I thoroughly enjoyed this Financial Times article on about Jerry Morris, one of the first scientists to pick up a positive correlation between health and exercise.  Apart from the enjoyable reportage, the short biography, etc, what struck me was the simplicity in the setup for which Morris was able to detect a pattern.  I will describe it slightly differently, just to show how ideally it is suited to a quasi-experimental approach.

Take a group of a few thousand people of roughly the same socio-economic status. Assign about one half of the group to one job, the other half to another on the basis of some random criterion.  Observe that members of the one group live longer than the members of the other seemingly as a consequence of doing the one job and not the other.  Try to interrogate the differences between the jobs, isolate this difference, and voilà.  So this is what Morris had: take the two people who work on buses in England during the 40s: a bus conductor and a bus driver.  Both men (and they were all men I believe) were from the same socio-economic background (class) and basically didn't require any particularly different qualifications - driving itself being quite common.  But, for some reason, the risk of heart attack for a driver was roughly double that of a conductor. What, then, separates the activities of a conductor from a driver to explain this? A conductor walks all the time and climbs between 500 and 750 steps during his shift every day, whereas a bus driver remains sedentary. This was the only factor that could legitimately explain the differences in outcome and was set up so well as to be an experiment, what we call a 'natural experiment' or a 'quasi-experimental' setup.

So, as an empiricist, the article appealed to me on grounds factual, biographical, and technical.  What an intriguing tale.  Well done to Kuper for the entertaining read.  Well done to Morris for establishing the connection between vigorous exercise and health. Coincidentally, I also support Morris's idea that exercise needs to be made more easily accessible - pedestrianised areas to encourage walking, bicycle paths to encourage cycling, accessible and inexpensive swimming pools for swimming, that champion of full-body exercises, and, I believe, accessible public transport to and from which one can walk and access resources like parks, pools and gyms.  We shall see.