Hail Aggressive Drivers

On the surface, driving seems like a competitive activity. Yes drivers cooperate to avoid accidents, but aside from that other cars mainly seem to be in your way – if they take a road space you wanted, it will take you longer to get where you wanted.

But appearances can be deceiving. I’m a relatively aggressive driver, i.e., eager to get places fast, and overall it seems to me that other aggressive drivers are more my allies on the road than my rivals. My main opponents are slow pokes – being stuck behind them slows me way down.

So when I choose lanes I’m mainly looking to avoid lanes with slow drivers. I avoid trucks and most anything weird – those have more chance to be extra slow. Yes I might feel a bit rivalrous seeing an aggressive driver jump before me to grab a choice spot. But mostly I’m relieved to find a good person to get behind – they are unlikely to slow me down, and they tend to choose faster lanes.

This seems a decent metaphor for the rest of life. Yes when you associate with competitive aggressive folks you may have to keep on your toes more, and expect them to sometimes grab stuff you want. But overall associating with them will help you to move fast – they will tend to go places, and take you with them.

Added 9a: Folks, I’m not talking about going much faster than traffic, I’m talking about avoiding cars going much slower:

The risk of having a crash is increased both for vehicles traveling slower than the average speed, and for those traveling above the average speed. (more)

There are now 1.13 fatalities per 100 million miles driven.  At 30 miles per hour, this means a fatality every 337 years of constant driving. So if driving 1% faster gave you a 1% greater risk of death, for most folks that would be a good deal time trade.  Anyone know what the actual speed-death elasticity is?

GD Star Rating
Tagged as: ,
Trackback URL:
  • Wophugus
    • Armok

      This is a great applied rationality lesson.

      • GT


        Interesting enough i was helping my 6yo with her homework which included a story which was exactly this lesson. (although not using cars of course). She’s in a Chinese school, though.

    • lastAve

      lane ending/late merge: In this case, merging at the end’s actually better for traffic. An exit only lane is very different though.

    • I don’t recommend “cheating” by cutting into lines.

  • richard silliker

    A rabbit is always a good thing when the possiblity of a radar trap is ahead.

  • nazgulnarsil

    if your time constraints are such that you need to drive fast to get somewhere you suck at planning.

  • Very scientific.

  • Hail whatever appeals to Hanson. Yeah, I drive fast and am impatient and competitive too, but I wont pretend everything I fart is some great virtue. Tell us a characteristic about yourself you dont think is so great.

  • Matthew Fuller

    You are signed up for cryonics. I would think you would drive more cautiously as a result.

    • Dorikka

      I am also curious about this.

    • Aggressive =/= less safe, if you compensate with increased attention.

  • Ibod Catooga

    Fun fact: I once regurgitated an entire negro.

  • Riley Jones

    Sadly, Robin is not alone in this opinion. Ryan Holiday seems to agree.

  • Matt

    Thank you for cooperating with me and distracting the police officers. I have the satisfaction of knowing I can drive just a little slower than the aggressive drivers.

  • Hail drunk drivers! How about that! Fuck you cowardly sober slow folks!

  • Russell Wallace

    Aggressive driving looks like a great strategy until you stop being lucky, and then maybe you get to spend the rest of your life knowing people you care about are dead because you tried to shave five minutes off your transit time. Not hypothetical, I know people it’s happened to.

    • Evan

      i speed all the time and the only accidents ive been in are when idiots on cell phones missed a stop sign or merged w/o looking
      i dont have exact figures, but talking on a cell phone is something like 10 times more likely to results in car accidents than speeding. its actually worse than drunk driving. yet i bet you arent nearly as averse to that as speeding.

  • Seem

    Driving is a relatively anonymous activity so the social risks of aggressive behaviour are lower. Try your aggressive strategy in a social situation involving repeat interaction (i.e. most areas of life where people build true – rather than imagined – status).

  • Aron

    I think this is what they invented Twitter for:

    “I drive fast. I live fast. It’s metaphorical bitches. You do the math.’

  • Rafal Smigrodzki

    Hail Robin! I share your opinion about driving but … could you comment on the trade-offs inherent in speeding, especially given the following: http://www.freakonomics.com/2011/03/28/does-the-highway-patrol-keep-us-safe/

    Personally I would speed even if the calculations above were correct.


  • feh

    But Robin doesnt have data on the risks of agg lane switching which may be more dangerous than faster or slower speeds. Not to mention increased road rage from people like me who are liable to bump him.

  • Granite26

    I’m right behind you, Prof.

  • Rafal Smigrodzki

    The speed-death elasticity is probably very difficult to calculate, given all kinds of selection biases present in data. Nevertheless, the fatality rate on the German autobahn is essentially the same as on US highways, despite dramatically higher maximum speeds and AFAIK somewhat higher average speeds. I also read about a NHTSA study (but cannot find the quote) that claimed that removing highway speed limits in the US would increase fatalities by 200 per year (which would be orders of magnitude less than deduced from the Oregon experiment). I calculated that obeying highway speed limits cost me about 6 months of my life – this is a much higher cost than even the most alarmist estimates of increased death risk from speeding, so for me the calculus is quite clear: Speed limits (assuming zero utility to extra time spent driving unnaturally slowly) result in humongous net loss of QALY, on the order of thousands of years lost for each year gained.


