Freakonomics Hucksters: “Save the Earth, Drive Your Car”

Remember those wizards of counter-intuition, the Freakonomics guys? You know, the ones who told their audience that it’s safer to drive drunk than to walk drunk? Well, in his latest piece for NPR’s Marketplace, which ran with the headline “Save the Earth, Drive Your Car,” Stephen Dubner talks to Clemson University’s Eric Morris and arrives at the ridiculous conclusion that driving is greener than transit.

Any counter-intuitive finding, true or not, seems like it can pass muster with Steven Dubner of "Freakonomics" fame.

The intellectually dishonest argument rests on the per-passenger energy consumption of cars versus buses. Buses are potentially much more efficient than cars, Morris admits. But many buses are underutilized: The average bus carries just 10 passengers, while the average car carries 1.6. As a result, Morris says, those traveling by bus consume 20 percent more energy per passenger than people driving in cars. (American trains, he concedes, are two-thirds more efficient than cars on this measure, but he qualifies that by saying the “number is warped a bit by the New York City subway, which is just a monster of efficiency.”)

So let’s say you’re an average, environmentally-concerned Joe, and you take this segment to literally mean that you should, in fact, drive your car to “save the earth.” How would that affect the environment? Well, the decision to take transit would consume essentially no additional energy — you would be using the system that’s at your disposal. While driving a car would spew greenhouse gases into the air that would otherwise stay in your fuel tank. It is pretty clear which choice is better for the environment, and it’s the intuitive one.

Midway through the article, after slagging transit with their big, attention-grabbing counter-intuitive point, Dubner and Morris admit that getting more people to use existing transit is unequivocally good for the planet. What they actually want to warn people about is building new transit, which won’t work “in places like Cleveland, Pittsburgh and Memphis” because the routes will be under-used. This, too, is incredibly dishonest.

Places like Cleveland have weakened the transit systems they were endowed with by creating every possible incentive to drive. If anything, the “hidden side” of this issue that Morris and Dubner play up for its counter-intuitive shock value — energy consumed per passenger mile — just points to the disastrous environmental consequences of planning communities around driving. The low ridership on Cleveland’s passenger trains is testament to poor planning, not an indictment of transit. Check out the pedestrian environments around some of Cleveland’s rapid transit stations:

Clearly, many American cities need to repair decades’ worth of damage to the walkable urban fabric that makes transit efficient and well-used. They also need to build better transit so new walkable development can flourish. When Cleveland made a substantial investment in upgrading bus service on Euclid Avenue, ridership shot up 47 percent and a wave of new development followed.

This gets to the biggest omission of all in Morris and Dubner’s argument. They never mention that transit helps create places where people drive less, walk and bike more, and live in much more energy-efficient homes. Professor Robert Cervero at UC Berkeley estimates that transit alone has the potential to curb America’s per capita carbon emissions by 20 percent. When you factor in the energy efficiency that comes along with the type of development transit supports, the number jumps to 30 percent.

Dubner notes at the beginning that New Yorkers have the smallest per-capita carbon footprint in the United States, but he never fully explains why places with good transit like New York, Boston, and San Francisco also have some of the lowest energy consumption rates in the country. There’s no reason places like Cleveland, Pittsburgh, and Memphis can’t join them as green cities with great transit.

  • mikesonn

    “while the average car carries 1.6”

    Where are they getting that data? Come ride a bike with me around the “green” Bay Area and you’ll see SOV after SOV after SOV with just a few cars carrying more than one person sprinkled in to make us all feel warm and fuzzy.

  • They also don’t take into account that individual cars mostly just sit idle. The production costs and energy of those vehicles is very underutilized, while many transit systems are squeezing every mile they can out of available rolling stock.

  • Anonymous

    This just shows you can always cherry-pick or make up a metric to “prove” whatever you want.

    A better analysis of this issue would also consider all the cars that wouldn’t need to be manufactured, all the roads that wouldn’t need to be built, and all the parking spaces that wouldn’t need to exist if more people chose transit instead of private cars. And once you reduce the amount of space devoted to driving and parking, you can have denser and safer cities where you can even (gasp!) walk or ride a bike, bringing your carbon footprint to nearly zero compared to the other options.

