I'm curious. What backs up the idea that the average number of children per individual will be 2 regardless of reproductive lifespan?

It seems that I've seen a somewhat(?) common pattern among people who have young families and get divorced that often times have another family (+2 children) when they re-marry.