Tuesday, May 11, 2010

On Drones: Lack Of Data Makes It A Difficult Issue

The other day I tweeted that the issue of drones is not an easy one for me, and is mainly typified by the color of gray, rather than black or white. Mosharraf Zaidi has a good column on this issue, and while there is some overlap in my thoughts and his piece, you'll just have to go ahead and take my word that we reached these points independently.

The reason this is such a difficult issue to have a definite pronouncement on is a lack of clear data on the topic. Essentially, we need to know three things about drones: one, do the local populations in the tribal areas support them or oppose them? Two, does the Pakistani population in general support or oppose them? And three, to what extent do the drone strikes actually hit the operational capacity of the Taliban and al-Qaeda, as opposed to the murder of innocent civilians? But questions one and three have extremely unclear answers. We only reliably know the answer to the second question, and that answer is that "settled areas" Pakistanis oppose drones in fairly overwhelming numbers (usually between 10% and 20% of public opinion poll respondents say they support the strikes, and between 60% and 75% oppose them).

Let's talk about why each of these questions matter, and why we don't know enough.

The first question is how the local populations think about drone strikes. This matters because the contested space between the Pakistani state and anti-state insurgents and militants lies, primarily, in the tribal areas. Put differently, it is no coincidence that the drone strikes operate where the state enjoys the least control. This fact, in turn, highlights the importance of local populations; counterinsurgency 101 speaks to the need to "win" the local populations, usually with a mixture of carrots (the "hearts and minds" strategy) and sticks (violence and intimidation), which allow you to exercise control to a greater degree.

The important point to note is that both sides in any civil conflict will employ both these carrots and sticks, and so it becomes a straight competition of who is doing the better job. So if drone strikes are, on the margin, compromising the Pakistani state's ability to win the allegiance of local populations, then they are bad news.

The problem is that we simply do not know what these local populations think. Opinion polling in the region is next to non-existent. The "data" we have on local populations' opinions on drone strikes is patchy and generally anecdotal. We simply don't know -- though some people pretend they do.

The second question is whether the Pakistani population in general supports or opposes drones. This matters because public opinion in favor of the war against the Taliban has fluctuated over the last six or seven years, and only relatively recently -- from the spring of 2009 on -- has it been in favor. Public support, in turn, is crucial because it allows the state to prosecute the war to the best of its ability. Nothing hamstrings a war effort more than flaccid public support.

On this question, we know fairly conclusively that the public is heavily against the use of drones, for a variety of reasons. As such, policymakers are advised to move carefully in escalating or even maintaining the presently high level of drone attacks.

The third, and perhaps most important, question is: what functional effects are the drone strikes having on the ground? This is murkiest question of all. Essentially, we have two points of view on this. One, articulated in this op-ed in the New York Times last month, argues that the drones are in fact very effective at killing and dismantling the top leaders of the Taliban and al-Qaeda in the tribal agencies, and minimize civilian casualties to the extent possible, especially relative to the much wider damage the Pakistan military wrings when it uses force.

On the other hand, there is good reason to be skeptical with such an argument, as I am. The most important point would be that the so-called data that the "effective drones" argument rests on is faulty and questionable at best. The main reason for that, in turn, is that it is next to impossible for independent observers to verify the casualty figures of "top leaders" and "innocent civilians" in these regions. Yes, the "effective drones" argument employs the best sources possible -- a variety of newspapers and online sources based in Pakistan. But even the best sources may not be good enough, because of (a) the lack of access to the struck regions in quick time, and (b) the intense amounts of political pressure -- from both sides -- for reporters to fudge their figures. (UPDATE: Here's a post on The Mob and the Multitude that makes this point much more comprehensively).

And if we don't know how many "top leaders" (whatever that means) these drones are actually killing, then it becomes incredibly difficult to assess their benefits relative to their very real costs.

So this is why it's incredibly difficult for me to know one way or the other whether I support drone strikes. The data that would actually help me figure it out either doesn't exist, or is badly and unreliably collected. I understand that drones strikes motivate strong passions, but for me, I'm very much in the unknown middle.


takhalus said...

farhat taj has done a lot of work on whats going on in FATA..again FATA isn't uniform but her work suggests there is quite a bit of support for the attacks in FATA ..especially in comparison to more destructive military operations.

Zulfiqar said...

