The Problem with CAFE

As I have been reading reports of the current debate over the pending energy legislation, it occurred to me that there is a fundamental problem with the approach to CAFE standards. The Washington Post reported on the issue in today’s edition:

Senate, House Turn Focus to Energy Bills

Senate Majority Leader Harry M. Reid (D-Nev.) said after a speech to the Center for American Progress yesterday that the increase in auto-fuel efficiency requirements, known as the corporate average fuel efficiency (CAFE) standards, would be the most controversial part of the Senate package. It orders auto companies to hit a 35-mile-per-gallon target by 2020 and improve mileage 4 percent a year after that.

“I know that the auto industry is still wavering on this issue,” Reid said. “I met with the CEOs of the big three automakers last week, and here is what I told them: The debate on raising CAFE standards should be over. It will happen. And perhaps if they had joined us instead of fighting us these last 20 years, they might not be in the financial mess they’re in today.”

So what’s the problem? Isn’t raising CAFE standards a good idea? On the surface, yes. And I absolutely agree that our fuel efficiency in the U.S. is terrible and must improve. But the problem I see is that this attempts to address the issue in the same way that windfall profits’ proposals attempt to address the issue of high gas prices. There are plenty of high fuel efficiency cars on the market now. The problem is, people aren’t demanding them. They want their SUVs and big trucks. What people are really after here is a free lunch. They think increasing CAFE standards will make everyone else drive a fuel-efficient vehicle. Or, they think this will result in a dramatic boost in the fuel efficiency of those trucks and SUVs.

I see the point made by the auto industry, because this will force them to make vehicles that people aren’t demanding. This is incredibly inefficient legislation. There are other ways to increase the demand for fuel-efficient vehicles, rather than forcing auto makers to increase the supply (that doesn’t happen to be in demand). What this may do is restrict the supply of inefficient vehicles, thus boosting prices for them. And a secondary effect may then be that people will turn to more fuel-efficient vehicles. But it is an incredibly convoluted way to achieve that goal.

As I said, this is analogous to the windfall profits’ measures that have been debated. People see this as a free lunch. The oil companies will be punished by having some of their profits taken away, thus somehow resulting in lower prices for all (as the companies respond to this punishment?) It is legislation aimed at the wrong problem. Oil companies make big profits because there is high demand for their product. And fuel efficiency is low because there is high demand for trucks and SUVs. The current legislation being debated won’t change that.

There is no free lunch. Increasing CAFE standards is not going to increase the public’s desire to drive fuel efficient cars, nor is it going to result in a 35 mpg Ford Expedition.

27 thoughts on “The Problem with CAFE”

  1. In the same way that the auto makers have helped to manufacture demand for heavy, low mileage vehicles, they can help to manufacture demand for lighter, higher mileage vehicles.

    I just don’t buy the idea that the auto makers are supplying what the public demands. They have fought increased mileage standards in every possible public relations and lobbying mode by equating big and heavy with safety and freedom of choice. Their efforts have resulted in making the highways less safe for many of us who choose higher mileage and freedom to consume less oil. Let there be no confusion about their efforts to manufacture demand for products that bring them more profit per vehicle produced.

  2. I’m down on CAFE and ready to scrap it (more on that later) but to answer your last paragraph:

    “There is no free lunch. Increasing CAFE standards is not going to increase the public’s desire to drive fuel efficient cars, nor is it going to result in a 35 mpg Ford Expedition.”

    It’s a pass-the-buck thing, which is why this isn’t totally true.

    The way CAFE will increase the public’s desire is by (passing the buck) forcing the auto companies to put incentives in place.

    They might charge more for cars above the CAFE mean, drop the worst models, and use profits from the guzzlers to put lower prices, and rebates, on the most efficient cars.

    I guess you could think of it as a “feebate” system, only run by each auto company on its own models.

    That’s stupid and inefficient, but that’s not my main complaint against CAFE.

