The Need for Speed and the Value of Human Life

Tod Kelly

Tod is a writer from the Pacific Northwest. He is also serves as Executive Producer and host of both the 7 Deadly Sins Show at Portland's historic Mission Theatre and 7DS: Pants On Fire! at the White Eagle Hotel & Saloon. He is  a regular inactive for Marie Claire International and the Daily Beast, and is currently writing a book on the sudden rise of exorcisms in the United States. Follow him on Twitter.

Related Post Roulette

73 Responses

  1. Kyle Cupp says:

    I liked this post a lot, Tod. Good questions worthy of reflection and introspection.

    To my mind, being pro-life, if it means anything, has to mean more than a political position on abortion and other “life issues.” It should be a way of being in the world, a habitual disposition, in a word, a virtue. A person who drives recklessly to some pro-life event isn’t truly pro-life, whatever his or her positions on the issues.Report

    • Tom Van Dyke in reply to Kyle Cupp says:

      Very nice, Mr. Kelly. I especially enjoyed the “totemistic” assay of those who “act for us as a symbol of life.” Nothing wrong with that, or who we might choose as our totems. [Which surely says more about ourselves than them.]

      But let us not cheapify “pro-life” with sophistic intent. It means “anti-abortion,” as in slaughtering innocents for the sake of the convenience. There are many other worthy causes that respect the gift of life as divine or otherwise inherently worthy: innocent convicts, convicted slam-dunk murderers, cetaceans, great apes, dogs [see Michael Vick], even those vermin-of-the-sea, otters. But it’s an injustice to the truth and accuracy to sledge them all together.

      As for your automobile device arguments, empirically there can be no argument, only the question of liberty and/or self-deception. “I” won’t kill anybody if I’ve had more than the parsimonious .08 blood alcohol level, or speed when it’s unsafe.

      Perhaps .12 and/or 90 mph as the cutoffs? Not to be combined? [Trying to work with you here on the rubber-meets-the-road level.] The problem is that the fussers are seldom content with setting outside limits such as these. MADD, having achieved righteous success against drunk driving, finally spun itself into near-obviation by agitating for the .08 that has made a criminal of anyone who takes a third glass of wine. Now the law is the enemy of a healthy chunk of the people instead of its friend.

      [Looking at your fussified bio, I realize you & I get along pretty damn well considering my dream date with you is smoking bigass stogies, playing the dozens, and taking in Blake Griffin leading my beloved Clippers to victory. And knocking back a few in the process. I cannot embrace yr Vision of the Anointed.]

      Because all I could think about whilst reading yr post was the late great Michael O’Donoghue, first editor of the National Lampoon and filmist of Mondo Video, doing his dark, grim and brilliant Mr. Mike on Saturday Night Live.

      Remember all those totally awesome hood ornaments American cars used to have? Then “they” scotched them, for mauling the occasional idiot pedestrian who didn’t know his place in the American car culture. Mr. Mike: “Personally, I think a really good hood ornament is worth four or five human lives.”

      That’s what freedom and America are really all about, Tod. Hood ornaments and smoking whatever you want and insulting each other and cheering for really pathetic basketball teams anyway. Plus consuming alcohol and operating motor vehicles.

      Within reason, of course.Report

      • Tod Kelly in reply to Tom Van Dyke says:

        I reread this whole comment three times. It is my favorite thing I have ever read of yours. The O’Donoghue quote, the “rubber meets the road” bit, the “Perhaps .12 and/or 90 mph as the cutoffs? Not to be combined?” and the fact that your taught me a new word (“assay”).

        Awesome.

        And on top of everything else, a dream date? Watching Blake Griffin?

        Pinch me!Report

        • Tom Van Dyke in reply to Tod Kelly says:

          You do bring out the best in me as an American, Mr. Kelly. For that alone, I am in your debt.Report

          • Tod Kelly in reply to Tom Van Dyke says:

            Your O’Donoghue quote reminded me of a pretty great essay by PJ O’Rourke about how to drive drunk. The two bits of advice that stuck with me were:

            1. Always use a rental car, because they go faster and turn tighter at high speeds and fit into too small places better than your own car, and my favorite

            2. You wanted to drive a really, really big car. If you hit a pedestrian in a Honda it was right there in your face. But hit someone in a Lincoln Continental and sure, it was horrible, but it was a long way away. Like children starving in Africa.Report

            • Tom Van Dyke in reply to Tod Kelly says:

              O’ Rourke, exactly. He’s grown quiet or wussed-out by having to draw a buck from PBS or Atlantic and such. Author of All The Troubles in the World, where he took a personal tour of the world’s shitholes to find free enterprise as its only buoyant feature; Parliament of Whores [no explanation needed], and Don’t Vote, It Just Encourages the Bastards.

              And of course, the classic text to which you refer from Nat’l Lampoon,

              How to Drive Fast on Drugs
              While Getting Your
              Wing-Wang Squeezed
              and Not Spill Your Drink

              “Even more important than being drunk, however, is having the right car. You have to get a car that handles really well. This is extremely important, and there’s a lot of debate on this subject – about what kind of car handles best. Some say a front-engined car; some say a rear-engined car. I say a rented car. Nothing handles better than a rented car. You can go faster, turn corners sharper, and put the transmission into reverse while going forward at a higher rate of speed in a rented car than in any other kind. You can also park without looking, and can use the trunk as an ice chest.”

