A substantial number of variables are partially correlated with county income growth in Pennsylvania in at least one regression. I examine the entire distribution of the estimates of the coefficients of possible explanatory variables and find that county in Pennsylvania has probably experienced rapid per capita income growth since 1970 if it had a college educated labor force, agricultural income was a large share of county total earnings, and its population was largely urban. Since 1970, income per capita probably grew slowly, or not at all, if the county's population grew relatively rapidly and voters were mostly registered Democrats.