Interesting perspective, and not one you hear about often: http://hbswk.hbs.edu/item/6420.html?wknews=052410
.
Although the title uses the words “Stimulus Surprise,” the only thing surprising about it, IMO, is that anyone who knows anything about economics would find it particularly surprising. The result makes perfect sense in the light of the Law of Displacement, as well as considerations of the effects of competition for factors of production and of general uncertainty engendered by changes in the interventionary climate.
I think you’re missing what was supposed to be the surprise. The surprise is that someone reported on it 
Right or not, I’m not impressed with the word “cause” in the title, “Do Politicians Cause Corporate Downsizing?” Because in a strictly statistical sense, the authors did not even attempt to prove causation. They only proved correlation. So they can say the results are consistent with various conjectures as to the cause, but they cannot say that politicians caused anything at all. This is clearly why they included a question mark at the end of the title, which is a bit of linguistic subterfuge usually associated with tabloid journalism or Fox News.
Before I’m attacked by the righties, remember that it’s this distinction between correlation and causation that is the foundation of all CO2 climate denier logic.
Since (for reasons I’ve outlined here in the past) I question the validity of that kind of empirical studies in the first place, I tend to agree with you about that. In order to determine which relationships are causation and which ones are mere correlation, you really need a different approach to economics–specifically, one that infers necessary consequences from much more fundamental principles, where the fundamental principles draw from our common experience as human beings.