Most people interested in economics (or in scandals) have heard of this – the famous conclusions of Kenneth Rogoff’s and Carmen Reinhart’s analysis of data about the relationship between public debt and GDP growth have been found wrong. They were based on flaws in handling the data base and in the application of statistic methods. There is a lot to be learned from this “scandal”, but I personally found Barry Eichengreen’s commentary particularly insightful. He calls for open-access to the data and algorithms used by researchers, as well as more reservation regarding causality interpretations.
Statistics are helpful. But in economics, as in other lines of social inquiry, they are no substitute for proper historical analysis.
In impugning the authors’ motives and criticizing the uses to which others have put their research, critics of Reinhart and Rogoff have taken their eye off the ball. The real problem is scholarly procedures and priorities, not motives. If the problem of procedures and priorities is addressed, the fact that politicians are tempted to misuse scholarly analysis for their own ends will take care of itself.
In other words, what is true of the economy is equally true of economic analysis. A crisis is a terrible thing to waste. [more]