Embracing the Certainty of Uncertainty

Many extremely smart people have famously made predictions that turned out to be really wrong. Consider this: "I think there is a world market for maybe five computers." -- Thomas Watson, chairman of IBM, 1943. Or this: "Stocks have reached what looks like a permanently high plateau." -- Irving Fisher, Professor of Economics, Yale University, 1929.

Nate Silver, blogger, statistician and psephologist (someone who studies elections), first came to public attention in 2008 through his unnervingly accurate forecast of the November 2008 US presidential election —correctly predicting the winner of 49 of the 50 states. Silver has a new book out later this month, “The Signal and the Noise: Why So Many Predictions Fail — but Some Don’t,” an excerpt of which was published last week where he talks about the enormous strides weather forecasting has made over the years.

Weather forecasting has made stunning progress over the past few decades. Landfall of two of the most devastating cyclones of recent times, Sidr (hitting Bangladesh in 2007) and Nargis (hitting Myanmar in 2008), were predicted 13.5 and 11 days, respectfully, in advance. The National Weather Service also made a nearly perfect forecast of Hurricane Katrina, anticipating its exact landfall almost 60 hours in advance. The introduction of Doppler radars for tornado detection in the United States in the 1990s led to a drop of 45 percent in fatalities.

While improved science, technology and huge investments have all played a role in our increasingly accurate weather predictions, Silver highlights an often overlooked issue: weathermen recognize and respect uncertainty in their predictions. The earth’s atmosphere is the ultimate complex, adaptive system. The mathematician Edward Lorenz laid the foundation of what is now known as “chaos theory” to describe the behavior of dynamical systems that are highly sensitive to initial conditions. A small difference in initial conditions (such as those due to rounding errors in numerical computation) can yield widely different outcomes for chaotic systems.

We live in a world where chaos, complexity and uncertainty reign- ranging from the frequency of human heart beats to the size of cities. Most of the estimated impacts of climate change, for example, can be attributed to increased variability and uncertainty around the hydrological cycle. Development itself is a messy and complex endeavor. We, as development professionals, need to recognize, embrace, and find effective ways of communicating uncertainty.