This must be a standard problem in utilitarian theory, but I don't know its name.

In case you haven't read my comment introducing myself, know that my ultimate social value is freedom, a sort of utilitarian calculus where utility is freedom. So to judge whether someone should live, the main question to ask is whether they want to live. (I forgot to say in my reply to MartinB that of course I am against medical treatment of those who do not wish it.)

But those who do not exist do not wish anything. So it doesn't matter.

If by ‘a great life’ you mean a life of great freedom, then I prefer that to the alternative life. But one can only judge what such a life actually is once the person actually exists and has wants. I support prenatal care only on the basis of a prediction about what people will want later, like wanting to be healthy.

It still doesn't hang together mathematically, since I should simply take expected utitlity/freedom. As I also said in my introductory comment, I don't really believe that any utilitarian calculus captures my values. I can understand decision theory once the utilities are assigned, but I don't understand how to assign utilities in the first place.