Streetlight Effect

A policeman finds a drunk searching under a streetlight. “I lost my keys,” the drunk explains. “Over there,” he adds, pointing to a dark alley. “Then why search here?” “The light’s better.” The joke names a serious error: looking where it’s easy to look rather than where the answer is likely to be.

Science calls this observational bias. We study what’s measurable, not what’s important. Psychology studies college sophomores because they’re available. Economics studies developed markets because data exists. Medicine studies conditions with clear endpoints. What’s hard to study — consciousness, informal economies, slow-acting diseases — gets neglected, regardless of importance.

The streetlight effect interacts with the mcnamara fallacy: not only do we measure the wrong things, we look for answers in the wrong places. A company tracks metrics it can gather, not signals it needs. A researcher investigates questions their methods can answer, not questions the field needs answered. The methodology drives the inquiry rather than the reverse.

The effect compounds over time. What’s studied develops better tools, attracting more researchers, generating more studies, improving tools further. Understudied areas fall further behind. The streetlight grows brighter; the dark alley stays dark. Academic fields develop not toward truth but toward measurability.

The antidote isn’t to avoid the light — sometimes the keys really are there. It’s to notice when you’re searching where the light is and ask whether that’s where the answer is. Methodology should serve questions, not determine them.

Practically: when facing a problem, distinguish between “where can I look?” and “where should I look?” The overlap isn’t guaranteed. Sometimes the hardest and most important step is figuring out how to illuminate the dark alley.