Where RCTs were not feasible, I believed that quantitative analysis would allow us to resolve every question, whether we were investigating the efficacy of selective serotonin reuptake inhibitors or the social return on investment of antimalarial bed nets. Armed with tools such as regression analyses, models and simulations, etc., I thought we could identify the right path to take at any crossroads we faced.
With the above in mind, this budding empiricist came to business school to get those answers.
Nearly two years and dozens of regressions later, I am coming away surprised. The right answers continue to elude me and, in fact, may often not exist. If data is king, as they say, then it’s certainly a mercurial and fragile one, easily besieged by unrealistic assumptions, imperfect proxies and the manipulation of those with preexisting agendas.
For example, take this simple question from my social enterprise class this past semester: Was Ben & Jerry’s right to sell out to Unilever? Contrary to my initial expectations, we cannot simply construct an algorithm that captures the entirety of the social, economic and financial impact of that decision, especially in relation to the unknown and unknowable counterfactual. Models are beset by inadequate proxies, valuations vary with the multiple you choose, and data can be sliced and diced to support the bias of their presenter, as I also learned upon evaluating John Lott’s “more guns, less crime” hypothesis in my law school class.
Looking back, then, I haven’t quite gotten the black and white answers that I originally came for. However, one could argue that I am coming away with something better. For one, Yale has equipped me with the tools to arrive at better answers than I would have had had before. More importantly, though, this experience has yielded a humbling lesson: Life is just too gloriously complex and unpredictable to come prepackaged with a single, “right” solution to every question we ask, and unraveling these complexities is ultimately what makes living—and learning—so interesting.