You are an economist looking to estimate a sickness production function.
a. What data would you need to construct the sickness production function?
b. What would be the unit of analysis?
c. Given that you cannot run an experiment, how would you get lucky to have some good variation in pollution to test whether it causes sickness?