There’s No Such Thing as a Free Lunch (Or Gas Mileage)

Would you call this a clunker?

Setting aside my esteemed co-writer’s obsession with hypermiling (not practicing, mind you, joking about), there really is no such thing as a free lunch – especially when it comes to fuel economy.  The Government, however, as demonstrated by recently finalized mileage laws, wants their lunch on the house.

As the saying goes, “you can’t legislate innovation”, and fuel economy standards give car manufacturers very interesting incentives.

Free Lunch and Free Beer

When people discuss receiving things for ‘free’, one common retort is “free as in beer, or free as in lunch?”.  Beer – in a sense – is considered a possible free item.  Lunch, on the other hand?  Not a free possibility.

The phrase “ain’t no such thing as a free lunch” dates back to the early years of last century, but it was popularized both by Science Fiction writer Robert Heinlein and Economist Milton Friedman.  Friedman’s formulation basically said that even though the costs might not be known, they exist – and they might not be monetary.  While ‘free lunch’ offers might come, at best, with a 2 drink minimum (with inflated prices!), at worst they come with some sort of commercial pitch.  The same concept applied to cars?  An increased fuel economy might come with reduced car weight or new designs but at the expense of something very important – passenger safety.

Materials Science

In my discipline, Computer Engineering, there is a sarcastic quote often used as an argument during any requirement gathering session: “Cheap, Fast, High Quality.  Pick any two.”.  I’m no mechanical engineer, but I do know that the fastest way to gain fuel economy is to reduce weight.  One EPA study pegs the number at around a 1-2% increase for every loss of 100 pounds.  Right off the bat, that’s where you see the first problem – momentum is mass times velocity.  Perhaps a compact car would fare poorly against a larger vehicle either way, but reducing weight and engineering well designed crumple zones can only get you so far.  New materials don’t necessarily help – deceleration injuries can be just as fatal as crushing injuries.  So yes, exotic materials will help – to a degree.  But momentum is a physical law which we can’t quite bypass (insert witty Higgs Boson comment here).

So, in a nutshell, early generations of cars working under new efficiency standards will be at a disadvantage to older cars still on the road.  One of the funniest things about Cash for Clunkers is the beneficial effect it had on clearing older, heavier cars off the road – there are less hulking anachronisms for compact cars to hit than there were 5 years ago.  Another important thing to note is that safety devices themselves add weight.  Air bags (which attack that acceleration problem discussed above) are heavy installations – they require an expensive package of materials and a gas propellant which weights something more than zero pounds.  Same for active suspension systems, seat belts, tensioners, and the dozens of other pieces of safety equipment.  In fact the only negligible safety equipment weight comes in the software on on-board computers.  So, where does a lot of the base weight come from?  The frame and superstructure, of course.

Take the Ford F-150, for instance.  Ford, in an effort to comply with new regulations, is redesigning the car from the ground up and basing their frame design on aluminum, not steel.  Now, don’t get me wrong – there are some very strong alloys of aluminum (such as the 7000s, used in aviation) – but at the same specification (assumed since the goal is weight reduction) steel is stronger than aluminum.  This is on top of that whole momentum thing I discussed earlier.  Will reduced weight necessarily make the truck more dangerous?  Yes, but I do expect safety technology to eventually catch up with whatever weight we aim for – but I don’t expect it to come for free.  (Not to mention, those same safety features would also be effective with current vehicles).  I am in no way saying an F-150 will become “dangerous”.  As one of the larger vehicles on the road, it will still have advantages in a collision with most traffic.  I think the largest effects will be on compact, sub-compact and mid-sized cars.  In fact, counter-intuitively, perhaps reducing weight in our current bifurcated fuel system (cars vs. light trucks) would actually make driving safer, even if making SUVs and trucks more dangerous?

Quantifying the Effect of Fuel Economy

Luckily for us, there was a paper covering this very topic on NBER.  Not only does it cover fuel economy changes, but also models driver preference – perhaps a dangerously driving Suburban Mom switched from a compact car to a land beast (making us less safe, in aggregate).  The author’s conclusion?  Every 1 mile per gallon increase in requirements leads to an excess of 149 deaths a year.  Currently, cars average 27 miles per gallon.  In 2016, the mandate is 34.  In 2025?  54.5 miles per gallon (but the measure should be closer to 43).  You do the math.

Maybe Google can save us with this?