  • CubistHamster

    I like aggressive drivers because they’re predictable. If your situational awareness doesn’t completely suck, it’s pretty easy to figure out what the guy going faster than the traffic flow and switching lanes every 5 second is going to do. Not so with everyone else.

    In fact, when someone tries to do me a favor (like slowing down to allow me to merge) it often seems to disrupt smooth traffic flow in both lanes that are involved, due to sudden and unexpected speed changes that are required.

    Frankly, I wish all drivers were aggressive, all the time. Altruistic drivers are the dangerous ones.

  • Dan Weber

    I call “aggressive” drivers those who try to squeeze maximum gain out of a zero-sum situation. On one path home, there are people who get into an exit lane and use it to pass slower traffic (on the right). Then they try to re-merge into the slower traffic. I took pride in not letting the jerks back in, but it probably wasn’t good for my blood pressure.

    Note that they didn’t improve traffic flow at all this way. The road ahead could only handle so many people — they just wanted to swap places with people ahead of them by passing on the right.

  • Pingback: Recomendaciones « intelib()

  • Douglas Knight

    Trucks are driven by professionals who are the opposite of aggressive drivers. Most cars want to drive faster than trucks, so should usually pass them. But in congestion, get in the right lane with the trucks. The aggressive drivers will take their negative sum games elsewhere and you’ll zip along.

    • Trucks are driven by professionals who are the opposite of aggressive drivers.

      Right — they camp the passing lane rather than help traffic flow smoothly.

      At least where i live.

  • Doug

    A related thing happens in quant/high-frequency trading. The presence of other high frequency traders using similar signals to yours can take the choice trading opportunities before you, reducing your profit.

    However the market ultimately moves in the direction of whichever way the participants push it. Thus the presence of more traders doing similar things to you can make your signals stronger and your strategy more profitable, especially after adjusted for risk and/or transaction costs.

  • Dave

    If everyone drove twice as fast,the time each trip took would be cut in half. This would reduce by half the number of cars on the road,making driving safer. Cutting the speed traveled by 50% would the double the length of each trip with a corresponding increase in traffic density,decreasing safety. Moral: Step on it,but don’t try explaining this to a cop.

  • pconroy

    I’m also an aggressive driver – on the rare occasions I drive – and in my 20’s was an aggressive and somewhat reckless driver, though I’ve never had an accident, not as much as a scrape.

    I totally agree that in fairly free-flowing traffic, best to get behind another aggressive driver – as they are more likely to find the quickest path through; however in stop-and-so traffic, better to get behind a large truck – as they can see above the fray, and spot what is causing the traffic to be so slow, like a closed lane, or traffic accident

  • Luke

    Safest thing is to not drive. Failing that, develop a good sense of situational awareness. If you are driving relatively fast you do need to be relatively more alert. This has trade-offs in terms of stress that need to be taken into account. I suspect better sleep (melatonin supplementation perhaps) could reduce deaths a lot more than slowing down.

    • mobile

      If the alternative is walking or biking, then the safest thing is to drive. If the alternative is staying home instead of going to a job that would give you a good income, then you are less prepared for health problems and the other vicissitudes of life, and once again you are probably safer driving.

      • This is somewhat context dependent. Are we assuming the same distance will be covered either way? Are we assuming that while you walk or bike, everyone else will drive? Are we measuring your personal safety or the average person’s safety as affected by your decision?

        Many cities in europe are largely car-free. I imagine living in one of these and walking would be safer than living in a city with cars and walking, perhaps even driving. Using public transportation would be safer than driving.

        Personally I have managed to move to a place where I can walk just a few blocks to work every day, which causes me to exercise regularly and is less stressful for a variety of reasons.

  • JB

    The problem with only looking at fatality statistics is that you don’t measure people that get into accidents where they don’t die, but have something else bad happen to them. What about people that become paralyzed. Or people that have their face disfigured. Or what about simply people that get whiplash and now have a bad back for the rest of their life. Chronic pain is one of the worst things that can happen to you. And how about totaling your car, but no injury. And now you have to spend money on fixing it and you raise your insurance rates. I think you have to add all of these things in the numerator when you are trying to determine a ratio of “bad” things happening.

  • mobile

    This report estimates that about 30-32% of fatalities are speed related, so the speed-related fatality rate may be something more like 0.35 per 100 million miles.

    • mobile

      An even better factoid comes from this report:

      . In 2000, the cost of
      speeding-related crashes was estimated to be $40.4 billion — $76,865 per minute or
      $1,281 per second.

      Now all we need is a good estimate of the benefits of speeding in 2000.

