The White House Council of Economic Advisers (CEA) has released a projection of jobs created by the economic stimulus bill. However, the method they used to get these numbers falls short of basic scientific standards. The CEA modeled a potential outcome for the economy without the stimulus, basing it on historical data. They then compared this computer simulation with the actual data about what occurred with the stimulus – and have declared that the stimulus worked.
This method would be met with a failing grade in any decent university economics course. A projection of a potential economic outcome based on a model must be compared with a baseline scenario produced by that same model. This is basic: the policy simulation must be compared with a baseline. Instead, the White House compared a simulation with empirical data!
This is absurd because any slightly different assumption in a simulation will produce a different result. When comparing with a baseline these can be controlled—we can see what assumptions went into the baseline scenario and which went into the simulation of the policy. There are no hidden tricks. But if a simulation is compared to empirical data, we have no way to know why the two differ. It could be anything.
What were the assumptions in this simulation, I wonder? Do they not know that they must use a baseline—or was this a purposeful political maneuver?