    I’m down on it because it is always in the future. It can always be adjusted (as with the current ethanol exceptions) or delayed, or modified. And certainly it does not press consumer choice now, this year.

    I would like to see energy and environment activists call for the dismantling of CAFE and the replacement with something more immediate, like gas taxes or feebates on current cars.

  3. Odograph – what Robert says is true, Americans will pay for performance, but not for added fuel efficiency. We have an affluent society, and people like to show their affluence in what they drive.

    Even the Japanese car companies have gotten in on the act. Toyota makes the Prius, but they also sell the Toyota Tundra pickup. Guess which one sells better and makes more money for Toyota?

    That is why a voluntary tax and rebate system won’t work. If you want to change consumer habits, we need to increase the cost of fuel or directly tax fuel economy, or both.

    You can’t discount safety either. If we were strictly talking about fuel efficiency and commuting, motorcycles make the most sense. Safety wise, they don’t.

  4. Come on, you should be over the safety thing, with all the evidence presented so far!

    One more: No Trade-Off Between Higher Fuel Economy and Vehicle Safety

    On the other, are you really misreading me? People choose in a market environment, and make economic choice based on price and utility.

    Utility may be size, or speed, or a fuzzy emotional issue … but it is always judged against price.

    People “want” but they play that against what they can “afford.”

    CAFE passes the buck, and makes Detroit, not congress change the question of what is offered, and what people can afford.

  5. I agree totally with the “Free Lunch” concept. Someone needs to educate the electorate that the only one who’s able to do anything about their fuel bill is they themselves.

    The other problem is the concept of “Jevon’s Paradox.” Briefly, it’s the idea that as you make a product or service cheaper through efficiency improvements, more often than not people will compensate by using more of the product or service. Often they will overcompensate and end up using more energy. It’s analogous to the observation that as TVs of a given size get cheaper, people don’t save money; instead they buy a larger TV. Driving has externalities other than fuel consumption and pollution, namely traffic congestion, sprawl, and traffic accidents which could all get worse if fuel economy improves.

    I think what needs to happen is that the efficiency improvements need to be driven by consumer demand. I would like to see a carbon “feebate” like the one proposed by the folks at the Carbon Tax Center. This drives behavior but eliminates the regressiveness of a straight fuel tax.

    In addition, there are some clear market failures in the efficiency department (i.e. people buy inferior products that cost more because they cost less up front) that could be addressed by efficiency feebates. Feebates seem like a much less blunt instrument than standards.

  6. Let there be no confusion about their efforts to manufacture demand for products that bring them more profit per vehicle produced.

    No, I agree that their motive is profit. No question. But I can see their point that the demand is for the large inefficient vehicles. After all, I can always buy an import and get great fuel efficiency. And if more people did that, then the domestic auto makers would be forced to adjust to what consumers are demanding. I think that what consumers think is that by demanding higher CAFE standards, they won’t have to pay any price at all, and we will just happily watch fuel efficiency climb. It may by making the inefficient vehicles more expensive, but as I said that is a convoluted way of boosting fuel efficiency.

    Cheers, Robert

  7. Or to the extent that a “feebate” results, they may think that someone else will take the incentive and buy a small car, resulting in a fleet improvement even as they careen around town in their huge SUV.

    Did something like that happen in the era of the Geo Metro? Did GM have to sell them at a subsidized price to correct their fleet average?

    Expect that again, if we get back-door and hidden adjustments through CAFE.

  8. Even the Japanese car companies have gotten in on the act. Toyota makes the Prius, but they also sell the Toyota Tundra pickup. Guess which one sells better and makes more money for Toyota?
    I agree with the point you are making, King, but, at least for May, you have the details wrong. Toyota sold just over 24,000 Prii, and the goal is to sell 200,000 Trundra’s a year, which would be ~16,700 a month. AFAIK, they are still short of the goal for the Tundra, although, in the Toyota way, they are closing in on it.

