> That's only true if the lane is as full of busses as other lanes are full of cars. 1) The density of people per unit road area is about 10 times greater than

> That's only true if the lane is as full of busses as other lanes are full of cars. 1) The density of people per unit road area is about 10 times greater than with cars. A bus lane only needs to carry 1/10th as many buses to come out ahead (and a bus occupies considerably less space than 10 cars). 2) The relevant metric isn't people (or vehicles) per unit road area. It's the lane's flux of people per unit time. On that metric, an free flowing carpool lane will beat a congested regular lane every time (even when the carpool lane appears empty to the layman). You can do your own back of the napkin calculation, but vehicles have to packed roughly 5 to 10 times denser in the congested lane to achieve the same vehicle flux per unit time. That means a free flowing carpool lane (>=2 people) only needs about 1/20th the vehicle density per unit area to come out ahead of a congested single-occupancy lane. It's just no contest with an uncongested bus lane: Buses will win on a people per time metric with only 1% as many vehicles per unit area. To visualize this, 100 closely packed cars (remember, they are not touching!) is about three city blocks long.