The samples for many of the individual states are small, so there's a lot of random error, but the differences are statistically significant (p=.005) and the general pattern seems reasonable. My earlier post looked at rankings on a question of how large a part politics played in the handling of relief in your community from the late 1930s, raising the question of whether the two rankings are correlated.
*There was a category for "neither/none (volunteered)", but it was very small. I treated it as missing, because I wasn't sure if "neither" should be interpreted as "none are corrupt" or "don't know."
[Data from the Roper Center for Public Opinion Research]