Last Saturday, at the Spatial Econometrics Conference 2007, James Lesage presented a paper showing that everybody misunderstood the meaning of coefficients in spatial regressions. And sometimes the difference is huge. It was an excellent presentation and you could almost hear the audience thinking: "Ay caramba, I have to rewrite everything...".
Only the abstract of the paper is available on-line. Stay tuned.
(Luckily, there were no spatial regressions in the paper (co-authored with Martin Brauch) that I presented there.)