He cites Brooks as proving that communication friction is quadratic in headcount. The exponent does not need to be exactly, or even nearly, 2; 1 + epsilon is a

He cites Brooks as proving that communication friction is quadratic in headcount. The exponent does not need to be exactly, or even nearly, 2; 1 + epsilon is already fatal. This is very closely related to Coase's ceiling.