Each of the authors posted the graph and said (essentially), “see we have too much debt” and subsequently suggested that the monetary policy advocated by Beckworth and Ponnuru would make things worse. David and Ramesh have responded to this criticism by defending nominal income targeting. However, I would like to put monetary policy aside for a moment and talk about data and how we should use data.
The graph illustrated above shows the path of household debt as a percentage of income. Joe and Nicole each say that there is too much debt. Okay, but how much is too much? In other words, even if I were to agree with Joe and Nicole as to idea that there is too much household debt at present, how would I know when it had sufficiently fallen to an acceptable level? Saying there is too much debt presumes that there is an optimal amount of debt and that current levels exceed that optimal amount. In order to assess whether or not there is too much debt, it is therefore necessary to have some type of framework for assessing debt levels. Such a framework would presumably require a sufficient understanding of banking, intermediation, liquidity, household preferences, etc.
When reading this, I couldn’t help but to think of this quote from Milton Friedman:
I don’t believe that statistics, as somebody has said, statistics do not speak for themselves. Alfred Marshall once said, “There is no person, no theorist so reckless as he who says that the facts speak for themselves.” The facts never speak for themselves. They have to be interpreted in terms of some understanding of where they come from and what the relation between them is.