This is at least as big a problem as the troops-in-contact air strikes, highlighted by the USIP report.
Getting back to the issue of which, I'll make just one critical point: the authors of the USIP report, Alexander Thier and Azita Ranjbar, mistakenly write that "Troop levels in Afghanistan have been insufficient given the geographic and demographic scope of the challenge, resulting in increased reliance on air power." Troop levels are insufficient, of course, but it's not that aerial power is substituted for the non-available manpower really. Aerial power saves manpower instead in a different sense: in a direct sense. The available manpower is saved time and again by it. Thier and Ranjbar point out elsewhere themselves, that when one meets overwhelming force, the best from a counterinsurgency perspective is to withdraw, not to call in an airstrike. In fact the problem is greater than that. Troops will usually come under fire when ambushed. In that case, the site of the ambush is likely to have been carefully selected by the attackers. They will be shooting from higher ground, or from the protection of walled compounds, and usually from several locations, not from just one position. In this case, to kill them without air strikes would only be possible if one would be ready to sustain casualties. A hundred soldiers pinned down behind APCs spraying M-16 bullets in the general direction of walled compounds cannot necessarily be that much more effective than ten soldiers doing the same thing. Adding more soldiers on the ground will therefore not mean less air strikes.
I'm no military expert, I'm just trying to use my logic - if somebody can explain to me where I'm mistaken with this argument, I'd be happy to hear that.
But more soldiers interacting with communities (if they do) and providing security (for everyone), pulling off a real ink blot strategy instead of a rhetorical one, can prevent the infiltration of areas in the first place, and that's one of the best ways to avoid some air strikes.