  • Elliot

    I was with you up until the line “Professor Robert Cervero at UC Berkeley estimates that transit alone has the potential to curb America’s par capita carbon emissions by 20 percent.” This is misleading cherry-picking, which is hypocritical just after you dressed Dubner down for doing the same thing. Cervero’s number is based on an entirely different development pattern, not just implementing transit, a factor which you even discuss. 

  • Guest

    They forget one of the most important environmental advantages of public transit: storage. A bus needs to be stored at just one place: the depot.

    On the other hand, a private automobile owner needs to store his car at home (either in a garage, on a driveway, or in the street), at the office (garage, parking lot, or on the street) and at all places where his car takes him (school, the grocery store, the doctor’s office, a restaurant, the movies, etc.). All of those parking facilities take energy to construct and maintain.  Even lighting an outdoor lot at a shopping mall 24/7 or running elevators at an airport parking garage creates energy costs that more than level the playing field between public transit and private automobile use.

  • @mikesonn 

  • Matt

    Now, lets not get the pitchforks out just yet. This is a very specific argument. The idea is that cultural norms are difficult to change. If a proposed transit infrastructure isn’t going to provide the incentives for people to stop driving, which has proved true in many cities, then it’s conter-productive to build the infrastructure. It’s a straightforward argument, even to a transit enthusiast like myself. This argument can even be used to by transit proponents to highlight important aspects that transit systems must have. These incentives (often speed incentives) must be specifically aimed at altering cultural norms, like bus priority lanes, or underground trains vs street cars, etc. Proposing a bus system that will run at 3mph (such as some of the busses in NYC) is probably not an effective way to change the cultural norms within a region, and will probably end up being counter-productive as far as pollution is concerned.

  • @mikesonn:disqus   They likely got the average vehicle occupancy from the National Household Travel Survey.  I did a run for you:

  • @9713004d17188da28fc6763a3705921f:disqus completely agreed. You’re speaking to the relationship between ridership and ease of use, reliability, speed, and comfort. Some transit initiatives don’t address this; and others do. 

  • @9713004d17188da28fc6763a3705921f:disqus If Dubner really wanted to provoke our thoughts with a “nuanced” argument, he wouldn’t have consented to a headline that says “Save the Earth, Drive Your Car.”

  • JK

    Viewing transit in isolation from land-use is simply moronic. No doubt a subway system in rural Wyoming would be enormously energy inefficient compared to a pick-up. Transit systems decoupled from smart zoning are battling massive hidden subsidies and incentives for car-use. In particular, minimum parking requirements bundle the cost of parking into the cost of just about every transaction. Plus, parking, even parking decks, reduce density and make walking less convenient, which in-turn makes the walk-transit trip take that much longer compared driving.

    These schmuckster economists know that when you subsidize a substitute good, it will increase demand for the substitute and reduce it for the good — making driving cheaper, reduces demand for transit. Society has made the direct cost of parking really cheap. Since cars spend 90% of their time parked, often in cities, where land is expensive, this is a very big subsidy. Let’s get some equal airtime for Don Shoup to smack these chumps around.

  • Larry Littlefield

    If an economist makes that case, simply tell him (in his language) that he is confusing average and marginal cost.  And on a marginal cost basis, transit is almost always better.

    Meanwhile, that 1.6 for auto occupancy seems rather high.  I wonder if they are counting the driver who is merely dropping off and picking up the kids as a passenger?  They surely are not counting the bus driver.

  • Matt

    Ben, everyone knows that the modus operandi of freakonomics is, “well, if you look at it this way, [enter absurd claim here] is actually true.” The marketplace chat is initially attention grabbing, but the content is perfectly reasonable. A bus system in rural Wyoming probably wouldn’t reduce carbon emissions, changing cultural norms is difficult, and building a transit system isn’t de facto environmentally cleaner. It’s a mildly interesting, yet fairly trivial result, much like most of the freakonomics topics (the cobra effect, selling beer to cut done on public drunkenness, wanting your car to be stolen if you have insurance, etc.).

