Violence and Evaluation: Why It Matters To Document Progress

My preferred field of research is in informal economies. This means, often, that information is very limited, existing data sets can be misleading, not cleaned up well, or just not complete. Unfortunately, a lot of the existing research is based on anecdotal evidence — I can prove some of the theories that I work with… after hours of compiling data from individual sources into my own data sets. Or going into the field and painstakingly collecting it myself.

I find that working with non-profits, especially those interested in reducing violence, yields similar challenges. The groups I work with and think about often devote their resources to the issues they are trying to address, which might make sense in the short term… but then we also run into issues where we can’t scale solutions or improve development models because there was never a system to document progress before/after a program was implemented and/or measure the impact that program had on the specific target groups over time.

What do I mean by this? Look at Ciudad Juarez. The documented homicide rate has decreased significantly since 2010, there has been a ton of investment in local social programs, the military left the policing programs to local police forces… but what worked? Many things happened at once. Which social programs were most effective and why? How do all of these changes in the local fabric of the city interact with one another? What failed? And what were the negative side effects of these changes? What are we not seeing in these new numbers? How do we evaluate “positive change?”

It’s nice that sometimes there is enough clear data from different accounts that we can draw some conclusions after the fact. Sometimes, we receive anecdotes that offer enough context that we can compare data from one story to data from another. This is an extremely slow process — compiling data from anecdotes and interviews, but it is possible.

I would love to see groups in all spheres of development, violence reduction, public investment, etc. being trained to document their findings betterĀ and making these records public. That would, of course, require them to disclose when their programs were not working… which is another public branding issue for non-profits, but would, overall, ensure that we can find better programs that really can scale to bring positive change.

A girl can dream.