Every once in a while, Foreign Policy magazine says something sensible.
"[Almost 3 million] internally displaced people (IDPs) fled on just a few hours notice -- before a military offensive meant to "flush out" the terrorists in the North-west Frontier Province's Malakand district.... [But the recent] attack on the Pearl Continental [hotel] forced international agencies to withdraw their international staff from Peshawar, disrupting assistance to the hundreds of thousands now living in government-run camps.
The IDP situation matters for more than its very real status as a humanitarian crisis. Between 80 and 90 percent of the IDPs are not in the camps; they are bunking with overstretched relatives and friends who receive no outside aid whatsoever. If the international community responds to their needs, these IDPs could present a potentially powerful constituency of civil opposition to extremism. They fled their homes because they reject the militants' worldview. If and when peace returns, they, as a resident living in the Federally Administered Tribal Areas told Crisis Group, will be the robust civil society that is so badly needed in the conflict zones."
I'm less blithe about the necessary connection between leaving and rejection of extremism. Many Arabs left the immediate conflict zone in 1948 per instructions, expecting to return over the Jews' dead bodies. The act of leaving, in and of itself, spoke to little more than a wish to be out of the line of fire. On the other hand...
It is pretty clear that this is all part of a systempunkt strategy by the Taliban. Destroy infrastructure and focus on choking help since NGOs refuse to take casualties. Then combine aid via charitable front with armed delivery groups, to be the only game in town. Leverage that for intelligence and jihadi recruiting.
Ultimately, this is where my previous discussions of the Cuban model come in, as a necessary model for addressing future conflicts. The required counter-strategy to the Taliban plan is a similar approach of unified armed aid, plus tightly coupled propaganda and intelligence efforts - the Cuban model in its full sense. Aid provided by armed groups that will take casualties, coupled with propaganda to stir up bitterness against the Taliban and al-Qaeda for ruining these peoples' lives, and cojoined with intelligence recruitment within durable family and clan networks that form the basis of these societies.
That outlines the beginning of the problem - and the opportunity.
The next stage of the problem, of course, is the local government. The only armed aid group in the area is Pakistan's military, which has corrupt ion issues and also has Islamist sympathizers within it. Some portion of any aid given can be expected to end up in Taliban hands, either directly, or as cash following black market sales.
There aren't many alternatives I can see. Foreign Special Forces groups are operating in Pakistan, but they're under constraints. That makes them a valuable piece that could direct foreign military aid delivery to the right village elders et. al., but only in a few places, and to a scale that almost certainly falls short of needs. It's still worth using them, judiciously, in this role. The question of how to work with Pakistan's own military and tribal lashkars in this effort, remains.
As for Pakistan's ISI, the only rational way to treat them is as the Taliban's patrons. Which means the intelligence recruitment will need to happen around them, and in spite of them.
That means another facet of Special Ops work, as well as CIA personnel planted and operating as aid workers. Which aid agencies don't like. But if you're not prepared to take casualties in a war zone, there isn't really a defensible argument for keeping out people who will do so.
Actually, given the stakes at hand in Pakistan, the aid agencies' concerns are basically irrelevant no matter what their position is on casualties. The number of dead aid workers that constitute a reasonable trade for a much lower chance of the Taliban and al-Qaeda getting nukes is a very large number.