I think it's a fairly well-understood causal inference problem to figure out how much effort (money) needs to be spent on other people for GDP growth from that one person to offset the money spent on them by society.

We just need someone to step up and collect the data.