One good explanation is from a brief I wrote back in 2004 on (see the brief "Terrorist Death-March") how terrorism suffers from diminishing returns against stable enemies. Simply, the more it is used, the less valuable it is (which is a good thing). The reason for this is psychological. Target Western populations (and the press) become inured to terrorism in much the same way they do with petty crime. Each subsequent attack has less of a psychological impact than the first. In order to compensate for this, a terrorism planner must make each subsequent attack even more damaging or symbolically devastating than earlier attacks. The result is a death march until entire terrorism campaign runs out of steam.
This approach in part explains why the US hasn't suffered another attack since 9/11 -- the other factors being improved security (debatable) and the break-up of camps in Afghanistan. Al Qaeda has not opted to attack the US is because it hasn't been able to muster an attack that could exceed 9/11 in damage. Instead, and this is explained in the brief I linked to above, it has moved to new targets that access new geographies and political dimensions (Madrid, London, etc.).
A good example of this framework in action is found in the recent revelations unearthed by Ron Suskind in his new book called "The One Percent Doctrine." An excerpt made available through TIME magazine has this valuable tidbit:Ali revealed that Ayeri had visited Ayman Zawahiri in January 2003, to inform him of a plot to attack the New York City subway system using cyanide gas. Several mubtakkars (NOTE: small, portable, chemical weapon delivery systems) were to be placed in subway cars and other strategic locations. This was not simply a proposal; the plot was well under way. In fact, zero-hour was only 45 days away. But then, for reasons still debated by U.S. intelligence officials, Zawahiri called off the attack. "Ali did not know the precise explanation why. He just knew that Zawahiri had called them off."