In one of the trashier John Waters films (a pleonasm, I know), Desperate Living, Peggy Gravel (Mink Stole) is an insane, rich, murderous housewife on the run from the law. During her escape through the woods, before being accosted by a cross-dressing traffic cop with an underwear fetish, she exclaims her hatred at the forest around her:
You know I hate nature. Look at those disgusting trees stealing my oxygen Oh, I can’t stand this scenery another minute! All natural forests should be turned into housing developments! I want cement covering every blade of grass in this nation. Don’t we taxpayers have a voice anymore?
That was the first thing I thought of when reading David Kilcullen, the grand poobah of anthropologically-enabled counterinsurgencies, explain the way roads decrease violence in Afghanistan. Then I remembered how much I mocked Ann Marlowe for suggesting the same thing. Clearly this deserves some further examination.
Did you know that the one thing keeping NATO and the U.S. from securing Afghanistan was a nice, thick layer of asphalt? I certainly didn’t. In fact, when a commenter in that Marlowe post claiming to be “a soldier on the ground” in Afghanistan repeated the causation between road-paving and fewer IEDs, I noted some contradictory evidence and asked for further data:
I’m afraid the data we have available doesn’t support the correlation between roads and security. One risk monitoring firm actually claims that good roads have actually made insurgent activity worse… Now this doesn’t specifically address the number of IED incidents; unfortunately the data I could find on that was really vague (aside from claims that the Taliban “rule the roads“), with the exception that Ms. Marlowe claims. Indeed, given the risks to PRTs and local construction teams, it wouldn’t make sense for paved roads—which require a lot of lingering time and present workers as easy prey to roving insurgents—to be constructed in a high security risk area. From my understanding of areas of road construction vis-a-vis IED attacks, an area needed to become safe enough for workers to actually complete construction.
I don’t doubt that once roads are paved, IEDs are harder to plant (fewer suitable hiding places), and they’re easier to find. So in that sense I was being unclear in my original phrasing in the post above. I meant to say that there is no positive causation, at least with the unclassified data we have on hand: nothing I know of points to “more concrete = less bombs” as a security strategy.
My pleas for more data to support the connection between roads and security went unanswered. Indeed, subsequent research has reinforced the dual-use dangers of roads in Afghanistan—going as far back as 2004, westerners and aid workers either avoided the main highways or traveled along them at unsafe speeds because of the danger not just from IEDs, but also bandits, militants, corrupt police, and plain old Taliban out for some white people to murder. That situation has invariably become worse, especially in the south but also in eastern provinces like Kunar and Nuristan.
I would challenge any aid worker or PRT official (and this is a serious request) to tell me how roads have improved the safety in their province. Can non-military groups travel around paved roads without fear of attack? Do VIPs travel via road, helicopter, or airplane whenever possible? These are far more important indicators of the security value of road construction than vague assumptions that roads are a way to reward villages as if they were dogs in training.
Indeed, Kilcullen starts his post with an allusion to the Romans’ use of roads in building Empire. The analogy saddens me, as I really don’t think we’re building an empire in Afghanistan, and the roads were also used by the many barbarian hordes to invade, sack, and eventually dismember Rome. Not the best analogy to draw.
But Kilcullen has field notes! To my horror, he claims the techniques used in the Surge in Iraq were first developed during his trip to the Kunar Valley in mid-2006. This was the time when Gulbuddin Hekmatyar was probably hiding there and directing attacks on U.S. and GoA personnel, and when the area was known as “enemy central” for its extreme concentration of violent extremists. Besides which, how would a trip to a rural and isolated valley with years of violent resistance to American rule indicate anything about tweaking urban counterinsurgency tactics, anyway? Argh.This is a side note.
Let’s excerpt those notes a bit.
The PRT’s main project at present is the opening up of the Korengal Valley, to assist in clearing out a former major stronghold of the enemy, and to bring development and governance to the area. The main push is centered on driving a paved road through the valley to allow forces to secure the villages, driving the enemy up into the hills…and affording freedom of action to civilian agencies so that they can work with the people to extend governance and development.