  • Michael Morris

    You make a good point that by one person using transit instead of driving the carbon footprint is reduced. But this segments point was more a criticism of unneeded transit, which is almost always inefficient as first. And what really fails is trying to force public transit on those not receptive to it. The marketplace piece was aimed more at planners and less on commuters, making sure urban planners know that buses running every 15 minutes with an average of 10 people are not a green solution, in the short term. We should be patient with transit systems, especially light rail and street cars because economies will slowly begin to build around areas with good transit as congestion and city populations increase.

  • Michael Morris

    You make a good point that by one person using transit instead of driving the carbon footprint is reduced. But this segments point was more a criticism of unneeded transit, which is almost always inefficient as first. And what really fails is trying to force public transit on those not receptive to it. The marketplace piece was aimed more at planners and less on commuters, making sure urban planners know that buses running every 15 minutes with an average of 10 people are not a green solution, in the short term. We should be patient with transit systems, especially light rail and street cars because economies will slowly begin to build around areas with good transit as congestion and city populations increase.

  • Michael Morris

    You make a good point that by one person using transit instead of driving the carbon footprint is reduced. But this segments point was more a criticism of unneeded transit, which is almost always inefficient as first. And what really fails is trying to force public transit on those not receptive to it. The marketplace piece was aimed more at planners and less on commuters, making sure urban planners know that buses running every 15 minutes with an average of 10 people are not a green solution, in the short term. We should be patient with transit systems, especially light rail and street cars because economies will slowly begin to build around areas with good transit as congestion and city populations increase.

  • Michael Morris

    You make a good point that by one person using transit instead of driving the carbon footprint is reduced. But this segments point was more a criticism of unneeded transit, which is almost always inefficient as first. And what really fails is trying to force public transit on those not receptive to it. The marketplace piece was aimed more at planners and less on commuters, making sure urban planners know that buses running every 15 minutes with an average of 10 people are not a green solution, in the short term. We should be patient with transit systems, especially light rail and street cars because economies will slowly begin to build around areas with good transit as congestion and city populations increase.

  • Bolwerk

    Dubner is actually way behind the times on this one. This debate has been coming up on Usenet transit boards for decades now. It’s usually instigated by randroids or hack academics. Some comments make Dubner look downright clever.

    Anyway some things to bear in mind about these comparisons:

    • Average passenger-mile energy consumption of transit buses is in the 4000-range as of my last reading a few years ago. It’s in the 3000 range for cars. (Long-distance buses, the kind that go city to city with a full load, achieve 800 BTU/psgr-mile.)

    • If you actually dig up year over year numbers, vehicle-mile energy consumption of transit buses used to be lower than they are today, but Uncle Sam’s regulatory regime apparently made buses heavier over the years.

    • Yes, any apples-to-apples comparison is full of it. People tend to drive several times further than they take a bus or train. If buses do have a per-passenger disadvantage, it’s eaten up by the POV’s higher-intensity use per-passenger-mile. I am pretty sure all driving is included in the car energy use number. Empty freeway driving is conflated with use in urban areas.

    • Also, the average for buses is obviously skewed by the bus systems that don’t see many passengers.  Some bus systems do very well, others do not.

    • Rail tends to be significantly more energy-efficient than buses or POVs on a per-passenger basis, of course.

    • Also, the nature of transit use: add a trip, and you more or less always reduce average use. The nature of cars: add a car and you increment use, or even make all other trips less efficient.

    • Someone else also mentioned land use, so I won’t belabor that.

    Well, see table 2.12 here for energy statistics.  Better to depend on than my memory. Regardless, Dubner should stick to baby names and sex.

  • Anonymous

    Americans gotta consume, writers gotta write.  Thanks for calling out this jerk.

  • Joe R.

    How about comparing the incremental energy use of both modes? Adding another passenger on a bus gives virtually no extra energy usage. If that same person decides to drive instead they use a lot of energy.  While we’re at it, add in the energy cost of otherwise unnecessary, often underutilized, infrastructure to accommodate passenger cars. The hard fact is auto use uses far more energy per passenger mile than anything else once you count everything. The energy use of the vehicle itself is just the tip of the iceberg.

  • IsaacB

    I could have seen Dubner’s argument coming. As Bolwerk said, stuff like this was discussed on Usenet in the latter 20th century. As you wisely picked up, and which needs to be shouted from the rooftops, is that transit enables you to create places (like Manhattan) where the main mode of transportation is walking. 

