You're right that it's a loose extrapolation at best, but it's rather the best data we have, or at least the best data of which I'm aware. You asked for citations for my claims, and I provided them; I'm showing what happens when people _stop_ working early, not when people never really start or don't for long.

That aside, a society in which no one supports himself but rather is dependent on the noblesse oblige of the state to survive is a dystopia I hope never to see.