But from the beginning, critics were already pointing out that cash for clunkers might not work as planned. For one thing, the fuel-economy requirements were fairly lax: A person could, in theory, trade in a Hummer that got 14 mpg and get a $3,500 voucher for a new 18-mpg SUV. What’s more, that slight gain in efficiency would be partially offset by the energy costs involved in manufacturing the new car. And on the economic front, critics argued, the program might just move up purchases that would’ve happened anyway — thereby providing little actual stimulus.

So were the naysayers right? It seems so. A newly updated analysis from economists at Resources for the Future finds that the actual benefits of the program were pretty meager. The paper examined U.S. car sales using trends in Canada as a control group, and estimated that about 45 percent of cash-for-clunker vouchers went to consumers who would have bought new cars anyway. In the end, the program boosted U.S. vehicle sales by just 360,000 in July and August of 2009 and provided no stimulus thereafter. What’s more, the program increased average fuel economy in the United States by just 0.65 miles per gallon.