Jiang and VanderWeele respond to "bounding natural direct and indirect effects".