As suggested by the EPA:

On a serious note, in the story, the farmer featured complained about how such a tax would hurt him financially. While that may be true, it should be irrelevant. If there is truly an externality here and it can be effectively measured and properly implemented, then there should be a tax on cows in the amount that is equivalent to the negative harm done per cow. That’s all that matters. It all hinges on the extent to which cows cause global warming and the extent to which global warming harms society.

I don’t know what kind of market cows are, but I would imagine it’s probably close to perfect competition suggesting that the tax would likely be borne by him, the producer rather than us, the consumers. They’re talking about $175/cow/year tax. On the other hand, this farm received $561,695 in federal subsidies from 1996-2006.

A good first step would be getting rid of these subsidies entirely, which would let the market correct itself inasmuch as it can before trying to internalize these negative externalities. It would make more sense to remove the existing distortion, see what happens in the market, and then layer on the Pigovian tax.

Let the number of cows produced decrease in a two-step process, rather than shocking the market all at once which would probably unfairly tarnish the idea of a Pigovian tax as a policy tool. (“Look what happened to the price of beef and dairy when we…!” etc.)

Via Greg Mankiw.