  9. I’m down on it because it is always in the future. It can always be adjusted (as with the current ethanol exceptions) or delayed, or modified. And certainly it does not press consumer choice now, this year.
    That’s right OG,
    For the immediate effect on consumer choice we have gas prices. Gotta love the free market, “where there is no free lunches”.

  10. Feebates are BY FAR the best approach, because they directly affect the key decision (initial purchase) and they work with consumer behavior instead of against it.

    It’s instinctive to attack fuel consumption by taxing fuel, but gasoline is generally not a discretionary purchase. We buy lots gasoline because our cars require lots of gasoline. Once you buy a certain car, your fuel consumption is pretty much locked in.

    Academic research and practical experience repeatedly show buyers overwhelmingly choose low upfront payments. This explains auto financing (especially leases), “no payments until 2009” TV loans and teaser-rate home mortgages. Back end costs such as fuel are way down the list, if they’re considered at all.

    Think of a feebate as an upfront fuel tax based on expected gasoline usage over a 150k mile life:

    15 mpg SUV – 10k gal – $10k fee
    30 mpg compact – 5k gal – No fee
    50 mpg hybrid – 3k gal – $4k rebate
    PHEV-40 – 600 gal – $8.8k rebate

    The above uses $2/gal for illustration, and produces feebates so large people would stand in line for efficient vehicles. A $2/gal pump tax, on the other hand, would have negligible impact as shown by demand inelasticity as gas prices rose recently from $1/gal to $3.

    Furthermore, feebates would produce only $80b of annual economic distortion (16m vehicles * $5k average feebate) vs. $300b of distortion for an ineffective $2/gal pump tax. And feebate distortion falls primarily on new car buyers, the group most able to handle it. Pump taxes, however, fall most heavily on less-affluent used car buyers because fuel comprises a larger part of their income. This is especially silly since used car buyers as a group have zero control over the type of car they buy — they’re restricted to the pool of cars being passed down from new car buyers.

    If we want to kick our oil habit it’s clear we must directly target new car buyers. A feebate accomplishes this more effectively and efficiently than pump taxes, CAFE, or any other system.


  11. Optimist – I can’t find a complete set of sales figures, but it looks like in 2005-2006 the Tundra was outselling the Prius by several thousand units a month. In 2007 that trend has reversed. I can tell you that Toyota (at least locally) promotes the Tundra more heavily than the Prius.

    Odograph – I was comparing motorcycles to cars. Small car vs. SUV safety depends on where you drive. I share my commute with big trucks packed into narrow lanes of bumper-to-bumper traffic. I don’t need a Hummer, but I want something more substantial than a small sedan. When we lived in a medium sized town I drove a Toyota Camry, 30+ mpg. In the UK I had a Ford Ka and a Mercedes A class. Loved both of them.

    I also have to deal with illegal immigrant bad drivers who often don’t carry insurance.

  12. King, there might be some situations where an SUV would be a net win. I drove a couple jeep cherokees for years without rolling them. But I also knew their limits (esp. the 1984 model made before the rollover controversy).

    The people most at risk are those who transition from a car to an SUV without understanding the new vehicle dynamics.

    You see them all the time, they are the ones going too fast, not giving themselves extra room for braking, and expecting to “maneuver” in response to danger.

    Or, sometimes you just see them lying on their side in an intersection.

  13. Someone told me that Flexfuel vehicles are not counted in the CAFE calculations. This is why you see pickups and Explorers with flexfuel stickers. Is this correct?

  14. Odograph – I drive a Jeep Liberty. It is a good commute vehicle, but mileage is not great. I get around 19 mpg, by driving judiciously. Now that it is off warranty, I’m considering aftermarket devices to improve fuel efficiency (modified air intake, exhaust, changing internal program settings).