  • JamesL

    Assuming that transit can only carry 10 people per bus because it currently, on average, carries 10 people per bus presumes that ridership is the goal of the services that exist today. In most places, this is not the case. Most agencies have ridership among their goals, but often subordinated to others like service area coverage or serving high-need populations. I believe Jarrett Walker once said that criticizing one of these predictably low-ridership routes for low ridership is like throwing out your microwave because it doesn’t make ice; that’s simply not what it’s designed to do.

  • Anonymous

    What about the emissions related to constructing and maintaining all that car storage space, you know, parking lots?

  • Guest

    I am still waiting for somebody to provide a good rundown on “passenger miles.”  
    For any given origin and destination, autos will have the same or fewer passengers miles than transit, since you often need to take one route and transfer to a second route slightly less directly on transit than you would go if you just drove straight there.  

    On top of that, there is a question about how mode might affect individuals’ choices in regards to destination and trip chaining.  If you ride transit, are you more or less likely to go do something at a location further from home?  Are you more likely to combine trips if you’re driving or taking transit?

    Without these answers, you really can’t interpret any metrics.  You can tell how efficiently the vehicles moved people a given distance, but you can’t tell how efficiently it completed their trips.

  • The last comment is instructive. If I am a transit user and I live in Noe Valley, I will take the 24 Divisadero 14 blocks to go to Clif’s Variety to buy a hammer and a pack of lightbulbs. If I am in my car, I will drive to Home Depot in Colma – several miles away. And since I have a car I will probably buy a lot of extra crap that I don’t need. Great for economic stimulus but very bad for the environment.

  • Miles Bader

    @abb249055208c7af4d35568e422dfd63:disqus Of course given completely geographically arbitrary endpoints, transit will probably require more travel.  But real endpoints aren’t arbitrary at all, they’re distributed in a way that’s highly influenced by the way people travel.  In a city dominated by mass transit, destinations naturally end up grouped around transit nodes and lines.

    There’s a feedback loop … the more people use mass transit, the more attractive it is for businesses and housing to locate in places convenient to that transit, which in turns makes using that transit more efficient and easier, and thus more desirable.

    Cities essentially exist in local minima, where their layout and transit are more or less adapted to each other.  Other minima may be much more minimal, but moving out of the current local minimum into a lower (better) one requires things to get worse before they get better.  In the long run doing this is a win, but in the short run this is hard, and so tends to be resisted, people being people, and politics being politics…

    [Of course even in a car-dominated city there are points of density that are somewhat suited to transit, and can be exploited for less painful short-term gain, but at some point the structure of the city itself has to change to really get the best result.]

  • Keith Laughlin

    This ridiculous argument can be refuted with a simple thought experiment. Imagine that a metro region with a decent public transportation system aggressively markets a Drive Your Car to Work Day and all transit riders shift modes to single occupancy vehicles. Congestion would increase, productivity would decline, air emissions would rise, parking would be scarcer, etc. How could anyone claim that such an outcome would be greener?

  • mediawatcher

    One of the problems lies with our historical land-use patterns — specifically, our proclivity for putting office campuses (campi?) in greenfields rather than where they can reasonably be served by transit. Make employment transit-accessible and transit ridership will likely go up. In the shameless-organizational-promotion department: New Jersey Future (disclosure: my employer) has two pertinent reports on this: Targeting Transit — — which examines development opportunities around train stations, and Getting to Work — — which looks at the problem of connecting jobs to transit.

  • Anonymous

    @02f865733dfa420bd887a80616cfcad5:disqus : your thought experiment is exactly what happened in Manhattan after hurricane Sandy. It became obvious that the streets just don’t have enough capacity for everyone to drive.

  • Charles_Siegel

    The book Freaknomics was written because the economist Stephen Levitt had lots of unorthodox ideas, and Stephen Dubner got the idea of packaging them as a book.  It made lots of interesting points, though it was more like a series of essays than a unified book.

    After its great success, they followed up on its formula by coming up with more wild, unorthodox ideas in SuperFreakonomics.  But unfortunately, they had exhausted all of Levitt’s interesting ideas, and so they just turned into contrarians, looking hard for orthodoxies they could deny in order to attract attention to themselves.

    This claim about transit is one more case of their looking for clever, unorthodox ideas, after they have run out of interesting ideas.

