What is the optimal size of a global terror organization?

It looks like leaders of global terrorist organizations like al-Qaeda could benefit from lessons in organizational theory and on the theory of the firm. As William McCants argues in Foreign Affairs, it looks like al-Qaeda may have expanded too fast under its current leader Ayman al-Zawahiri, thereby resulting in the HQ’s loss of control over its subsidiaries, franchises and affiliates in the Middle East, Somalia and the Maghreb.  

As the political scientist Jacob Shapiro observes in his new book, The Terrorist’s Dilemma, all terrorist groups suffer from infighting for one basic reason. If they want to achieve their goals and to avoid being captured or killed, leaders of secretive violent organizations have to give their commanders in the field some measure of autonomy. When the field commanders become too independent, the leadership attempts to rein them in through various bureaucratic measures.

Without a doubt, Zawahiri is trying to rein in his unruly affiliates. What is striking is that Zawahiri created much of the problem himself by trying to expand al Qaeda too broadly. The one affiliate that Zawahiri did not push into a new arena of jihad, the Yemen-based al Qaeda in the Arabian Peninsula, has, unsurprisingly, avoided infighting. Zawahiri has now allegedly appointed AQAP’s leader, Nasser al-Wuhayshi, as al Qaeda’s ”general manager” and thus his eventual successor. Zawahiri had little choice but to promote from the ranks of AQAP, given the current disarray across the rest of al Qaeda.

But the organizational woes of al-Qaeda and affiliates should not give comfort to the global community. As McCants reminds us in his conclusion, dealing with a centralized terror group with an address (or quasi address) is better than trying to fend off lots of competing franchises [see here]: 

Zawahiri’s knack for creating factions and his unwillingness to part with them when they misbehave could help al Qaeda’s opponents blame the entire organization for the atrocities committed in its name. Over time, perhaps the bloody collage will dampen enthusiasm for joining al Qaeda and even horrify its members. But in the near term, Zawahiri’s poor management is not necessarily a boon to the United States and its allies. The various factions of a once-unified al Qaeda could compete with one another over which group can mount the biggest attack on the West. Whatever the case may be, Zawahiri’s inability to manage al Qaeda’s sprawling organization offers a preview of the infighting to come after his inevitable death.

Anyone know a good paper with a principal-agent analysis of terror organizations?