In Search of Recession Causes

The message of the last chapter is that a persuasive causal inference from nonexperimental data requires the following:

• (1) Statistically clear temporal orderings from historical data (first this, then that)

• (2) An unambiguous but-for hypothetical (an intervention, a change of the past)

• (3) A compelling story that turns observed correlations and temporal orderings into beliefs about what might have been

Let us see if we can find these three ingredients and cook up some conclusions about the causes of recessions. This is not going to be easy. We are entering a land ruled by Uncertainty and Ambiguity. If you expect to be comfortable here, you will need Faith.