Throughout the years the office buildings I have worked in generally practiced regular fire drills.  Sometimes you would get an advance notice (at least the managers would) and sometimes not.  The purpose was to make sure the occupants of the building could exit in an orderly manner following an alarm.

On the surface this seems like a good practice, however,  I would like to explore the value of this practice a little further.

Whats the impact and likelihood.

The risk event  (e.g., delayed exit during a fire, building collapse, attack by a jet airplane ) certainly has the potential for a high impact.    A delay or being trampled could result in death or serious injury.  For most of these risks though there are a number of mitigating factors that reduce the risk from occurring in the first place.  Things like regular inspections of buildings for fire hazards, early warning, special engineering for earthquake impact reduction and fire suppression devices are just among a few of these mitigating processes.   As a general rule, I would argue the likelihood of the risk event resulting in the need to exit rapidly is reduced to low by the various mitigating factors employed by building owners (recognizing some more than others, but building codes require a certain minimum level of mitigation.

So based on the assessment of high impact versus low likelihood how valuable is the pre-risk event mitigation of a fire drill.  One way to view this question is to consider the effectiveness of a fire drill.

My Experiences with fire in large buildings

I have only responded to three real fire alarms in my lifetime and all occurred in buildings where I had not participated in a fire drill  (two were hotels and one was an apartment building I lived in).  In all cases the building occupants exited rapidly, orderly and without incident.  Also, I can say with confidence the majority  (like me) had never participated in a fire drill in those buildings.

Unnecessary Risk Mitigation

Considering the effort, loss of productive work time and minimal risk mitigation value I am not a fan of fire drills.  Even without a practice drill the likelihood of someone not being able to exit or a mad panic occurring which delays the ability to exit is low in my mind.  And the impact would probably be quickly resolved.

And my final point would be the places where rapid exits from a building has resulted in people being trampled or injured usually occur at events (concerts, dances, etc) and generally none of these places (where the risk event has actually occurred in the past) conduct drills.  This of course is mainly because it is not practical.