The wonder of DSGE

Two charts showing the American Recovery act. In the first we have the ARA matched against the log of gdp growth. In the second I removed the log function. If we assume the exponential growth mode, then the ACA looks aligned with growth, top chart.  Actually it looks like ACA could have caused immediate growth, or it could have caused growth three years later, we do not know since exponential growth is a complex variable and has lags.

If we ignore the exponential assumption, we see the ACA had little affect, as in the second chart.  There is seems growth reached equilibrium and stayed there, ACA having no effect.

We know that total spending from DC stayed the same, after interest expenses are taken into account. ACA affects are likely an illusion. Why are economists so confused? In an undersampled system we get dead cat bounce. So the Romer method of finding multipliers will, by its nature, identify dead cat bounces. If Romer then assumes the economy is completely sampled, DSGE applies and the log growth operation accentuates changes.  Applying some exogenous change in spending, the changes will naturally line up and fool the economist.

What happens when interest rates suddenly drop and DC passes out money to the states? The money goes into the pension funds to make up for lost interest income.

So I vote a banana to John Taylor who first exposed this problem, though the poor fool still believes in DSGE.