I have written a lot about the Cash for Clunkers law, and the fact that it was a hit with its beneficiaries because it bought cars that blue-booked for just under $1500 for two or three times that amount. Other studies have shown that the program did abate some CO2, but at ridiculously high prices per ton.
But I have found a reason to love the Cash for Clunkers program: it is a fabulous demonstration project for just how utterly pointless government stimulus programs can be. Stimulus programs tend to be hard to evaluate in our complex economy -- sort of like trying to calculate the effect of a butterfly flapping its wings on world climate. But since cash for clunkers only lasted a few weeks and hit only one industry, we can learn a lot about the effectiveness of government stimulus.
Here is the US Census data for auto dealer sales (source). Thanks to my friend Scott who first pointed me to the analysis:
The dotted line simply averages the sales for the month of the clunkers program and the month after. I think it is pretty clear that we spent a few billion dollars making some used car owners happy (by overpaying for their vehicles) but did absolutely nothing to move the trend line in auto sales, as the program appears to have just pulled forward purchases rather than stimulated new ones.
Update: Welcome Instapundit readers. This is all in the family blogging day, as my son just started up his own blog with a post ranking baseball players. Feel free to give him grief for being a Yankees homer.