“…In the south, counties with a higher percentage of non-white population have traditionally suffered from low infrastructure investment (new roads, water systems, etc. ), and percent black in a county serves as a good proxy measure for infrastructure conditions (see Lyson 1989;Singlemann et al 1993;Talley and Cotton 1993;). Therefore, regression models will contain a binary regional variable to test differences on the effects of highway spending on south or non-south counties (SOUTH).…”