Distinguishing between False Positives and Genuine Results: The Case of Irrelevant Events and Elections

Abstract

Graham et al. (2023) attempt to reassess previous findings on the electoral effects of droughts, floods, tornadoes, and college football games. Reassessing previous findings by collecting new data, implementing new specifications, and preregistering these efforts is a valuable exercise, and we hope future researchers build on this model. This practice could be even more productive if researchers present independent evidence and think about how their various tests distinguish between different possibilities. We focus our discussion on the purported effect of college football games on elections in the United States. Although Graham et al. state that their results support the conclusion that college football games affect elections, we show why the new evidence they offer should decrease our beliefs that there is a genuine, substantively meaningful effect.