Recently there have been arguments in favor of quantitative easing (QE) in the US based on the use of counterfactuals, i.e. arguments based on questions about what would have happened to GDP or unemployment had QE not been in place. I argue in this post that such arguments naively imply unjustified causation.
Often counterfactuals are employed to establish or denote causality. Here is a typical example:
If the light switch is not turned on, the room will be dark
The above counterfactual is true based on statistical evidence gathered through inductive reasoning of the following kind:
(1) The room is dark
(2) Turn the light switch on
(3) The room is no longer dark
(4) Observe that happening many times
(5) Then turning the light switch on causes the light to go on and the room is no longer dark.
However, there are exemptions to this rule, although it is true with probability close to 1. Suppose that a high power transmission takes place near the room and there is induction, a phenomenon where power is induced to the electrical circuit even though the switch is off. Then, the light can go on without turning the switch on. Anyone who has operated a transmitter knows this.
One of the many problems in using counterfactuals in justifying QE is that there is only one data point. In order to establish causation one needs many data points to calculate confidence intervals. The next time QE is employed, due to the lack of established causality, the outcome may be different. Thus, any attempt to justify QE on the basis of counterfactuals amounts to naive claim unless causality is established within reasonable bounds.
I also believe that QE was very helpful in averting deflation but beliefs worth little unless they are justified. And justification is difficult in the case of counterfactual thinking unless there is statistical significance, which cannot be obtained with only one event, i.e. its application during the current crisis.