cial-airline pilot could dip the nose of the aircraft at a crucial point of
takeoff or could inexplicably land on an already occupied runway
against instructions from the control tower; the driver of a car could
suddenly swerve into a tree or into oncoming traffic; a passenger in an
automobile or corporate jet could incapacitate the operator, causing a
deadly crash; the pilot of a private plane could, despite all radio warn-
ings, plow into another aircraft. Thus the alarming climb in crash
fatalities we find following front-page suicides is, according to Dr.
Phillips, most likely due to the Werther effect secretly applied.
I consider this insight brilliant. First, it explains all of the data beauti-
fully. If these wrecks really are hidden instances of imitative suicide, it
makes sense that we should see an increase in the wrecks after suicide
stories appear. And it makes sense that the greatest rise in wrecks should
occur after the suicide stories that have been most widely publicized
and have, consequently, reached the most people. And it makes sense
that the number of crashes should jump appreciably only in those
geographical areas where the suicide stories were publicized. And it
even makes sense that single-victim suicides should lead only to single-
victim crashes, whereas multiple-victim suicide incidents should lead
only to multiple-victim crashes. Imitation is the key.
But there is a second valuable feature of Phillips’s insight. Not only
does it allow us to explain the existing facts, it also allows us to predict
new facts that had never been uncovered before. For example, if the
abnormally frequent crashes following publicized suicides are genuinely
due to imitative rather than accidental actions, they should be more
deadly as a result. That is, people trying to kill themselves will likely
arrange (with a foot on the accelerator instead of the brake, with the
nose of the plane down instead of up) for the impact to be as lethal as
possible. The consequence should be quick and sure death. When
Phillips examined the records to check on this prediction, he found that
the average number of people killed in a fatal crash of a commercial
airliner is more than three times greater if the crash happened one week
after a front-page suicide story than if it happened one week before. A
similar phenomenon can be found in traffic statistics, where there is
evidence for the deadly efficiency of postsuicide-story auto crashes.
Victims of fatal car wrecks that follow front-page suicide stories die
four times more quickly than normal.
Still another fascinating prediction flows from Phillips’s idea. If the
increase in wrecks following suicide stories truly represents a set of
copycat deaths, then the imitators should be most likely to copy the
suicides of people who are similar to them. The principle of social proof
states that we use information about how others have behaved to help
us determine proper conduct for ourselves. But as the dropped-wallet
112 / Influence