Most of the 677,842 sales were simply taken from the near future. That many older vehicles were traded in — and, as required by law, destroyed. Gayer and Parker accept as reasonable an estimate that the cost per job created by the program was $1.4 million. Although the vouchers did not come close to covering the cost of the new cars, voucher recipients seem not to have reduced their other consumption. This, say Gayer and Parker, suggests that participants in the program “were not liquidity constrained,” which is a delicate way of saying “there was no change in other consumption patterns,” which is a polite way of saying that “cash for clunkers” merely caused people to purchase vehicles “slightly earlier than otherwise would have occurred.”

Because the program was not means-tested, it had only a slight distributional effect of the sort progressives favor: Voucher recipients had lower incomes than others who bought new cars in 2009. Against this, however, must be weighed the fact that the mandated destruction of so many used vehicles probably caused prices for such vehicles to be higher than they otherwise would have been, meaning a redistribution of wealth adverse to low-income consumers.

As for environmental benefits from Cash for Clunkers, the reduction of gasoline consumption was small and “the cost per ton of carbon dioxide reduced by [the program] far exceeds the estimated social cost of carbon.” But it was — herewith very faint praise — more cost-effective than the subsidy for electric vehicles or the tax credit for ethanol.