Getting Things Wrong Really Does Help, as Long as You Keep Trying to Get Things Right: Developing Theories About Why Homicide Rates Rise and Fall

Abstract

How can we come up with new ideas in criminology that have a chance of being better than the ideas that have come before? That’s a humbling question for scholars like me, whose insights have been born of failure. It’s a daunting thing, even in a world of tenure, to work for a decade on a project, only to find that the data have made a hash of your hypotheses. As I put it to friends at the time, “my theories died a horrible death in the face of the evidence.” It’s tough to end up with a bigger puzzle than the one you started with and no responsible way to publish what you have found, because the data don’t make sense in light of your initial theories or the theories of your colleagues. But that’s the way many advances in knowledge come about, especially in criminology, where official data are of such poor quality and so limited across time and space that we have to gather our own data, project after project, without knowing what they will reveal. And the data, once gathered, almost always surprise us—indeed, I believe they will always surprise us if we are open to what they can tell us. But that’s a wonderful thing, because it means, as I remind my students constantly, that we don’t have to be geniuses to do original work in criminology. We just have to work long hours and trust that better hypotheses will emerge from the data.