> First, is there any basis for assuming a more fuel efficient car is in fact more expensive to produce on a per-unit basis? Great question! There's actually s

> First, is there any basis for assuming a more fuel efficient car is in fact more expensive to produce on a per-unit basis? Great question! There's actually some pretty good literature on it from energy and policy economists. The auto industry, the EPA, and the NHTSA all pull from very similar general background estimates. "A 1991 study performed by Greene[1] concluded that, for every one mile-per-gallon increase in vehicle fuel economy, the average, per-vehicle cost would be within the range of $100-200 dollars, in constant 1985 dollar terms, and corresponding to a range of $200-400 in 2010 dollars. A review of CAFE by Klier and Linn[2] found that numerous other studies have confirmed these figures. Using the range found in the literature, the projected 12.7 MPG improvement in fuel economy would cost on average somewhere between $2,500 and $5,000 per vehicle." So... that's the industry perspective. But that's a simple review of the background literature, so it's not that radical. The EPA and NHTSA dip only slightly below the low end, getting down to $2,000 using additional assumptions on that baseline, such as learning curves from past R&D making newer developments cheaper. But the regulators still estimate a positive cost per vehicle. The debate between regulators and industry is the extent of that cost, not whether that cost exists at all. It's a fascinating area of research, notwithstanding the minefield of biases one has to sort through. [1] Greene, David L. (1991). “Short-Run Pricing Strategies to Increase Corporate Average Fuel Economy.” Economic Inquiry 29(1): 101–114. [2] Klier, Thomas and Joshua Linn. (2011). “Corporate Average Fuel Economy Standards and the Market for New Vehicles.” Annual Review of Resource Economics 3(1): 445-462.