A Few Useful Graphs On Violence In Pakistan

Via the Brookings Institute, there are some interesting trends on the Pakistani security situation. There's a lot more to their report -- it includes data on a variety of issues, not just security -- but I want to just put up three graphs here, and you can go see the rest on your own (you can click on each of these images for a larger view).

The first one collates the number of terrorist attacks in Pakistan since 2006, broken down by province. There are basically two main spikes: the first is in the spring/summer of 2007 -- is it a coincidence that that was the time when Musharraf became embroiled in a foolish panga with the Chief Justice? -- and the second is in the summer of 2008. The first spike also followed the Lal Masjid fiasco, which probably had more to do with violence in that year than any other factor. It is striking, though, that both spikes came at times of political instability and/or transition, though for obvious reasons we can not really draw firm conclusions based on the existence of just two trends.


Of course, this graph stops before the current spike in October and November in both the NWFP and Punjab, but the overall trend is way too volatile to make any definite pronouncements. The only thing I can safely say is that Pakistan, on average, is no more or less violent than it was eighteen months ago, which is a bit shocking. That, and the complete absence of violence in Sindh (touch wood).

Here's the second. It graphs the number of attacks by type since last year.



This graph basically backs up what I said above: that Pakistan isn't suffering overwhelmingly more terrorist attacks now than it was last year. I don't know why it feels like it is more violent.

Actually, I think this third and final graph gives a bit of a clue. It graphs the number of fatalities due to violence in Pakistan. Since we safely assume that the absent October and November numbers would continue the upward tick, it appears that even though the number of attacks is not significantly higher, the number of deaths certainly is.



In other words, the militants are getting better and more effective at killing people. From memory, this certainly seems to be the case. The last few months have witnessed a wave of attacks that kill large numbers of people -- between 20 or 40. Most attacks I recall from before this summer were smaller, killing 5-10 people.

Which leaves us with an obvious and painful question: how are they getting better at what they do when the military operations against them is supposed to be curtailing their capacity?