Saw this post on Greg Mankiw's blog just now. Here's a rephrase.
What Mankiw said: where we are, the Laffer curve is concave down.
What Bloomberg heard: where we are, the Laffer curve is decreasing.
Maybe I've just been tutoring too many calculus students recently, but this seemed a pretty humorous way to put the misunderstanding.
(Detail for math wonks: the Laffer curve in real life is certainly much more complicated than it is in a typical economics class, where it's derived from linear supply and demand curves. Mankiw says -- and I'd tend to usually agree -- that a marginal increase in taxes will result in less revenue than static models would tend to predict.
A static model is based on a first order approximation, i.e., extrapolating out on a tangent line. So Mankiw is claiming that the actual Laffer curve falls below the tangent line approximation, i.e., it is concave down.
Bloomberg misinterpreted Mankiw's claim, and thought he was repeating the nutcase claim that the current US tax rate is on the far side of the Laffer curve, i.e., that revenue would fall with an increase in taxes.
It all boils down to the difference between first and second derivatives.)