This is actually a very interesting question for the simply reason if one asks the question when does an insurgency end one must equally ask--what is the exact ecosystem of the insurgency look like and or what is even the ecology of an insurgency.

All of the current intel analysis focuses on people, places, and things---not on how does the ecosystem function, hows does it communicate, what happens when it splits or a top leader is killed, how does it respond to the local population it survives in---all extremely valid questions.

This blog has totally ignored the the ecological based research released in Dec 2009 in Nature which actually would answer the question-when will it end as it would tell you how it even began and describe it's actual ecosystem.

Out of the article "How will it end?

From 'How Insurgencies End'

Government victories often cause the insurgency to splinter, leaving behind
small elements of irredeemables that may or may not represent an ongoing
threat. Tracking these splinter groups can provide tremendous
insight into the nature of the insurgency ending. We noted that, when the
government is winning and the insurgency is in its "tail" phase, often a
small group of insurgents splinters away from the cadre or leader-

In some cases, this splinter element is an irredeemable fringe unwilling to
negotiate or enter into an amnesty program. In others, the splinter is
formed from an irredeemable core leadership element. In both cases, the
splinter groups are intent on continuing the struggle against the government
in one form or another. They may
try to sustain or reignite the insurgency immediately, or they may be
willing to hibernate until they see an opportunity to reemerge.
Counterinsurgents should be able to tell a great deal about their long-term
prospects by observing how the general populace perceives this splinter
group.

This statement above begs the question from an ecology perspective, how do you track 'splinter groups' - are these a sign of weakening insurgency, or as a sign of evolution?

This was answered in the research:
Redistribute: When a group is broken the components are redistributed amongst the other groups in the system. The redistribution is biased towards the most successful remaining groups.

To understand the concept of ecology based research the following must be understood:
Ecology of an Insurgency:

The scientific study of the way that living “organisms” in this case “organism” is defined as an insurgency cell, group, or organization interact with their environment and predators (the counter insurgent).


Ecosystem of an Insurgency:

An insurgent ecosystem is a system whose members (members defined as being either an insurgent group or groups) benefit from each other's participation via symbiotic (mutually beneficial and self-sustaining) relationships.

The main goal of an insurgency ecosystem is to generate common ventures. It forms when many small and potentially diverse (origin, tribe, religious belief, etc.) insurgent groups join together to fight a common predator (the counter-insurgent or state).

Insurgent ecosystems attract and retain members (groups) due to network effects:

• The benefits of the ecosystem (shared ventures) are so great that groups won’t leave it (although temporary departures to avoid targeted pressure from counter-insurgents are possible).
• The ecosystem’s features (i.e. immediate access to shared resources) make it easy for new groups to form and participate.
• The growth of the ecosystem results in an exponential increase in benefits (i.e. more segmentation and specialization) for all of the member groups. IE Attacks by one group creates opportunities for other groups. The buying of resources (ie small arms, explosives) creates a market for groups to sell into and makes it easier for other groups to get access to the resources.
• An ecosystem can have groups directly fighting each other through direct battles - but it can also have indirect fighting (or competition) between groups for access to resources (people, money, strategy etc).


Once an ecosystem is established in a particular region/area, it becomes very difficult for the counter insurgent to eliminate it. The presence of multiple groups means that the counter insurgent must divide its efforts. Operationally, a focus on one group leaves other groups to operate freely and success against one group yields very little overall benefit. Removing leadership does not mean that the group will cease to exist. The leadership may be replaced by other parts from the same group or other groups. Or a new group will move into the space left open by old group. Strategically, the diversity of the groups in the ecosystem (different reasons for fighting) means that it isn’t possible to address a single set of issues or grievances at the national level that would reverse the insurgency (via negotiated settlement, repatriation, etc.).

So the answer to the question "when will it end" ---it will end WHEN the insurgent wants it to end based on his view of his own ecosystem.

It is up to the counter-insurgent to "shape that ecosystem" to the degree that the group in that ecosystem has no other option than to disengage.