My attempts to read _How Much and at What Cost_ have ground (er) to a halt (oh, man, the transportation puns. They are foolishness.), primarily over my Issues with the failure to model demand changes in response to price signals. After some discussion with my husband R., who said this made the conclusions unreliable, I went looking for someone who _had_ modeled demand changes in response to price signals for fuel -- I mean, it's happened, so it can be quantified. A google search found me a frighteningly wonky book aimed at transportation planners, _Transportation Decision Making: Principles of Project Evaluation and Programming_, which I have ordered (used), which was probably stupid because whenever I order textbooks, I find them largely unreadable. But it definitely has tables with short- and long-term demand response to price signals on fuel. Also, it has some hilarious paragraphs about accounting for pedestrian and cyclist operating costs. No, not the bike lanes: the shoes and the bike and accessories for the bike. Even funnier, they _do not_ account for the extra cost in food for the human-power part of the walking and biking -- they explicitly say that counts as a _benefit_, not a cost, particularly if it replaces other exercise. !!!
I agree, actually, but will note that when economists are coming up with numbers like GDP, it is _not_ to the GDP's good for someone to quit spending money on fuel, a car, and a gym membership, and replace it with spending (much less) money on shoes and/or a bicycle, better outerwear, etc.
Because food and fuels can both be quantified in kCal, it is actually possible to figure out human mpg to compare to a car (but somewhat time consuming).