Consider a regression equation for the health of an individual:

where health is some quantitative measure of a person's health; age, weight, height, and male are self-explanatory; work is weekly hours worked; and exercise is the hours of exercise per week.

1. Why might you be concerned about the estimated effect of exercise on health being subject to reverse causation?
2. Suppose you can collect data on the distance between each person's home and the nearest gym (disthome). Discuss whether this variable is likely to meet the conditions for a valid instrumental variable for exercise. If you think that disthome does not meet the requirements for a valid instrumental variable, think of an alternative.