Which of these is not like the other: craft beer, national-brand light lagers, liquor? In case you missed it, the Beeronomist reflected on the Modelo/InBev merger, and not in the vague, half-assed way I did. One of the most important--and largely unexamined--aspects he raises is this:
[You first] have to define the market. By all accounts the macro brewers are losing market share much faster to spirits than to craft beer, so shouldn't you include them when you are measuring the amount of competition? AB InBev will certainly argue that they should.
For example if a decrease int the price of vodka creates a sizable drop in Bud sales you could make a convincing argument that vodka is a real competitor to Bud and that vodka distillers should be counted as in the same market as ABInBev.Patrick doesn't say this directly, but implicitly raises this possibility: "craft" beer (broadly defined) may not actually be in the same market as Bud, Miller, and Coors--but those companies could be in a market with, say, vodka or tequila. I would love to see data on the kind of example he cites. In a way, I can see how this may actually be true--not just a legalistic excuse. The motives people have for drinking vodka seem to align more closely with the reasons people drink Bud--and differ from the reason people drink IPA.
There's more in Patrick's post and if you're at all interested in the subject, go have a look.