There is an old joke about a man searching for his keys under a streetlight. A passerby offers to help, and after some fruitless minutes asks whether the keys were dropped here. 'No,' the man admits, 'I lost them in the park. But the light is better here.' Social scientists tell the joke on themselves, because it names a temptation built into their work. Data are unevenly lit. Some phenomena—retail sales, standardized test scores, hospital admissions—are recorded meticulously and are easy to study; others, like informal caregiving or undocumented labor, leave almost no trace in the record. A researcher who follows the data will drift, almost without noticing, toward the well-lit questions, and a discipline that rewards publishable results will reward that drift. The danger is not that the streetlight studies are wrong. It is that, accumulated over decades, they quietly redraw the map of what counts as important, until the dark park—where the keys actually lie—drops off the edge of the field's attention entirely.