Have you ever considered safety when considering mileage rules?  What do you think of the numbers above?  Do you think that the benefits in saving fossil fuels, including what may very well be lives saves through indirect effects, are outweighed by increased driving fatalities?  Have you ever had a free lunch?  How did you pay for it?



  1. freeby50 says

    In 1980 cars averaged 16 MPG. In 2009 the average was 23.8 MPG.
    In 1980 fatalities per million mile was 3.3 and in 2009 in was 1.1.
    So in a 3 decade period cars got 50% more efficient and fatalities per mile dropped by 2/3.
    If we can do that for 30 years I don’t see why the trend would reverse all the sudden…

    I don’t know why people feel a need to argue against fuel efficiency. Why does increased MPG make people want to find fault with it? Yes if you drive around in a friggin tank you’ll win a collision with someone on a bicycle.. Safety isn’t about who has the bigger car. Have you seen a 1959 Chevy have a head on collision with a 2009 Chevy?

    • says

      There’s actually a hitch you can see in the early 80s after the first CAFE standards – for a time fatalities per mile did change (I’ll link it to Joe’s reply).

      I’m not denying that cars got safer – tire and wheel technology, suspension (active and passive), frame structure (and for small cars, crumple zones), computers – I’m saying that that stuff would have happened anyway. It’s a similar argument to my stimulus article. I could draw a second line without fuel economy where the number of highway fatalities would be even lower. Yes, it’s great that the line is down and to the right. I’m saying that the slope could be even greater.

      I’m not arguing against fuel efficiency. I’m arguing that it’s ridiculous to mandate innovation in fuel economy. First, gas is elastic. Look at miles driven by year around the gas peaks in 2008. I have argued here that a gas tax makes more sense than CAFE standards – let people decide if they’ll pay the penalty. (Disclosure: you know I’m a Pigovian). Second, cars may reduce in size compared to last year’s efforts, but not compared to the stationary objects. It’s safer to hit large things with large things.

      I watched Youtube on those a while back – even 90s Volvos (regarded by many that I know as a “very safe car”) didn’t seem very safe compared to today. It’s the effect of all those other innovations, which I argue would have happened anyway. I think it would be better to compare a land tank from 2009 with a subcompact from 2009.

      I have other examples than cars, of course. Stay tuned, because you know I’m down to write an article on washing machines and toilets!

      • freeby50 says

        One point about the physics argument:

        ” It’s safer to hit large things with large things.”
        Its safer to hit large things with LARGER things. (bigger wins)
        Its safer to hit smaller things with large things. (tanks driving over bicycles)
        But its safer to hit stationary things with smaller things.

        Its safer to hit a brick wall with a smaller car than a larger car. Damage to occupants in vehicle collisions with stationary objects is dependent on the force of the collision which increases with the mass of the vehicle.

        Bigger cars are safer when they’re running into little cars. You’re bigger so you win. But bigger cars are not safe running into brick walls, trees, etc

        Thats assuming you hold everything else equal. But generally when you compare big cars to little cars everything isn’t equal. Small cars are cheaper and have less features in general versus larger cars which are usually more expensive and have more robust safety features.

        • says

          You’re right, but with caveats. Here you go, straight from the EPA’s summary of the NHTSA findings on vehicle size:

          • in crashes with a stationary object additional mass may be sufficient to knock the object,
          such as a tree or pole, down, allowing the vehicle to continue moving and reducing its delta
          V than if it was completely stopped by the object. In a previous study NHTSA estimated
          that the object is knocked down in about 25% of frontal collisions with stationary objects
          (Partyka, 1995).

          Also, we should mention that momentum matters, yes. But deceleration injuries are force/time, so the length of the vehicle being crushed also factors into the equation. But 2 vehicles of the same size with the same crush space and properties, the lighter one is better.

          Another interesting thing – cars weigh a lot more than they used to (slide 5). I chalk a lot of that up to safety innovations that weren’t software.

      • freeby50 says

        ” I’m arguing that it’s ridiculous to mandate innovation in fuel
        economy…. I have argued here that a gas tax makes more
        sense than CAFE standards – let people decide if they’ll pay the penalty”

        I can certainly see the argument that gas taxes make more sense than mandating standards. To me it comes down to where you apply the change on the supply side or the demand side.

