Do distributional preferences reverse on a dollar? An experiment

In settings where other-regarding motives are likely to be (and some would argue, should be) at the forefront of our minds, how much of our behavior can still be explained by narrow pecuniary self-interest by itself? In an experiment where subjects are asked to vote between two income distributions that have diametrically opposed effects on the group as a whole, I find that self-interest still appears to dwarf the combined effects of other-regarding motives in influencing the votes of the vast majority of subjects.