The author has truly pointed out the problem of non-existent data, which has made it difficult to ascertain the opinion of local people about drone strikes. The survey that Mr. Mosharraf Zaidi referred to is also unsubstantial to decide what is true and what is false.

fatima-ahtesham said...

I agree with Zulfiqar

Raza said...

I spoke with with three US "drone-experts" a few weeks ago (including one of the guys who pushed Obama to dramatically escalate use of the tactic). They all basically said that although the strikes have been effective for the past few months (apparently the Uzbek militants have gotten their asses kicked by the drones), they've reached a point of diminishing returns as militants have figured out how to avoid them.

But they also said that given the weaponry and resources available to both the US and Pakistan, the drones are the most precise way to kill militants.

Personally I think the strikes notch-up a pretty decent bad-guy body count. But bad-guy or not, it's gotta scare the living shit out of anyone in North Waziristan to hear a drone flying around. And that fear is enough to make someone flip (Nicholas Schmiddle describes the fear factor in NYT piece a couple of months ago).

Oh also, the drones strikes are barely legal. You have to apply a very generous interpretation of international law to allow for the strikes.

Anon_for_a_good_reason said...

Drones strikes are not an attempt at counter-insurgency.Its a temporary & partial fire-fight counter-terrorism strategy of decapitation and disruption of AQ+Taliban command & control. "Temporary" as in until Pak Army makes up its mind.Nothing can substitute for boots on the ground.May be, if a major US mainland attack is successful,then US might also add an element of Afghan Army special forces to covertly raid AQ safe havens.

TLW said...

The "data" we have on local populations' opinions on drone strikes is patchy and generally anecdotal. We simply don't know -- though some people pretend they do.

I feel the second sentence deeply applies to Farhat Taj. The stuff this lady comes up with does have an anecdotal ring. She may be prob ably the best source available but I still have my doubts about her.

Maybe Rahimullah Yousufzai would be better?

And Anon-For-A-Reason is right that this is only a stopgap measure.

@ Raza's points. This is a good quote:
the strikes have been effective for the past few months (apparently the Uzbek militants have gotten their asses kicked by the drones), they've reached a point of diminishing returns

Back to front; there is a concept in economics called diminishing marginal utility. I get that feeling too about drone strikes; that the marginal utility of the program is decereasing with each strike.
And the reason that Uzbek militants are easily targetted is because, well Everybody Hates Uzbeks.
But seriously, Uzbeks aren't local, even less local than Arab Mujahideen who settled down and married Waziri and Mehsud girls 22 years ago. So it's easier to strike recent forein militants.
But in favour of the program is the obvious attacks on Baitullah Mehsud.

Now all we need is Hakimullah Mehsud.

Let's wait till the drones get Qari Hussain and nine-lives Hakimullah.

takhalus said...

Raza barely legal has rarely been a US justification to not do certain things..

I don't think you can compare Farhat Taj and Yousafzai, the latter tends to write in broader strokes and does't cite facts very often..he is a good writer and his contacts have been miled by western journalists for their own books.

Taj does tend to make some very inaccurate statements but she remains the only one who has focussed on the region.

mkj said...

And three, to what extent do the drone strikes actually hit the operational capacity of the Taliban and al-Qaeda, as opposed to the murder of innocent civilians?

I think this question has two parts. 'Whether The drone attacks kill AlQaeda/ Taliban leaders as opposed to Civilians?' and 'Whether the killing of AlQaeda/ Taliban leaders reduces their operational capacities?'. Both the questions are difficult to answer but important in their own right.

Juan Cole in a recent post argues that organizations are weakened by death of their leaders whereas movements are not. He thinks Alqaeda is an organization whereas Taliban are a movement. Again there is little data/ studies to back these claims, but if true it further complicates the matter.


Rabayl M. said...

Perhaps this was the blog post you were looking for?


Ra1 said...

Even Cyril Almeida has had a post on the same topic last week...

Ahsan said...

Thanks Rabayl, that's the one.

Sadia said...

The lack of data does make it hard for us to analyze the effectiveness of drone attacks, as the terrain is inaccessible to our security forces which are why drones are used at the very first place. The figures of civilian casualties are often exaggerated and as Farhat taj opines that the militants cordon off the area and remove their bodies. However it is important that these attacks are carried upon real time information so that they are more effective.