The road project involves a series of negotiated agreements with tribal and district elders – the approach the PRT is taking is to make an agreement with the elders to construct the portion of the road that runs through their tribal territory. This has allowed them to better understand the geographical and functional limits of each elder’s authority, and to give the people a sense of ownership over the road: since a local workforce has constructed it (and is then paid to protect it) they are more likely to defend it against Taliban attacks. Also, the project generates disputes (over access, resources, timing, pay, labor etc) that have to be resolved between tribes and community groups, and this allows Afghan government representatives to take the lead in resolving issues and negotiating settlements, thereby connecting the population to the provincial and local administration and demonstrating the tangible benefits of supporting the government.
This tells me things I did not know before:
- The Taliban attacks roads.
- The Elders control the Taliban.
- The Afghan government’s local representatives are effective at moderating tribal disputes.
None of these assertions are born out by reading press reports of the many challenges facing development efforts. Which doesn’t mean they’re not true, just that Kilcullen doesn’t really explore why his experience (or assertion—I can’t tell which it is at this point in his writing, as it seems more theoretical than experiential) is at such variance with the years of published anecdotes on the topic. For the record, where he was is a very pretty place:
Credit: Panoramino user Joelpac, in his beautifully shot gallery of Kunar
Once the road is through and paved, it is much harder to place IEDs under the tarmac surface or on the concrete verge, and IEDs are easier to detect if emplaced. The road provides an alternative works project to prevent people joining the Taliban, the improved ease of movement makes business easier and transportation faster and cheaper, and thus spurs economic growth, and the graded black-top road allows friendly troops to move much more easily and quickly than before, along the valley floor, helping secure population centers and drive the enemy up into the hills where they are separated from the population – allowing us to target them more easily and with less risk of collateral damage, and allowing political, intelligence, aid, governance, education and development work to proceed with less risk. Road building is not a panacea, but the way this PRT and the local maneuver units are approaching this project is definitely a best practice
Again, it is difficult to parse this fully having never been to the Kunar valley, but this simply does not track with the non-military reporting coming out of the area (and what of the experience with roads and IEDs in Iraq?). Unemployment is usually not a primary reason people seem to join the Taliban, though it is a reason they’ll contract out their services to a militant group—which is not, it is important to note, the same as “joining the Taliban.” In many cases, better roads have made it easer for the “Taliban” (Kilcullen doesn’t seem to realize that drug and timber smugglers use these roads as well, and they can be just as violent and capricious) to push further into communities and launch attacks.
Is it good tactics to move along a valley floor? I could be missing something, but one of the sticking points in “The Bear Went Over the Mountain” by Les Grau was the ease of setting up killboxes at either end of a valley used as a hiding place by mujahideen. This happened again and again, and accounted for a surprising number of Soviet military successes. Did traveling along a well-marked road on a valley floor in a U.S. military convoy suddenly become a good idea since 1989?
It is comforting to see Kilcullen admit road building is not a panacea, because it is not. But that is certainly the implication of his essay here.
And a further sidenote is really quite necessary: I think road building is a fantastic idea. As he says, roads really do spur economic growth and enable a more thorough interconnection between communities. Roads are a great idea for their own sake, in other words—people need infrastructure. But pursuing road construction as a security strategy is just shortsighted and seems to confuse causality: as has been pointed out before, construction is only viable once security has reached a certain minimum state to ensure workers aren’t held ransom or murdered. And won’t insurgents innovate? That has been the hallmark of the fight so far: they innovate tactics much faster than we do. Relying on huge static construction projects to undermine them is simply shortsighted—that’s the point I’m driving home here.
Let’s look at Kilcullen’s next entry.
[the Brigade Commander] said that parties of enemy still infiltrate in this area, coming down by night from the hills on the Pakistani side, crossing the river on truck inner tubes, spending a few days attempting to do armed propaganda work in the villages on the Afghan side of the river, then moving up into the hills to avoid our patrols. Many of them congregate NW of the river, just beyond the [XXXX] valley, in a district we rarely visit and which remains a pocket of insurgent activity – but one the Brigade tolerates because there is no access from this isolated valley to the rest of the population and their focus is on securing the bulk of the population rather than clearing terrain, and because they lack the forces to secure every part of the province and have therefore sensibly “triaged” their AO.
This tells me the biggest problem is a manpower shortage, and not the presence of too few paved roads. Sure, the roads make it easier to move through cleared areas (and in cleared areas the economic benefits of roads are certainly much more apparent), but if there really is a problem with the Taliban attacking roads as Kilcullen said, then how could they be constructed in a non-secure area as a security strategy? There is something missing from this account, and I’m not sure what it is.
