To help illustrate this point, here is an interesting chart from Clusterstock that shows the effect of the federal government's "cash for clunkers" program. The chart shows that a significant portion of the currently reported GDP is a result of this program:
according to the BEA the spike you see [in the chart] added 1.66% to the U.S. GDP growth figure reported. Thus without it, GDP growth would have been only 1.89% (3.5% - 1.66%) in Q3.However, they correctly point out that all the program did was rob from future demand. In other words, individuals that would have purchased cars next quarter, purchased them now. Therefore, although current GDP is up, future GDP will be correspondingly down.
Another interesting CNN article reports that auto sales analysts at Edmunds.com performed a study which shows that the government actually spent $24,000 per car - not the $4,500 per car that was the headline per car subsidy. Why? Because many of these people would have bought new cars anyway at some point in 2009. They calculate that only 125,000 of the cars sold under the program were cars that would not have ultimately been purchased with the subsidy. Therefore, the government's expenditure of $3.0 Billion divided by 125,000 cars works out to $24,000 per car.
Next quarter, we won't just be returning to business as usual for auto output. Don't forget that Cash for Clunkers pulled future auto demand, ie. some of Q4 demand, into Q3. Thus Q4 is likely to be very weak since many people who planned to buy a car in Q4 probably took advantage of Clunkers and bought in Q3.
Next quarter, not only are we unlikely to get Q3's boost, but motor vehicle output data could subtract from GDP as well.
Thanks government. Good work!