        However I think that you can mandate innovation in fuel economy. Honestly I think the car companies are certainly able to improve efficiency but don’t do so unless required. They argued for decades that it was too expensive or too hard to get MPG up. But then when gas prices go up magically the car makers make it happen. THis is an area where car companies take a short term outlook and do whats cheap and easy but the govt. can force them to innovate.

        Whether or not people think the government should meddle with the market is another matter.
        Personally I think higher MPG standards are good for all of us. Consumers will save more money in the long run and we’ll cut foreign oil imports as well.

        • says

          Well, the point of the article is that nothing comes for free. Like my little anecdote about “Cheap, Fast, High Quality. Pick any two.”. You could say the same thing about any choice which we weight higher than others. (“Want it quickly and at a high quality? You’re going to pay!” for example.) You’re of course free to decide that the trade offs are worth it – and, of course, gas taxes would also leave a safety gap because it artificially elevates miles per gallon versus other inputs.

          I just can’t see the ‘mandate’ argument working, in practice (it makes great politics, of course). Pretend my company decides tomorrow, “all programmers must write 100 lines of code a day”. Maybe in a fit of productivity we hit those lofty targets for a few days – especially the people working on new code. But those people maintaining code or writing less complex code? They’ll either be arguing for lower standards (“let us count comments as lines of code!”) or unwinding all their loops, or using some other trickery.

          • freeby50 says

            Car makers can increase MPG. So yes the government can effectively mandate an increase and have it happened (as they’ve been doing for decades previous). Sure there is a cost and tradeoff for higher MPG. But I think its silly to assume the cost will be dead people. The costs will likely be higher purchase prices costs or lower power engines. I don’t imagine auto makers will automatically turn to lowering safety and lower weight isn’t the only solution. So to assume that higher MPG = lower weight = dead people is a little drastic and unsubstantiated.

          • says

            Whoa – back up a second here!

            You’re free to disagree with me all day, but if you have evidence that this is “drastic and unsubstantiated”, the burden is on you to produce evidence to support your position.

            Since we agree that, with the exception of some stationary collisions it’s safer to be in a large vehicle, you must be saying that MPG increases won’t lower weight more than weight would already be lowered. I disagree here, and I’m not speaking from some fringe place. Straight from the IIHS’s position page (slide 10 of the presentation I sent you):

            • Vehicles will get lighter and smaller as fuel prices increase

            One benefit of size-indexed CAFE is to keep larger, safer cars affordable longer for all income brackets


            • However, some people will die in motor vehicle crashes that would have survived without the downsizing

            – Large cars still will have fatality rates half as large as smaller cars and the smaller (on average) fleet that results will not be as safe as it could be


            In the article above I linked a WSJ piece on the F-150 switching to an aluminum frame and targeting a weight reduction.


            The Autoblog link above discussed car manufacturers setting up lightweight structures divisions (Lotus, BMW, Jaguar are listed, but it is implied they aren’t alone).


            I linked to the study by Mark R. Jacobsen of University of California, San Diego where he states in his abstract that “Each MPG increment to the standard results in an additional 149 fatalities per year in expectation.”

            I can’t follow what else you could mean by that comment, “the government can effectively mandate an increase and have it happen”. Remember that CAFE standards were effectively locked in place and little changed since Reagan’s term (and high MPG vehicles have still emerged, possibly due to the price of gas alone). Do you mean you think the car makers have the technology to produce cars with the same absolute safety levels as today with twice the MPG, yet aren’t releasing said technology to the public? I’d prefer you explain, because the only place I’ve ever heard that position is in the conspiracy theories surrounding the EV1.

            Am I missing something? Why do you think what I’m saying is unsubstantiated? Look – I hope you’re right and it’s as simple as de-tuning the engine and running some new software (on par with undervolting a processor, dimming a monitor, or some other ‘free’ updaete), but I don’t think it is.

        • sullivanb3 says

          I agree that the MPG standards are good for all of us. I believe it is a little stupid for the government to get involved though. The market definitely speaks for itself when you see ten prius’s in a 5 mile drive.
          Toyota definitely hit the nail on the head.

          Conspiracy warning: when the government bailout happened, rather than putting so much risk in gas efficiency, they may have tried to level the playing field.

  2. AverageJoe says

    Here’s my question: do those death toll numbers lower after a few years or stay the same? I’d be interested to see if it’s a new car/new tech thing or permanent.

