TL:DR - Assuming utility functions, if you value animals above the meat they produce, the utility from paying people not to eat them and not eating them yourself is always greater than paying people to eat meat and eating it yourself.

Here is why a person following a utility function couldn't get out of fulfilling these types of moral responsibilities by paying for them -

Killing an animal [K utils]

Eating an animal [E utils]... and the cost of acquiring the animal is included in this

So, for a "vegetarian", the relevant information is that util losses from killing an animal outweigh util gains of eating one. In other words...

K+E<0

To want to eat meat, you need to make K+E greater than 0.

Now, you're essentially proposing paying some money in order to Save an animal. Let S be the utility generated from not Killing an animal, as a function of how much you money you Pay.

S = -PK-P

Now it is true that you can generate a net positive utility by Paying sufficient money to offset K+E, and that's why this intuitively seems like a good idea...

K+E+S > 0 if P is large enough...

But here is the problem: since K+E is a negative number, not killing and eating is always better than killing and eating. In math terms -

K+E+S < S

for all "vegetarians". Therefore, no utility-maximizing agent who is currently a vegetarian is going to choose to start eating meat and paying others not to eat it.

At best, if they value animal lives above all other causes they might donate to, they would continue not eating meat and also not pay others to eat it.

The only person who would eat meat and pay for others to not eat it would be a person who 1) valued meat more than animals 2) valued animals more than all other causes they could donate to. This would be a very small set.

Still - this approach does work for humans, for the purpose of guilt-alleviation. It would probably be a net good if people started doing it - it is akin to carbon offsets,.