Social scientists have found that while some group members will leave after the date for a doomsday prediction by the leader has passed uneventfully, others actually feel their belief and commitment to the group strengthened. Often when a group's doomsday prophecies or predictions fail to come true, the group leader will simply set a new date for impending doom, or predict a different type of catastrophe on a different date.Niederhoffer and Kenner say: "When you have gone far out on a limb and so many people have followed you, and there is much "sunk cost," as economists would say, it is difficult to admit you have been wrong."
In Experiments With People: Revelations from Social Psychology, Abelson, Frey and Gregg explain this further: "..continuing to proselytize on behalf of a doomsday cult whose prophecies have been disconfirmed, although it makes little logical sense, makes plenty of psychological sense if people have already spent months proselytizing on the cult's behalf. Persevering allows them to avoid the embarrassment of how wrong they were in the first place."The common-held belief in a catastrophic event occurring on a future date can have the effect of ingraining followers with a sense of uniqueness and purpose. In addition, after a failed prophecy members may attempt to explain the outcome through rationalization and dissonance reduction.
Explanations may include stating that the group members had misinterpreted the leader's original plan, that the cataclysmic event itself had been postponed to a later date by the leader, or that the activities of the group itself had forestalled disaster. In the case of the Festinger study, when the prophecy of a cataclysmic flood was proved false, the members pronounced that their faith in God had prevented the event. They then proceeded to attempt to convert new members with renewed strength.
http://en.wikipedia.org/wiki/Doomsday_cult