Like the other Regional Command-East commanders, he is all about development and governance. Having fought a hard kinetic fight to gain control of the province in 2005-6, during [Colonel, former Brigade Commander] Mick Nicholson’s time, the focus has now shifted to economic and political issues, with ANA/ANP doing the bulk of the security work, supported by a smaller US footprint and by local agreements and neighborhood watch forces.
Again, dipping into popular mass media reports from 2007 and 2008, I do not understand where a Brigade Commander could say with a straight face that his biggest concerns in Kunar are electricity generation and crop rotation. In the paragraph above, Kilcullen was saying there remain pockets of insurgent activity and known border crossings. Are all these reports about bombings and combat in Kunar just false? Did the war there end in 2006, and now all that’s left are adventure-seeking reporters busy hyping small gun fights as epic battles?
To repeat: I am missing something here.
As for his 16 key points… well, very few of them flow from the narrative he constructs in bits and pieces from his two field notes excerpts. This again doesn’t mean they aren’t true, but rather that there is no evidence in what he writes to suggest they do. There is no information about the impact roads have on cross-border infiltration, especially considering his assertion just paragraphs above that there are still routine border crossings by militants. Similarly, if the BCT is stretched so thin it can’t go after known areas of insurgent control, how is it that they can drive all militants into the hills? This is probably the case for some villages near the FOB, but for the whole province? This is a doubtful claim.
The points about access are wonderful, but they don’t address the security issue. Even if a convoy can go 80 mph along a road and thus a platoon can cover a vast territory, that only applies to the road itself. If the insurgents are fanning out into the countryside, then you’re left with a classical rural insurgency—and neither Nepal nor India have had much luck in combatting theirs.
From what I’ve seen of the roads in this part of the country, which are often hacked out of mountainsides, the difficulty of hiding an IED is a strange claim to make. Even along a valley floor, unless all scrub and rocks have been cleared from the area (which in some cases is true), there are still plenty of places to plant IEDs. If this only applies to broad valley floors, that is a limited test case. More convincing are reports that villagers simply do not wish to plant IEDs but are coerced to, so they choose to plant them in obvious places to facilitate discovery.
In Kilcullen’s conclusion, there are a lot of pretty words, but no discussion of causation. He saw well constructed roads, he got solemn assurances that the local BCTs and PRT were focused on governance and construction, and he seems to have drawn from that the idea that roads spur security gains, and thus counterinsurgency success. He also seems to dismiss the dual-use problem of road building:
Thus, the mere building of a road is not enough: it generates some, but not all of these effects, and may even be used to oppress or harm the population rather than benefit it. Road construction in many parts of the world has had negative security and political effects, especially when executed unthinkingly or in an un-coordinated fashion. What we are seeing here, in contrast, is a coordinated civil-military activity based on a political strategy of separating the insurgent from the people and connecting the people to the government. In short, this is a political maneuver with the road as a means to a political end.
This is entirely different than the argument he was laying out above, and not at all connected to the practice of road construction he described—the relationships and social flows relayed secondhand from PRT and Brigade staff talked about benefits in a general and theoretical form, and not any concrete ones. There is no reason here to think that better governance and better security are positively related to road construction. They certainly coincide and might even correlate. But there is no reason to assume causality.
And this is just bad form:
In summary, like the Romans and other counterinsurgents through history, U.S. forces in Kunar, in a close and genuine partnership with local communities and the Afghan government (most especially, a highly competent and capable Provincial Governor), have engaged in a successful road-building program as a tool for projecting military force, extending governance and the rule of law, enhancing political communication and bringing economic development, health and education to the population.
The Roman experience in exercising control over restless outer provinces, and the experience with how their roads were turned against them, can hardly be the lesson Kilcullen wants drawn from his analysis. It is good to see him cautioning against generalizing from this one area (a caution Ann Marlowe would do well to follow), but at the same time, the security benefits of road building are totally oversold and not backed up by evidence, only theoretical discussions of how roads should function within a larger counterinsurgency campaign.
None of this is meant to knock the very real benefits of road construction. Roads are one of those critical pieces of infrastructure that make the rest of the nation-building project go a lot more smoothly. In fact, there should be a great deal more road construction all around. But the military benefit of roads here is simply ridiculously oversold.
This Topic Continues: