My guess is that having children is far less cost-effective at getting more EAs than university community building, because of far greater reach with the same time and resources from the latter, and not far worse individual counterfactuals. To elaborate on the last point, it’s true that someone who’s never born would not become an EA anyway, but if someone isn’t interested in EA by the time they leave university, it’s not overwhelmingly likely that if they’d have become an EA through university community building, then they would have gotten into EA eventually anyway (and that the gap in time to becoming an EA is small enough in expectation).
My guess is that having children is far less cost-effective at getting more EAs than university community building, because of far greater reach with the same time and resources from the latter, and not far worse individual counterfactuals. To elaborate on the last point, it’s true that someone who’s never born would not become an EA anyway, but if someone isn’t interested in EA by the time they leave university, it’s not overwhelmingly likely that if they’d have become an EA through university community building, then they would have gotten into EA eventually anyway (and that the gap in time to becoming an EA is small enough in expectation).