Thomas Friedman’s usually pitch-perfect commentary on energy and security hit some high notes Sunday, but it also went off key twice, in disappointing ways.

First, the sweetest passage from his New York Times column:

Reader support makes our work possible. Donate today to keep our site free. All donations TRIPLED!

By doing nothing to lower U.S. oil consumption, we are financing both sides in the war on terrorism and strengthening the worst governments in the world. That is, we are financing the U.S. military with our tax dollars and we are financing the jihadists–and the Saudi, Sudanese and Iranian mosques and charities that support them–through our gasoline purchases. The oil boom is also entrenching the autocrats in Russia and Venezuela….Finally, by doing nothing to reduce U.S. oil consumption we are only hastening the climate change crisis.

Now, the ear splitters:

Grist thanks its sponsors. Become one.

We need a gasoline tax that would keep pump prices fixed at $4 a gallon, even if crude oil prices go down. At $4 a gallon (premium gasoline averages about $6 a gallon in Europe), we could change the car-buying habits of a large segment of the U.S. public, which would make it profitable for the car companies to convert more of their fleets to hybrid or ethanol engines, which over time could sharply reduce our oil consumption.

What’s wrong with that? I like the sound of it and the general drift. And I don’t take issue with very high fuel taxes, if they are offset with cuts in conventional taxes and if their regressivity is offset.

But Friedman goes sharp here in one important detail: he suggests that fuel taxes are mostly about changing car-buying habits. In fact, they’re a pretty poor instrument for influencing purchases. (They do a better job of influencing day-to-day driving decisions, but even there, other things might be more effective, such as pay-as-you-drive insurance.)

The reason fuel taxes fall down as an incentive for buying more-efficient vehicles is that consumers are notoriously short-sighted in their estimates of the value of future energy savings. Typically, consumers will only pay extra for vehicle fuel economy if the investment pays itself off within three years. (In economic lingo, their "discount rate" is well about 50 percent for future energy savings.) According to Rocky Mountain Institute (huge pdf, see page 139), consumers underinvest in fuel economy by roughly 60 percent–at any given fuel prices.

Grist thanks its sponsors. Become one.

So simply raising fuel taxes, as Friedman suggests, will certainly factor into consumers’ calculus and encourage them to buy slighly more efficient vehicles. But it won’t have the mammoth effects he suggests. The cost of fueling remains fairly modest: perhaps one-eighth of the cost of owning and operating a vehicle. Hiking fuel taxes might raise that fraction to one-fifth or even one-fourth, but it won’t make fuel-economy the overriding factor in purchase decisions.

As RMI writes, “The difference between a 30-mpg and a 40-mpg car, though consequential for society, will seem to the short-sighted buyer to be worth only about the price of a set of floor mats.” So, if Friedman gets gas up to $4 a gallon, the difference may double to about the price of two sets of floor mats. But that’s not going to set off a dramatic change in car buying habits.

Fuel taxes won’t close the yawning chasm between consumers’ payback requirements and the economically justified level of efficiency investment. But feebates can: they make the lifecycle energy costs of new vehicles powerfully evident in the one price that matters most to car buyers–the sticker price.

These point-of-purchase incentives charge fees to the buyers of inefficient vehicles and refund the resulting revenue as rebates given to the buyers of efficient vehicles. The fees and rebates are directly proportional to the efficiency of the model, so there’s a continuous and market-wide upward tug on fuel economy. They’re on the docket in Canada, the United Kingdom, France, and — most recently — Connecticut.

So here’s Friedman rewritten:

We need a gasoline tax vehicle feebates to turbocharge gains in fuel economy. that would keep pump prices fixed at $4 a gallon, even if crude oil prices go down. That we would change the car-buying habits of a large segment of the U.S. public, which would make it profitable for the car companies to convert more of their fleets to hybrid or ethanol engines, which over time could sharply reduce our oil consumption.

Next, Friedman goes flat by most of an octave, inserting an ill-informed note on nuclear power: “We need to start building nuclear power plants again. The new nuclear technology is safer and cleaner than ever."

Nukes have largely died from market forces, so why any market-savvy thinker like Friedman would want to resurrect them (presumably through further subsidy) I cannot fathom. Our energy future should flow from least-cost planning that tries to incorporate social and environmental costs such as climate change and national security.

Under least-cost planning, nuclear power does not fare well. Efficiency and renewables beat nuclear power on price alone in almost every application, because nukes are extremely expensive. Add environmental and security costs and nukes fade to a footnote. Generating electricity through a technology that creates wastes still deadly for thousands of years seems like folly. Peppering the countryside with such facilities in an age of global terrorism seems, well, like lunacy. (A recent National Academy of Sciences report on the terror threat of one aspect of A-plant management was summarized in the Washington Post yesterday.)

Fortunately, Friedman finds the tune again at the end. I’ll give him the last word:

And we need some kind of carbon tax that would move more industries from coal to wind, hydro and solar power, or other, cleaner fuels. The revenue from these taxes would go to pay down the deficit and the reduction in oil imports would help to strengthen the dollar and defuse competition for energy with China.