Sources and Methods #10: Jason Lyall


Jason Lyall 101:

Personal Website (you can find links to all his research here)

Yale Professor Page


Jackson Institute at Yale

Institution Social and Policy Studies

Show Notes:

3:58 - I’m interested in the effectiveness of violence - when it works, whether it works - and what it does to the victims. And I’m interested in insurgents and how they react to violence, and how does it affect them? Do they ramp up their attacks, or scale down their attacks?

4:40 - Counter insurgency is not a new practice. But it’s often been approached in a fairly crude fashion. What we’re trying to do is bring some new literature to the table from spatial economics literature or survey literature and better understand it.

9:00 - Study looking at artillery fire that the Russians were lobbing more or less randomly into populated settlements. They were doing this as a way of trying to separate the people from the insurgency, showing people what the consequences of continued support for the insurgency would be. Most of our theoretical literature would say that indiscriminate violence is incredibly counter productive and would not be be effective - would generate new insurgents and new grievances - so you should see this uptick in violence. What’s interesting in the Russian case is that you don’t see that uptick. It actually seems to have decreased insurgent violence in the periods after the shelling. So it’s an outlier case... And I would say it is what has led to the dismantling of the insurgency in Chechnya.

12:12 - I think the way I would read that paper - we should not assume that violence has any one kind of effect. We should be looking for the different effects, and looking for the conditions of those effects.

14:30 - The metrics used in this research is really good for short term effects, but entirely different methods may be necessary to take the long term view and understand those effects. The field lab I run is trying to merge those two perspectives.

17:58 - We’re not good at trends, we’re not good at regional patterns, we’re not even good at generalizability inside the same case we’re studying - and these are problems.

22:00 - It is possible to get the data in difficult areas, but you have to be really really smart and careful. You can do it, but it takes a tremendous amount of time. And the risk of failure is incredibly high. In Afghanistan, I’ll typically have two or three projects going at the same time with the understanding that one of them will likely fail.

23:04 - If you adopt the right methodology, you can get more data than we ever thought possible.

26:22 - I have a love/hate relationship with the Asia Foundation. I know what they’re trying to do, but I have questions about how they’re doing it.

27:32 - The latest Asia Foundation Report here.

27:49 - Surveys for me are about mapping patterns down, rather than getting really deep insights into what a particular concept means.

29:00 - For local level questions or regional questions - I think these are a nice middle ground, using a mid-range methodology, where I can do lots of things but not everything - is the way to go.

29:35 - Smart polling should be a key part of our toolkit, but the thing that we usually lack is baseline data. By the time we start caring about a place, it’s too late to run baselines. So we’re forever behind, because we never know what was there before. So I’m a big proponent of using smart, tailored surveys. Specific questions. Specific areas.

30:54 - With that, these surveys just measure attitudes, not behavior. Important to remember that.

33:01 - In Afghanistan, I think we vastly underuse SMS as a tool.

36:40 - Explaining Support for Combatants during Wartime: A Survey Experiment in Afghanistan (co-authored by Lyall and includes the Endorsement polling technique). A primer on  the list technique in polling.

45:55 - Bombing to Lose? Airpower and the Dynamics of Violence in Counterinsurgency Wars, published in August 2014 by Lyall, found:

Evidence consistently indicates that airstrikes markedly increase insurgent attacks relative to non-bombed locations for at least 90 days after a strike.

We’d be hard pressed to find an example where an insurgency could be destroyed from an air - where air power alone has actually done this.

55:56 - I build project-specific infrastructure when I get to work. Sometimes I’m working with satellite imagery. Each project has it’s own methodology, research. A problem I’m having right now is that there’s too many platforms in play - we’re working to get everyone on the same

1:00:40 - Morning Routine:

  • I knock off around 2 or 3AM. The last thing I do is write a note to myself about what I should be focusing on the next day. If I don’t get centered quickly, I find that I lose the day, getting distracted by email, etc.

  • So that note shows me - which I keep with me - what I need to do, what I have to get through that day, it’s like my touchstone.

  • I set aside an hour or two every Friday about what I want to accomplish the following week, and these notes each morning help preserve my space to think.

  • It’s often a physical sticky - it tells me what I’m doing, and where I am in the master plan. Otherwise I write these things in Evernote.

1:04:22 - Also use:

  • RunKeeper, to remind myself to keep working out.

  • I’m a big fan of turning off my social media, letting it go dark. I use a software blocker often for up to a week at a time to keep me off of it.

Jason’s Book: The Peripheral by William Gibson. And The Book of Strange New Things by Michel Faber.  

Jason’s Film: Guardians of The Galaxy

Jason’s Song: Alt-J’s Taro