    • says

      Here’s a graph since 1921. For the record, the first CAFE standards were in 1973 (oil crisis). I would also say that’s the magnitude of the mandate makes a difference – if I saw “hey, turn 23 into 23.5” that’s different than “turn 20 into 55”, for example.

  3. says

    I don’t see why they don’t just mandate 200,000 miles per gallon so that the entire life of my car requires a mere gallon of gas?

    Free lunch? Think of the insurance rate increases! Car insurance, life insurance and health insurance.

    • says

      Or my managers could mandate that I write code with no bugs. (Or I could write articles with no errors!) Hey, I’d do that if I could!

  4. Joe says

    I like Pigovian taxes and subsidies. Every highway should be a toll road. Gasoline taxes should be higher. Those things subsidize or tax the desirable outcomes, regardless of the innovative input by which the goal is pursued. Perhaps sugar cane ethanol (which is economical, whereas corn ethanol is straight-up stupid) would become more widely available.

    Notably, aluminium, fibre glass, and plastic, are increasingly common in cars not just because of fuel efficiency requirements but because they’re much cheaper to cast and manufacture with.

    • says

      You hit me right in the agreement part of my brain. Bravo.

      I’m not ready to say “raise gas taxes now!” in the US, at least until we, you know, make sure those funds don’t sneak over to general revenues. We also have to be sure that we want to change consumer choice – because here is a fundamental rule that “when you tax something you get less of it” (insert income tax arguments nearby). But, yes, I think it’s better to achieve the goals with a gas tax than with a straight-up mandate. (I also think a negative income tax is preferable to a minimum wage.)

      Oh, I agree – and there is a reason aviation uses the 7000 alloys of aluminum instead of steel. I know that the shift to other materials, whether for performance, efficiency, or whatever will continue to happen – I just hate mandates to use a certain thing. Thanks for your comment!

    • 101 Centavos says

      I’m partially with Joe on Pigou-style taxation. Wish/mandate in one hand and do something else in the other, and then sit back is not as efficient. Too many unintended consequences.

      • says

        My only question is “how do you pronounce Pigou and Pigovian?”. I took a shot on one of my podcasts, sounding like “Pig-oh-vee-an”. Cameron told me like 3 pronunciations are appropriate.

  5. sullivanb3 says

    would be worth noting that while using lighter/weaker metals, one would sacrifice a little bit of frame integrity, but with increase maneuverability, accidents may be more easily avoided.

    Software improvements are one of the best additions to any electromechanical system. Information is free!

  6. says

    Ask all of the workers from the Chevy Volt plant who recently got laid off how they’re enjoying their free lunch. They get to collect government unemployment checks now because the free market didn’t want to buy enough of their electric cars. Put me on the side of opposing all government mandates. The free market works every time. If gas prices went up to $15/gallon and electricity stayed reasonably cheap, Chevy wouldn’t be able to build enough Volts to keep up with demand.

    • says

      Politically, if the Government forces Chevy to pump out Volts they should also buy Volt fleets. I’m not saying it’s good economics, but it just seems like GM doesn’t have much of a choice.

  7. Potato says

    “you can’t legislate innovation”

    To be fair, the US CAFE system doesn’t require innovation. The targets can be hit by selling the exact same cars as today, but just shifting the mix (more compacts, fewer SUVs). They can be hit by innovating (more hybrids, electrics, and diesels). Or they can be ignored completely, which comes with a monetary cost. BMW, for example, doesn’t even attempt to hit the CAFE standards: they just pay the fines every year and jack up the price of their cars accordingly.

    • says

      You’re right – the loopholes are numerous; I read a few articles breaking down how car makers could hit high 50s with a fleet average in the 40s or even 30s. We could also have follow on effects (unintended, I assume), similar to when previous rounds of standards revealed loopholes in light trucks and pretty much made a market in SUVs. (I enjoyed that read)

      As for your point – they may not be de jure standards, but they certainly are de facto. BMW might be able to get away with thumbing their noses at the standards and paying the fine, but I can’t see how, politically, one of the Big 3 (2?) could do it without inviting a massive political storm.

      That said, if we really want to change consumer behavior, we should have a larger gas tax. It’s been 18.4 cents a gallon for so long that many states have overtaken it, haha. That way people can see what gas costs… and adjust their preferences accordingly. Will we become Europe, with tiny turbocharged Diesel cars? I don’t know – but if fuel consumption is something we actually care about, it’s better than CAFE (also, Jevon’s paradox!).