    My commute is 80/20. The first 12 miles easy residential/highway. The last 3 are bumper to bumper congested interstate. At one point this morning I was surrounded on all 4 sides by big rigs. Then a lady in a Ford E series delivery van rode my bumper, literally 6′ behind me for more than a mile. I try to leave the 2 second cushion between myself and the vehicle ahead, but if you do people will cut into it. We have agressive lane changers here. They cause a lot of accidents and don’t really get there any faster.

    I would consider a hybrid for my next car, maybe a Ford Escape. I’ll be in the market again in about 2 years. Fuel economy will definately figure into our decision.

  15. I can tell you that Toyota (at least locally) promotes the Tundra more heavily than the Prius.

    But maybe that’s because Toyota is selling Prii as fast as they can make them, without needing to promote them. If this is true, any money spent promoting them (Prii) would be utterly wasted.

    I don’t watch a whole lot of commercial TV so I’m not the bext judge, but I don’t believe I’ve seen a single TV ad for the Prius in the 3 or 4 years since it was introduced. I *have* seen multiple news stories saying that Toyota dealers (at least in the past) haven’t been able to keep up with demand for Prii. And, anecdotally, this was repeated by the saleman at the local Toyota dealership when we were buying my dad’s new pickup (Tundra or Tacoma, can’t remember which) last year.

  16. Mike C – would agree that in the last couple of months Prius (is Prii the plural?) has been selling well. I check inventory here in the Houston area every couple of months. At the end of 2006 there were about 800 available at Houston area dealers. Today there are only 64.

    Toyota promotes Prius somewhat as part of it’s image advertising. But not the heavy radio/TV newspaper model promotion.

    Curiously, Prius sells pretty well but the Camry hybrid not as much. Honda is discontinuing the Accord hybrid. This leads me to believe the market is for first adopters and “conspicuous envirnomnentalists”. This might also explain why they don’t heavily promote the brand. Heavy promotion probably doesn’t increase the Prius customer base, but is likely effective in getting someone to switch from Ford or Chevy trucks to a Tundra.

  17. The earliest (Prius I < 2004) was probably for those conspicuous environmentalists. The (Prius II >= 2004) has been making a bit of a transition. When I bought mine in 2005 the question they asked a the dealer was “do you drive a lot?”

    Toyota recently decided they could push Prius to the genuine mainstream, and started new advertising for that. I believe they did get it in the top ten models sold in the last published-month’s data.

    As far as which models succeed and which fail … I think the market figured out something sooner than the car companies: The reason you buy a hybrid (generally) is to get 50 mpg. The two (4 passenger) cars that will do that are the Prius and the Civic Hybrid. Within that group of 2 the Prius has a couple advantages. It is slightly larger than the Civic (a bonus in space for the same mpg) and has a fold-down rear seat (another bonus in space).

    People aren’t dumb. The Prius gives you the most for 50 mpg, and it might be holding onto its hype partly for that reason.

    As a Prius owner it strikes me that the car could easily serve a huge share of buyers! No reason it shouldn’t be the top selling car in America.

  18. Odograph – Figured you for a Prius owner. What is your actual mileage? I’ve heard it can vary depending on your driving habits.

    Until recently CA allowed single occupant hybrids to drive in the HOV lanes. I understand they ran out of special licenses. If that was true in TX I would consider trading my Jeep in on one. It would cut about 15 minutes per day off my commute. Since big trucks can’t run in the HOV lane, I have less of a safety concern. My 1987 Toyota Camry wagon got about 33 MPG. If I could get a conventional sedan that got 30-35 MPG it would be hard to justify a hybrid.

  19. I was down to 46-47 when i did a too-short commute. I’m back at 49-50 now that they moved the office.

    I wrote a bit about the dynamics of prius mileage here.

    If you work with oil plants I’m sure you can relate to the ‘control problem’ they have, and the reason they run the engine more on short trips.

