I read that, in the recent Connecticut primary election, Lamont did better in the richer towns and Lieberman did better in the poorer towns. But the exit poll showed little correlation between income and vote preference (see page 5 of this document). Putting these two facts together, I think this implies that, within towns, Lamont did better among poorer voters and Lieberman did better among richer voters.
I’d like to do more analysis (as in here and here) but I don’t have the poll data and so can just speculate.
P.S. Boris pointed out Mark Blumenthal’s comments here and here on exit polls and voting in Connecticut.
One random speculation:
Lieberman won among those with less education, Lamont among those with more. Lieberman also did better with older voters, Lamont better with younger. Several reports suggested a very high college student turnout as well.
Perhaps the low-income Lamont voters were disproportionately likely to be college and grad students. Students tend to be low income while young but high income later, and may have some views that are similar to high income educated people that they will likely become.
Of course, more analysis would be needed.
In any case, income is correlated with age, so it's possible that Lieberman did better among the poorer people of a certain age, but did better with people the older that they were, which canceled each other out. Possibly confounding this even more is that current college and grad students tend to be temporarily poorer than those who went to work straight out of high school or college (although they make up for it later) but more likely to vote for Lamont.
Yes, I agree that income is not a particularly important variable in Connecticut (as we discussed in our earlier paper, of course). I just thought that this between/within towns comparison was interesting since commentators seemed to be discussing it, and making the common aggregation error of attributing between-town differences to between-individual characteristics.