      (well, plus a few other cost estimates: increased fuel consumption, for one)

    • Rafal Smigrodzki

      If you assign zero utility to the extra time spent driving unnaturally slowly due to a speed limit, and you assume the the average speeding driver speeds by about 10 miles, all you need to know is the number of miles driven (available from the NHTSA), and the fraction of mileage spent speeding (also should be available from the NHTSA), and then the average utility of time not spent driving … and presto, you can easily calculate it. Are you game for the exercise?

  • Telnar

    One confounding factor when trying to figure out the effeciency of higher speed driving is that it requires more attention for a given level of risk. Some of the slower drivers might prefer to be able to devote more of their attention to conversation or a podcast, and therefore prefer to drive in a way that requires fewer decisions.

  • J

    “the fatality rate on the German autobahn is essentially the same as on US highways”

    Germans also have extremely rigid enforcement of lane discipline laws. I’d love to see that here, but I’m not holding my breath. Unfortunately, without it, autobahn-like speeds are pretty dangerous.

    “In 2000, the cost of speeding-related crashes was estimated…”

    Take a look at the first sentence of the report you linked, My experience has been that police officers have an unreasonably broad idea of what constitutes “too fast for conditions”. NHTSA reports citing that issue are propaganda used to justify speed traps and unreasonably low speed limits, not valid data. I suspect if you could break the data down you’d find that nearly all crashes that don’t involve intersecting traffic are attributed to excessive speed, at least in part.

    • Rafal Smigrodzki

      You are absolutely right, lane discipline is very important, and Germans are much better drivers than Americans – but this is in part because of the speed limits here which teach bad habits, rude behavior (slow driving in the fast lane) and encourage inattention.

      The poor quality of American drivers I have observed since moving here is thus the effect of speed limits, and at the same time it is used as justification of speed limits. Catch 22.

      • J

        I suspect technology is just as much to blame. When police in the US were given RADAR, it gave them the ability to detect a specific, quantifiable violation that was nearly impossible to contest in court and highly resistant to spurious civil rights complaints, and they pretty much stopped enforcing other traffic laws. Drivers in the US haven’t always been as rude or inattentive as they are today. I’d say the widespread adoption of dashcams might turn the tide, but low speed limits generate so much revenue, we’ll need to outlaw government entities keeping fine income before that can happen.

  • Wonks Anonymous
  • Aggressive driving works well inside cities (off the highway.) A study I saw about 5yrs ago found that, worldwide, cities with more aggressive drivers had much higher avg traffic flow.

    Only trouble is, aggressive driving only works on highways outside the city grid when the road is relatively empty. When it’s congested, traffic jams are almost entirely due to the common aggressive behaviors: tailgating and lane switching. The jam shown in http://www.trafficwaves.org video is caused by a fight between mergers versus blockers (or “cheaters” versus “vigilantes.”) If those drivers opened up and tried to attain “zipper” merging, those lanes would flow fast like gear teeth.

    If you want to drive fast in congested traffic, the best tactic is to avoid triggering traffic jams: temporarily suppress aggressive behavior. One website likened this to the “Gravel-pit Etiquette” used by professional drivers in large construction projects. In those situations the “cheaters” will end up getting fired. By trying to get a couple positions ahead, they completely foul up the overall flow, and the rapid moving “gear teeth” patterns all grind to a halt.

    On congested highways it’s *impossible* to drive faster than average. For example, during rush hour the cars are spaced 1-2 seconds apart, so in order to shave 5min off your commute, you’d have to pass 300-600 other drivers. WEIRD, EH? If you’re not passing hundreds of other cars, then you’re not driving faster. That’s where dishonest human nature cuts in: we abandon our goal of actually getting to work earlier. Instead we try to pass a handful of other drivers, and if successful, we congratulate ourselves as if we’re winners in a race. In a race, ten positions ahead is significant. You might come in first rather than being ten cars back. But on a commute, ten positions is completely insignificant: it’s shaving 10 or 20sec off your commute.

    The trick is to judge when aggressive driving works, and when it’s counterproductive. On key to this is to notice “race mentality” in yourself. On open highways you can speed, and easily pass hundreds of other drivers in an hour of driving. But as soon as you notice yourself working to get one car ahead, or notice that “my position in line” has become important, that’s when attempts to drive faster will only throw a wrench in the gears.

  • Pingback: Conductores agresivos y embotellamientos fantasma « Notas marginales()

  • Andrew

    According to the WHO 2004 estimates, the total amount of DALYs due to road accidents is 985 per hundred thousand. This compares to a total of 1,384 for cancers for example.
    I am reading this blog for the first time in years, and I have noticed that Robin has talked about health economics. But maybe this suggests we should be concentrating more on road safety than certain health priorities? The speed vs safety elasticity is certainly an interesting question, but you have to also consider effects of commuting times. Such as behavioural effects – choosing where to live and work.

    Some people may actually enjoy their commutes, their private time to listen to music, or chat socially with friends. Perhaps the time spent commuting time lets us process our thoughts for the day. Who knows, but there are lots of questions that can be asked.

    Admittedly I’m also an aggressive driver with a turbocharged car and I agree with Robin’s behavioural characterisation of aggressive drivers.