    (one benefit on my new commute is that I can use my stickers on jammed up days – maybe a couple times a monty)

  20. Yes, I can relate to the control problem. Some are promoting plug in hybrids. Do the gasoline engines start up immediately on PHEVs? I seem to remember that PHEVs also modify the computer programs.

    I like the technology – particularly the CVT transmissions. US automakers are starting to employ CVTs on more models. This move alone could increase fuel efficiency by 20% or more. CVTs had a problem driving heavier, high torque vehicles, but better materials have improved the durability of CVTs. I see they are being offered on 3 liter V6 engines now.

    I think my next commuter vehicle will be a truck. I’ve never owned one before. Looking at the 2.3 liter Ford Ranger with manual transmission. Gets 29 MPG highway – not too bad. Can pick up one for about $13,000 before haggling. Just have to convince my wife.

  21. Curiously, Prius sells pretty well but the Camry hybrid not as much. Honda is discontinuing the Accord hybrid.
    It should be pointed out that the Accord hybrid was a “performance” hybrid, as opposed to an “economy” hybrid (everything else). I guess Honda figured they’d take the new technology and do what auto companies have been doing for the last two decades: make the car faster, as oppossed to more efficient. Well, people connect “hybrid” with “economy”. I suspect when customers noticed there is almost no improvement in efficiency with the Accord hybrid, most of them concluded it was not worth the investment.

    Honda made a strategic error! Who knew they could?

  22. Accord hybrid – you are correct. It has essentially the same fuel economy as the 2.4 liter, 5-speed manual Accord Sedan. (Hybrid: 28/35 vs. Sedan 26/34) Yet at $31,000 it costs $13,000 more than the base Accord, and about $10,000 more than the Prius.

    It will be interesting to see how many units are sold of the same vehicle with the hybrid drives. Camry has both, Ford Escape 2008, GMC Yukon, Lexus. Will people spend as much as $7,000 for marginally better fuel economy?

  23. “Do the gasoline engines start up immediately on PHEVs?”

    The Toyotas have different systems than the Hondas. The former can have the engine fully “off” but the latter idle the engine when the electric motor is taking the load.

    Generally my Prius engine will start as I get rolling (10 or 20 feet?), but get this (control problem again) your cabin temperature controls affect this was well. If the user requests a warm car on a cold day, the computer will run the engine to get some heat.

    I’ve noticed that I’m stopped at a light an my engine is still on … what’s the deal? … turn down the heater, off goes the engine.


  24. The control systems in the Prius are essentially “dumb” in that they only sense the state of the engine and battery and then make adjustments. When you kick on the A/C or heater the Prius fires up the engine to conserve on battery.

    I think the way to go is an electric hybrid tied into a satellite navigation system. I would make the drive train totally electric and use a smaller engine strictly for recharging the battery. When you left your house in the morning you would tell the navigation system where you were going, and it would figure out whether it needed to turn on the engine or not. You could have presets for common destinations (like on a radio) such as home, work, grocery store. The system could monitor fuel usage and tell you where the nearest gasoline or quick charge station might be.

  25. I am beginning to believe that the future of transportation is powering a Solid Oxide Fuel Cell with same old gasoline and diesel we’ve been driving on. Seems BMW is already working on it.

    The fuel cell can be used to power a highly efficient electric motor. All the benefits of high energy density of hydrocarbons and highly efficient electrics combined! And you won’t need a battery anymore!

    And do take note of this: Solid oxide fuel cells have so far been operated on methane, propane, butane, fermentation gas, gasified biomass and paint fumes. Sounds like she’ll take anything that burns!

  26. Well, I said “werid” as in weird experience. I’ve programmed motion control systems and can imagine how it works internally. It would be easy enough to do “weightings” for the engine on condition. A big differential between cabin temperature and requested temperature (for heating) seems like an obvious waiting.

    Programming those “take me home” functions sound a lot harder ;-), but there are a generation of kids trying with the DARPA challenge. In 40 years that could be big.

Comments are closed.