  • Indeed, I’ve seen this argument before and it comes down to looking at a tiny slice of the entire life-cycle costs. Thanks for addressing the marginal environmental cost in your write-up; there are at least two other huge fallacies:
    1. As Guest mentioned below, per-trip measures and per-passenger mile measures differ substantially because auto trips are much longer than transit trips. On a per-mile basis, even walking is very energy intensive.
    2. The environmental impact of building the car vs. building the bus is never addressed.

  • IsaacB

    Oh, and anecdotally, it does seem safer for a drunk individual to drive than to walk (and to hell with the people s/he hits). That does not make it good for society at large. 

  • Anonymous

    The whole argument is totally, ridiculously stupid.

    Let’s assume that everything Freakonomics said is right. It doesn’t mean jack for an individual’s decision to drive or take transit, because it’s not like the transit agency decides to run 1/10 of a bus for every new passenger they gain. If drivers shift to transit, many of them will be accommodated by buses/trains that are already running so the incremental pollution is not 1/10 of a bus with 10 passengers on it, but the incremental pollution caused by that passenger’s weight affecting vehicle performance.

  • If only economics had some form of analysis suited to this kind of problem, a way of comparing what happens if one person (someone “at the margins”, you could say) decides to drive or take transit. What are the consequences of her decision for the transportation system as a whole? Are pollution and traffic slightly better or slightly worse? What factors influence this fence-sitter’s decision to drive alone or take transit, and how can we adjust them to produce better transportation for the greatest number of people?
    Yes, if only there were something called “marginal analysis”, the esteemed Freakonomists would have some way of fairly treating their topic instead of emptying yet another silo of manure onto the heads of their stunningly naive readership.

  • Joe R.

    @twitter-139098684:disqus It’s good you mentioned the environmental cost of building cars. I’ve thought for a long time that personal autos represent a huge inefficiency in that they’re a capital asset which is not utilized for 23+ hours of the day. For that reason alone building public transit vehicles which are used 12 to 20 hours per day for up to 50 years represents a much better allocation of resources. And then you have the missed opportunity cost of land used for auto travel which could have been put to some better use. To top it all off, despite all this inefficiency, cars often don’t even offer the best travel times because they’re fighting each other with their sheer numbers.

  • A useful model for transportation planning is wholistic, that is, of the four modes of urban/suburban travel – cars/trucks, mass transit, walking, bicycling – all must function adequately or none will function optimally. When roadway infrastructure is designed solely to serve cars/trucks, this is an (constitutionally inequitable) impediment to the other travel modes, and, traffic volume becomes an impediment to the optimal function of the cars/trucks mode.

    Those who advocate for transit, particularly for light rail, should apply “mixed-use” urban planning scenarios to surburban as well as central city development. Today’s sprawling metropolitan areas require transit systems such as light rail that can ‘comfortably’ cover long distances. The worst traffic congestion is generated in the suburbs where the need for redevelopment is much greater than in central cities. As suburban station areas develop healthier mixed-use economies, their residents need for cross-county travel is reduced, the overloading of transit during rush hours is also reduced but complimented with an increase of transit use during off-rush hours.

  • I don’t get it. So buses can be more efficient, but they’re underutilized… so the answer is to drive more, therefore making the buses even less utilized?
    Seems that the Freakonomics writers are more interested in getting page clicks than providing wisdom.

  • Ants are smart, Ants are stupid.  Have you ever look at an ant, at the individual level, ants are stupid, one is pulling on a twig one way, the other ant pulling on the same twig the other way.  But look at the big picture here, If you look at the ant Colony as a whole, you see how smart the ants are.

    I know that in the coming local transit planning meetings, this will no doubt be used by those who see transit as an evil USSR/UN conspiracy to become collective.  What I have taken from this story is that a bus not being fully utilized does waste energy.  But a system of Buses/trains being used in tandem reduces the need for owning a personal vehicle which in whole reduces energy usage.  

  • Transit agencies don’t run buses to save the planet – they run them to move people from point a to point b – bit of a difference there.

