IN 1965, when New York City was hit by a “crime wave” (which later turned out to be, in part, a consequence of improved record-keeping), the city increased the number of lights in crime-infested streets. Two of my fellow sociologists described the new anti-crime measure as a “gimmick”: it was cheap, could be introduced quickly, was likely to produce momentary results, but would actually achieve nothing. “Treating a symptom just shifts the expression of the malaise elsewhere,” one sociologist reminded the other, reciting a favorite dictum of the field.  Criminals were unlikely to be rehabilitated by the additional light; they would simply move to other streets. Or, when policemen are put on the subways, there is a rise in hold-ups in the buses. So goes the argument.