm. Only a Northern Kentucky cluster in Boone, Bracken, Campbell, Gallatin, Grant, Kenton, and Pendleton have their purchasing power downgraded to show a dollar as worth 98¢.

Some costs certainly are lower in eastern Kentucky, but I'm still having a hard time thinking the proportions are right. Especially, I'm having a hard time imagining that newly-minted teachers will believe that a starting salary of  $21,583 in Wolfe County is as good a deal as $30,000 somewhere else.

I can't say with certainty that there's a mistake here, and I can't offer an alternative method, so I just want to put my uneasiness out on the table before commenting on some of the other findings.

Source note: To figure out the cost-of-living adjustment, I downloaded the complete data file offered here, and divided the "Cost of Wage Adjusted PPE" column by the original "Per Pupil Expenditure" column.