[From Brookings Institution´s Tech Tank blog, April Fool´s Day 2015]
I feel like I need to see a shrink. A work shrink, that is, one that can help me address some deep-seated issues and conflicts I’ve accumulated through my years working in economic development.
My right brain tells me that we need to take holistic approaches to development; that it´s futile to build marketplaces if there are no roads that lead farmers to them, to give traders microloans if they don’t have basic commercial skills, or to invest in primary education if we cannot tend to the children’s health. My right brain is full of ideas, but knows that there are no miracle cures, no silver bullets. It prods me to try different things and not lose faith that things can get better. Development is a process not of putting individual balls in motion, but of balls colliding in complex, reinforcing ways.
Yet my left-brain wants me to move methodically through all the issues, sequentially, cutting them into small chunks so that the impact of each can be assayed through purposeful, careful, randomized (i.e. clinical) experimentation. It urges me to avoid chasing grand theories, and instead to cash in small impacts here and there. My left-brain wears a white coat; it demands verifiable evidence that can be attributed to specific factors, not general observation.
I’m delaying the visit because I know what the shrink will tell me: that I don’t have to choose, that I should engage both sides of my brain at the same time – it’s the power of and.
But how? My right brain accuses my left-brain of being on a futile search for not one but a whole sequence of silver bullets – for isn’t the expectation of a single, small thing having an impact by itself the very definition of a silver bullet? My left brain retorts that the right brain is trying to get to El Dorado without a map; should we not properly call that a wild goose chase? I’m at a mental stalemate.
I thought I’d reconcile them by proposing to do clean experiments not only individually on each of the n possible development interventions, but also on each of the 2n combinations between them. Get evidence not only on each ball, but also on each collision scenario between any number of balls. But nobody will give me the money or the time to do that. If we assume (rather harshly) that there are only 20 things that may matter at all in development, then that will require us to do over a million (=220) Randomized Controlled Trials (RCTs). There probably aren’t even enough econ PhD students in the whole world to slave on these RCTs. That can’t be the way.
I learned from Daniel Pink’s book A Whole New Mind that I am better off building up my right brain credentials in the coming years. With the increasing delegation of routine tasks and data analytics to machines, the balance of (human) power will shift from the reductionist, analytical left-brained to the holistic, empathetic right-brained. Suits me fine: I am good at numbers, but I like interpreting them through stories. I feel that the development world is veering sharply to the left-brained, because the spinning of detailed hypotheses, research plans, and evidence-bases is so attractive to donors. This may validate the main conclusion of the book: in the data- and computing-rich western world where correlations are a dime a dozen, the right brained rule. But in a data-scarce development situation, the left-brained are supreme.
Talking to the shrink is not just a matter of seeking professional guidance on which is the most impactful approach. There is also an element of personal branding and self-esteem involved. Who wants to get in front of the data science train and be accused of being faith-based? Equally, who wants to be dismissed for being (oh that awful word, outside academia) academic?
In fact, I need to go to the shrink mainly because both brains are making me (professionally) depressed. The right-brain because, the chance that we will stumble upon the right cocktail of interventions and the left-brain because it deals with components rather than entire systems, and we are trying to measure small impacts with such coarse-grained rulers.
Anybody know a good development shrink?