Nice! I like that we are clear on the disagreement now. 

Let me substantiate my point then with a couple of examples. 

- If you ought to plant an apple tree, it follows that you ought to plant a tree. 

- If you ought to donate to GiveWell, it follows that you ought to donate to charity. 

And so on. 

Whenever you ought to do an a specific action of kind A, it follows that you ought to do an action of kind A. (This follows by existential generalization, if you want to go down into the symbolic logic of the argument.) 

Furthermore, it can be true that you ought to do an action of kind A, even though, for some specific action t of kind A, if is not true that you ought to to t. 

For example:  

• It can be true that you ought to teach your children manners, even if it is not true that you ought to physically punish your children until they learn manners.
• It can be true that you ought to bring your mother a gift for her birthday, even if it is not true that you ought to give her a Ferrari for her birthday. 

And so on. 

... That's why I don't agree with your last point "I believe for this to hold you would need to know that [CAFOs < 0] is impossible, not just that [CAFOs > 0] is possible."

Hi Clara, 

The logical shape of my full argument is this: 

If [we ought to maximize net aggregate welfare] then [we should build more CAFOs of the kind in which animals have above 0 welfare].  

I also hold that:

If [we should build more CAFOs of the kind in which animals have above 0 welfare], then [we should build more CAFOs], since we cannot do the former without doing the latter. 

Provided that if-then is transitive, it follows that: 

If [we ought to maximize net aggregate welfare], then [we should build more CAFOs].

For these reasons, I continue to believe that the logic of the abstract is sound. 

As I said, I can see that stylistic preferences could draw one towards wanting to make the difference between CAFOs in which animals have net positive welfare and CAFOs in which they don't explicit in the abstract.

Hi Clara, 

I appreciate that there are different writing styles and that there certainly are other good ways to write the abstract and the conclusion of this article. 
 
However, to clarify, the claim that net positive welfare is possible in CAFOs is not an assumption, but a premise in my argument that I provide evidence for. It's not a norm in academic journals to discuss every premise of an argument in both the abstract and the conclusion of an article. That would defeat the purpose of these sections, which are meant to provide brief overviews with tight word limits. 

I agree with you that, if the conclusion was conditional on an unargued-for assumption, this should be highlighted prominently.

Thanks for your feedback, Clara! Not quite. My argument rests on the claim that it is possible for animals in CAFOs to have net positive welfare (not that they actually do). This creates an empirical and practical question for people wanting to maximize net aggregate welfare: figure out which exact living conditions ensure net positive welfare, and, based on that, which exact living conditions allow us to maximize net aggregate welfare. I have a long section in the paper were I discuss empirical research in that area, in particular the excellent work by Cynthia Schuck-Paim and her team  at the Welfare Footprint Institute. As far as I can tell, with respect to layer hens, there is little evidence that lives in cage-free CAFOs would have net negative welfare.