Perhaps it's because I'm ... a bit too deep in the diffusion literature at the moment, but when I saw this post commenting on how brilliant it was to use typos and other errors to track the diffusion of legislation from Mexico to Venezuela, as well as Henry's followup here, all I could think was, "Old hat. Eldridge Dowell did this in 1939. Duh."
Just kidding. I mean, just kidding that this piece should be better known. But he really does use the errors. And if I had more time, I could track down the more recent and less totally obscure reference that's on the tip of my tongue. (Is it Berry and Berry? Or is it Walker 1968?)
(I should mention that I've learned a tremendous amount from Gilardi's work, both theoretical and practical. In fact, I should probably pay him royalties for some of "my" Stata code for the analysis of dyadic datasets.)
DOWELL, ELDRIDGE FOSTER 1939 A History of Criminal Syndicalism Legislation in the United States. Baltimore: Johns Hopkins University Studies in Historical and Political Science, Series 57, No. 1.
NB. I think the commenter at the original link misses the point. There is indeed a theoretical benefit to calling this diffusion. In fact, it's as pure a form of emulation as you're apt to find! And it reinforces the point made by Beth Simmmons and Zach Elkins about the different mechanisms that underlay diffusion mechanisms in the first place--namely, that there are real differences between diffusion that takes place because of learning, because of emulation, or because of coercion. The bigger question that the Venezuela case raises is, rather, about the state's capacity for "reinvention" or for "innovation" (depending on whether we prefer to think in Mohr's terms or Rogers's).