dr. kilcullen? >> in fact, dr. jones' organization, the rand corporation, did a study of impart what would be the effect of removing a sanctuary on the chances of -- the chances of success. i am pretty sure that you have a very significant improved chance if you can reduce the sanctuary. it is roughly about 86% of cases where you can successfully destroyed the insurgent sanctuary in enabling country, the government wins. but if you fail to destroy the site to work, you still win in about 60% of cases. it is very advantageous but it is not essential. i think we should bear that in mind when thinking about what we have expect from pakistan. but that was a also bear in mind the history of our relationship with pakistan which you know better than anybody else. which have some realism about our expectations of -- which have some realism about our expectations. i do not think we will see a significant lack of support, certainly not for the haqqani network, possibly not for others. >> and is that because they perceive a v

dr. kilcullen? >> sir, in fact, dr. jones' organization the rand corporation did a study on removing a sanctuary on chances of success in a counterinsurgency environment. i'm quoting from memory but i'm pretty sure that you have a very significant improved chance if you can reduce the sanctuary. i think it's roughly about 86% of cases where you can successfully destroy the sanctuary in a neighboring country the government fails. but if you fail to destroy the sanctuary you still win in about 60% of the case. so it's absolutely not essential to destroy the sanctuary. it is very advantageous but it's not essential. so i think we should bear that in mind when thinking about what we expect from pakistan. i think we should also bear in mind the history of our relationship with pakistan which you know better than anybody else. and have some realism about our expectations of what they will actually do in response to this series of events. i don't think we're likely to see a significant drop of support. certainly not for the ha

dr. kilcullen? >> yes, senator. i think we have covered this to some extent.ust to rehash, do we need to do it? yes. because if we lose the war, that all of the money we have spent so far would be to not. aught. can we afford it? that is the purpose of focusing on the drawdown right now. 80 percent of the costs of the war is the cost of u.s. combat involvement. another 10% is ansf development and -- involvement and 10% is civilian assistance. the devil in the detail is the issue of demobilization. if you expand the afghan security forces to 378,000 and order to win the combat phase, what are you doing with all those armed guys afterwards? what is the plan to put them into productive labor rather than having them on the street with weapons? that has been the achilles' heel of most foreign security assistance programs of this type. it is something the people in afghanistan are well aware of. but it is something that we are willing to engage with as a priority problem as we get closer to 2014. >> so you then agree with the assessment that we need to increase the