The planning fallacy is a well-documented tendency of persons to underestimate the amount of time which a task will take, quite often by a factor of a third or more. Numerous instances have been set forth in studies wherein a group of persons tasked with writing a report of a set length or writing a program of certain functionality have erred in their time-to-completion estimates by predictable ranges (and, naturally, the error rate increases as the task becomes one less familiar).

The standard explanation is that people are overestimating their working speed. And so, despite their studious attention to the task before them, the time taken stretches beyond that which is expected. But I suspect that in at least some appreciable subset of instances, the dynamic develops decidedly differently. Rather than devoting themselves to the task, the tasked person -- expecting at some subconscious level that the job to be done will encumber all their time for the period between its commencement and the deadline -- desires to just screw off and do as they please. And so, at some slightly more conscious level they rationalize to themself, 'the task will only take half the assigned time, so self, you can screw off and do as you please for the first half of the time.'

Simply put, it is possible that some people build in procrastination to their durational expectations, generating a sort of counter-self-fulfilling prophecy. For it may be the very act of believing things will turn out early which inspires procrastination which makes things end up turning out late. And if it turns out that the task only takes 75% as long as expected, instead of 50%, then even though the doer could have done it well within the time allotted (had they begun when they first could have), they end up going 25% past the deadline simply because they lopped their spare time off the wrong end of the span.

As an example, I began write this up during the 2011 Ironnoder.