Flutterby™! : Can you trust an economist?

Uwe Weinhardt asks "Can economists be trusted?", in which he talks about how economists SIF (structure information felicitously) catch his additional response to user comments:

With the method called ordinary least squares estimation this economist was forced to reject that hypothesis; ditto with two-stage least square estimation. With maximum-likelhood estimation, however, he could infer from this data set a sizeable causal effect of insurance coverage on obesity — so large, in fact, that he wondered whether the effect could be this large.