Steven Levitt, author of this summer’s bestselling “Freakonomics,” says there are lessons for businesses to be learned from his unusual data analysis methods. First, use your data. “Often, data generated for one purpose is useful for another. ‘Freakonomics’ describes the case of an entrepreneur selling bagels in corporate offices who kept meticulous records to track profits and losses—data that eventually yielded insights about white collar crime and the effects of office size on honesty.” Ask different questions, and don’t mistake correlation for causality, says Levitt, citing the case of a feudal king who, upon hearing that disease was greatest in areas with the most doctors, figured that reducing doctors would reduce disease. Question conventional wisdom, particularly when it’s derived from simplistic thinking. Respect the complexity of incentives. “You can’t imagine all the ways humans will connive to beat a system.” And finally, use your data to guard against cheating. “Just as companies don’t sufficiently capitalize on the data they have access to, they aren’t exploiting ∑ opportunities to think about fraud or theft in their businesses.”

Inc. Aug 2005