Barry Pump shows that state level union density is uncorrelated with progressive taxation but reasonably well correlated with social welfare spending:
It would be interesting to know more about this. There’s a difference, after all, between how generous a program is and how expensive it is. One program might hand out money to poor people. Another program might hand out money to poor people if and only if they pass regular drug tests and criminal background checks. The latter will likely be more expensive, since it’s harder to administer, but the former seems more generous by any reasonable standard. Unions have reasons to lobby for generous, efficient programs (like Social Security, which labor is a key backer of) but also have reasons to lobby for making programs labor-intensive and letting service providers capture as large as possible a share of the appropriations. It would be interesting to see evidence on which effect dominates. Maybe a chart with “personnel costs as a share of state expenditures” on the y axis would be illuminating.
It also might be interesting to see this done with fewer states. The chart as it exists seems to me to be largely showing that southern states have both “right to work” laws and low levels of spending, which seems like a clear case of joint causation by the underlying malady of right-wing political culture.