There seems to be some confusion about the difference between welfare loss and GDP loss. Welfare loss (in partial equilibrium) is approximated by the Harberger Triangle. In the graph above C.
Partial equilibrium GDP loss is how much smaller the market is. In the graph above that is A - B. We measure the size of the market as quantity X price.1
It is entirely possible for the Harberger Triangle to be large and the GDP loss to be negative. That is, a tax could decrease welfare but increase GDP.
Moreover, in general equilibrium GDP loss can only come from decreased production. For most of our models that means that either there has to be less labor supply or less capital. Same labor, same capital implies same GDP.
Why is this important.
First, in most cases I believe GDP losses are significantly smaller than welfare losses. I may be able to prove this always true but I have to think about it.
Second, the standard answer that economists give when talking about the losses from taxation, tariffs, regulation, etc is a welfare measure. Yet, policy makers hear a GDP measure. In the back of policy makers minds they are already converting GDP to welfare. If you give them the wrong number they will convert twice and really overestimate the loss.
(1) Just as in Harberger case we assume that taxes are lumped summed back and spent on something else.
Monday, May 7, 2007
Posted by Karl Smith at 5:20 PM