It’s Complicated

Evolutionary biologist E.O. Wilson was on NPR talking about the behavior of ants. This was on Freakonomics Radio, where Freakonomics authors Steven Levitt and Stephen Dubner explore economic and behavioral theory on everyday events. I won’t go into ant behavior, but it segued into Levitt saying that evolutionary biologists and economists try to strip things down to their simplest form and explanation, while fields like literature and history and social science try to understand complexity and inter-relatedness. I laughed out loud, driving alone in my car around the Beltway. I was a Literature major, I read history for fun, and my professional work is planted firmly in social science. I am a sucker for complexity.

Why should this matter? That moment on the Beltway helped me understand a tension in my programming that has bugged me for years. I often have lively debates with colleagues, donors, and researchers about project design. It’s thrilling stuff, really! Usually we are talking about the Nigerian Urban Reproductive Health Initiative, which is admittedly a complex project, designed to address a complex problem, the low use of contraceptives in urban Nigeria, where maternal and infant mortality and poverty are high.

The discussion usually goes like this:

The other guy: That Nigeria project works to increase contraceptive use! How awesome! So tell me, what is the driver of change?

Me: Well, there are lots of inter-related barriers to contraceptive use, so we address each of them. Demand for contraceptives is a major driver, but if you don’t have trained providers and contraceptives in stock and if the policy isn’t supportive it is going to fall apart. So we address all those things simultaneously using interlocking strategies.

Silence.

The other guy: So, if you had to choose ONE of the interventions to call the key driver, what would it be?

In that moment in the car, in a flash I came to see this reoccurring discussion as a valid meeting of two world-views and types of training, rather than as a cynical attempt to pick apart my beloved program.  (If I have had this conversation with you, and if I got “salty,” as my son would say, I humbly beg your pardon.)

It is the job of certain people to strip things down to their simplest form, and it is the job of other people to see problems as complex systems. It isn’t that economists deny the existence of complexity – Levitt says that economists like him (or ant biologists like E.O. Wilson, perhaps) acknowledge that they can’t get into someone’s head and know the complex soup of motivations that drive behavior, and so they don’t try. They simply look at the behavior, and the measurable factors outside of it. And it isn’t that complexity loving people like me don’t also like to see straight lines between inputs and outputs, but that we think entering into the real mess of human motivation is unavoidable in order to solve big problems.

Perhaps the recent embrace of behavioral economics by social and behavior change is an example of a meeting of the minds – maybe not a Vulcan mind meld, but a good old fashioned handshake. It seems to me that this application of economic theory to individual behavior is great for single actions – pick this rather than that product from the grocery shelf, sign up for insurance so you don’t lose the promised bonus. Each problem is an island, and each can be individually influenced. Social science approaches, on the other hand, see problems as islands in a vast sea, shaped by a shared climate and ecology and ideally managed with root causes in mind.

As for me, I am going to work on having more sympathy for the request that I strip program design down and describe one straight line between an input and an output. I’ll try to address the motivation of the request – to make an intervention simple enough to replicate, for example – rather than simply re-affirming that the program works as designed.

I will try very, very hard to resist replying: It’s complicated.