Keep in mind the former is an index—a numerical representation of categorical data—and therefore quite subjective. It is therefore more important to find experts whose judgment you trust in this matter, whereas cost of living is purely quantitative and can be trusted to anyone who can use Excel. There are a couple of studies I threw out because they ranked Michigan’s roads better than Minnesota’s, which is laughably untrue.
It appears these are inversely related, with high cost of living also predicting bad roads.
Road conditions are, to my mind, an excellent indicator of existing social capital (in the state’s recent history). Building and maintaining infrastructure requires a reasonably non-tribal government (concerned more with administration than identitarian demagoguery and spoils) as well as a large working class that is reasonably functional.
Cost of living is a pretty good indicator of cosmopolitan social cachet. Areas with low cost of living tend to be low social class with a lot of working class whites effectively integrated with African Americans. Areas with high cost of living tend to be classier where minorities are more mixed and there are more Hispanics who speak very little English, who furnish the labor force. These areas may have more diversity in some cases but they are more effectively segregated, having stark lines between the classes (hence “classy”). It is inversely related to religiosity, so that states where working class people go to church also have a lower cost of living, lower class, and better social integration.
I expect that this inverse relationship is the opposite of what is found at the national level, where borders much more effectively delineate tribal boundaries rather than mere preferences.