As you may or may not know, health insurance can be really expensive to buy individually. To compete with other businesses for the best workers, companies often provide

their employees health insurance as part of their compensation package. But this is costly for

companies to do and they$$re always looking for ways to reduce those costs. A some$-$ what

common proposal $($seen here at UHart actually$)$ isto provide employees with highly subsidized

or free gym membership$/$exercise classes to encourage workers to be healthier which should

reduce some of their medical care costs. Say you want to study whether this strategy actually

works. You get data from a company who does this and you estimate a regression where

annualspending on healthcare is a function of whether someone used

the gym$/$workout classes there or yi $=\backslash$beta $$

$0+\backslash$beta $$

$1$x$1$i $+\backslash$epsi $.$