Comparative Politics

Field Work in Conflict Settings: Or, the Anatomy of an (Almost) Ambush

Jason Lyall Oct 4 '11

Is there a set of best practices for fieldwork in conflict settings?

I recently returned from Afghanistan, where I was piloting another survey experiment, this time looking at how (co-)ethnicity shapes an individual’s willingness to share sensitive information about wartime activities. Several of our villages are in Konduz, a one-safe northern province that has now been largely carved up into different fiefdoms controlled by local militia (arbaki) or the Taliban.

As part of my due diligence, I traveled to Khan Abad district with my survey team to participate in training sessions and to conduct site inspections of the survey actually being carried out. That’s when the problem arose. While we had negotiated prior access to our villages with the local militia commander, word had leaked from a careless enumerator that “two foreigners” (my program manager, Prakhar Sharma, and myself) were present. Turns out  a rival arbaki commander, one who controlled the main route to our chosen village, had also heard the news and was now setting up an ambush to catch us as we exited the village. Since I’m writing this blog post, you know that things ended well, a fact due to the quick-thinking of my own security manager, a local with very good ties to the arbaki who caught wind of the ambush in time and re-routed us to safety.

This got me thinking about best practices in fieldwork–if there are any. A quick scan of the literature finds a wealth of work on the ethics of field work in conflict settings (“Do No Harm”), but not much in the way of practical advice (other than “Don’t Be Stupid”). It may be that settings are so different that little of worth travels, but I’m not so sure about that. In the interests of sparking a conversation, then, I offer a few practices below that I’ve found helpful during fieldwork in Afghanistan and the Northern Caucasus. Additional advice and suggested readings from any discipline would be welcome in the comments.

(1) Maintain information security. Looking back, this is where I messed up, since we never instructed our enumerators that “loose lips sink ships,” as the saying goes. Keep movement and lodging locations a secret, and vary your routes. If you have a team, make sure they understand the importance of not revealing details of field sites, arrival/departure times, etc.

(2) Don’t be a tourist. Make sure you have a local on your team who can scout conditions in advance. The fluid nature of these settings means that you need up to date information to make good decisions, so make sure your partners (NGOs, survey firms, local academics) can deliver if you can’t.

(3) Interpreters. Best if you speak the language. If not, then you need an interpreter who’s an advocate for you, not your local partner. You need unfiltered information, so find an independent interpreter if at all possible.

(4) Hope for the best, plan for the worst. Be sure you know how to get out of somewhere, not just in. Do your logistics homework. What happens if your vehicle breaks down on the Salang Pass, for example? (Not a hypothetical example)

(5) Know when to quit. I had to pull the plug on site visits to a second district, Imam Sahib, when our advance party found multiple Taliban roadblocks outside the district center earlier than expected. Indeed, children were openly collecting “taxes” for the Taliban at additional roadblocks. Given the risk involved in moving through multiple checkpoints, I felt my presence endangered the team, and for little gain, and so we opted to scrub the site visit.