  • I think Ms. Schmidt misunderstood the gist of Dubner’s argument, which is pretty simple:  you have to run the numbers to prove that building a mass transit system is indeed more efficient in a given situation over cars; you can’t make the assumption a priori.  I doubt if Dubner was really arguing against building new transit systems altogether, but rather against the common tendency where wishful thinking and a prejudice in favor of building new mass transit projects where the numbers don’t pan out.  San Jose is considering a light rail extension to Los Gatos that would cost $175 million that will probably support 200 riders per day, according the transit agency’s own projections.  How does this help the environment?  (On the other end of the spectrum, there’s some BART infill station projects that would cost less, and benefit thousands.)  Green transit advocates should be on the front lines of those demanding the most environmental benefit from limited financial resources.  While there is a place for being visionary (where the numbers don’t support a new system yet, but it could 10, 20 years down the road), it hurts the credibility of transit advocates when they turn a blind eye to boondoggle projects.

  • Dave Fisher

    This is a silly article that has heavy bias all over it. If you READ the Dubner/Morris article they fully disclose what they are and aren’t covering. As Angie was willing to admit it doesn’t even say that you shouldn’t take public transit – rather encourages it.

    What Freakonomics did here is say that advocates of public transit shouldn’t assume it is more environmentally efficient that current alternatives. The data say this is true. Angie may be able to provide evidence of ridership shooting up after investment, but Morris/Dubner aren’t saying that isolated incidents don’t work. They just say that with currently established programs they aren’t as efficient.

    They also state that in some of the really high density areas mass transit is very efficient, like the NY subway is an example. Many mass transit advocates will compare us to Europe and say we are behind when Europe is a different animal altogether with its much higher population density. Mass transit is evidently not as efficient in a place where people have room to build houses where they want.

    We can’t let our biases interfere with honest data gathering. Dubner might be good at selecting topics that will grab our attention, but that doesn’t invalidate his research. 

  • Yes and no. It’s standard Dubner. You have to truly look at the holstic sitatuation. if the bus can’t fill to a reasonable level (say a passenger load average of 8 passengers at all time) it won’t be more environmentally friendly. If it is on a corridor that can’t sustain that, it’s likely better to bike (or in odd cases drive). However it also depends on the car.

    But I digress, the case with the Freakonomics crowd is that one is encouraged to do the research and figure it out. What we’re told by the media and public leaders is rarely all there is to the story. Matter of fact, I rarely see anything from public leaders or media that is EVER the entire story. Read, learn and research, then do it again. I believe that’s the most resounding takeaway when reading anything form the Feakonomics crew.


Department of Energy Gets Basic Math Wrong in its Rail Analysis

When it comes to the carbon consumption of cars, trains, and buses, the U.S. Department of Energy’s (DoE) Transportation Energy Data Book [PDF] is an indispensable resource. But this year’s Data Book contains an eyebrow-raising error in its analysis of rail’s energy use. (Image: DoE) Page 66 of the Data Book, reprinted on the DoE’s […]

The $125,000 Commute, Not as Uncommon as You’d Think

We know car commuting is expensive, more expensive than is commonly recognized. That’s because every time we get in the car, we’re not presented with a bill for gas, maintenance, insurance, lost productivity, etc. Let’s say the tabulation was all there: the cost of your car commute. Maybe even included with mortgage paperwork, so everyone […]

Mr. Money Mustache on Retiring at 30 By Riding a Bike

His claim to fame is that he retired at age 30. He swears that you can achieve greater financial freedom too, if you follow his example by eliminating unnecessary expenses and investing wisely. He calls himself Mr. Money Mustache. And he says nothing is more essential to his philosophy and wealth-building strategy than riding a bike. Mr. […]

Here’s How Much Safer Transit Is Compared to Driving

Keep this in mind the next time a high-profile train crash generates more press coverage than a year’s worth of car wrecks: Despite the media sensationalism and overwrought regulatory responses that follow such events, transit is already a lot safer than driving. Looking at traffic fatalities per mile traveled in the U.S., analyst Todd Litman […]

How the Self-Driving Car Could Spell the End of Parking Craters

Here’s the rosy scenario of a future where cars drive themselves: Instead of owning cars, people will summon autonomous vehicles, hop in, and head to their destination. With fewer cars to be stored, parking lots and garages will give way to development, eventually bringing down the cost of housing in tight markets through increased supply. […]