These points indicate some of the possible reasons for estimated task durations being shorter than practice.
Clearly, if the estimator is over confident, does not appreciated the complexity of the situation, is unsure of the deliverables or makes inappropriate assumptions he may well underestimate the time required to complete a task.
This is probably a greater problem than over estimation.
The project will take longer than expected and consume more resource and finances as a result which were not allowed for.
In similar fashion to overestimation the use of a group of suitably experienced people will help get the estimation right.
Not understanding the complexity of the task is major a factor in being over confident in estimating durations.
This could be a lack of experience or poor (or non-existent) training procedures.
Some times a desire to please the boss and to seem a superior worker makes people go for faster task completion.
Again, this can be a culture problem.
If the boss makes people feel incompetent personnel may try to show they can complete tasks in unrealistic times.
This is an aspect of leadership which is covered in more detail in 'The Complete Leadership package'.
Again, poor communication and understanding of what must be produced can give rise to poor estimation.
Ensure that the information is available and that the workforce understand it.
If people are unaware of the latest assumptions and constraints, or any new developments, estimation problems can result.
Managers are all too aware of deadlines for a project which may be totally unrealistic.
There may be a deadline to prepare the necessary information to gain project approval.
If this is too tight, not enough time is spent on the planning and inaccurate estimates may result.
This could be a major problem for contractors preparing bids.
It is not unusual to hear of major projects coming in at 4 or 5 times their original cost estimates, as well as being late!
There is a temptation to cut task durations anyway to meet deadlines without any real justification.
Some tasks cannot be done any faster no matter how much time is spent on them.
It may be appropriate to use a formal estimation procedure to help provide consistency in approach and hopefully more accuracy in estimation.
If you are a manager don’t inflate estimates from lower levels in the Work Breakdown structure.
This can have an accumulative effect.
At times the Project Manager or department head must defend decisions on task durations.
There is little point in saying a task will take 3 weeks, just to look good, when it is known to take a 4 weeks.