              That last bit I have used to great fun and profit. This particular text changed my life, Tod. Some say the Bible; others say Immanuel Kant, although they never actually read him because you actually can’t.

              Peej’s seminal essay can be read here in all its actual glorious and rational fullness.

              http://www.heretical.com/miscella/reptile.html

              DO NOT Drive Fast on Drugs While Getting Your Wing-Wang Squeezed before reading this. Word up.

              Thank you, Friend RTod, for reminding us that to understand our own lives, we again and again and again must return to the classics. Our mutual friend RufusT surely agrees.Report

      • Patrick Cahalan in reply to Tom Van Dyke says:

        > taking in Blake Griffin leading my
        > beloved Clippers to victory.

        This tells me quite a bit about you, sir. A Clippers fan. The greater Mr. Van Dyke comes into more clarity.Report

  2. Ryan Davidson says:

    Written like a true risk analyst. Human beings are *really* bad at doing this without training. The tendency is to over-estimate infrequent, catastrophic events and underestimate common, “low” severity events.

    The US has experienced something in the neighborhood of 3,200 terrorism-related deaths since 1968. Maybe more if you could things like lynching as “terrorism,” which certain parties do. Almost fifteen times that many people died in car crashes in 2005 alone.

    I think part of the reason people make such a stink about security theater but don’t actually care all that much is that unlike vehicular black boxes or breathalyzers, the TSA doesn’t actually bother us all that much. Ten minutes in line and ten seconds in the scanner is way less intrusive, especially since most Americans don’t actually fly all that much.Report

    • unlike vehicular black boxes or breathalyzers, the TSA doesn’t actually bother us all that much. Ten minutes in line and ten seconds in the scanner is way less intrusive, especially since most Americans don’t actually fly all that much.

      Which is why it’s easy in principle to get outraged by the TSA but almost no one is really doing much about it.

      The thing is, I suspect that if something happened to get us over the hump and mandate black boxes, it wouldn’t be very long until we began embracing them. I can certainly tell you that at least one lawyer in every auto versus auto case would like having that data.Report

      • Tod Kelly in reply to Burt Likko says:

        ” I can certainly tell you that at least one lawyer in every auto versus auto case would like having that data”

        Would they? Or would bring down their value, since you would rely less on their legal expertise and ability to steer a jury (for the bigger jury cases)?

        I’m curious what your take on that is.Report

        • Burt Likko in reply to Tod Kelly says:

          Let’s say there was an auto v. auto lawsuit in which liability was disputed. The black boxes from the two autos would make reconstruction of the accident much more reliable.

          Don’t worry about us lawyers. In a high-dollar value case with disputed liability, we’ll find a way to make ourselves useful; the boxes would provide us with even more information about which to argue. In a low-dollar value case, the boxes would make determining liability that much easier.Report

        • Ryan in reply to Tod Kelly says:

          I’m currently a defense attorney doing almost exclusively torts. Auto accidents are my bread and butter.

          Liability is only one of the things that can be disputed in an auto crash. I’d say more than half of the time we don’t even bother, because it’s damn well clear who is at fault. And where liability is questionable, in my experience it’s almost always the defendant who has the better part of the argument. Because insurers have to pay defense costs, they’ve got every incentive not to waste time on issues they can’t win. When a defendant disputes liability, the outcome most of the time is either a defense verdict or a significant reduction in damages due to attributing some portion of fault to the plaintiff. So I’d say that the defense bar would really like black boxes, but the plaintiffs bar probably wouldn’t. Ambiguity almost always works in the plaintiff’s favor.

          But even having the black box data and a breathalyzer won’t necessarily determine liability. Was the light red or not? Who had the right of way at a stop sign? Was the defendant vehicle in the center lane or the right lane? Black box isn’t going to tell you any of that, even with GPS, as it’s only accurate to a few meters, which is more than enough space to raise liability issues. Even if the black box has time stamps, that isn’t necessarily going to resolve anything, because differences of just a few seconds between chronometers can produce drastically different outcomes.

          That aside, what’s disputed in almost every case, whether or not liability is in question, is damages. Health care is expensive, to be sure, but figuring out which bills are actually attributable to the accident is both controversial and downright hard. I’ve seen plaintiffs try to get the defendant to pay for every single medical bill they incur after an accident, including diabetes, cardiac problems, and arthritis, all of which were present before the accident. Clearly, insurers don’t want to pay for that stuff.

          So even if we can completely remove liability as an issue in auto accidents–which we can’t–there will still be significant fights about damages. The tort bar will still be in business.Report

          • boegiboe in reply to Ryan says:

            Black box isn’t going to tell you any of that, even with GPS, as it’s only accurate to a few meters, which is more than enough space to raise liability issues. Even if the black box has time stamps, that isn’t necessarily going to resolve anything, because differences of just a few seconds between chronometers can produce drastically different outcomes.

            You don’t need chronometers if you’re using GPS; the GPS signal carries the exact time. Also, a single GPS may only be accurate to a couple of meters in real-time, but it wouldn’t be hard to record the satellite information being used by the GPS to determine location. Then, after the fact, that data could be compared from the two black boxes for a very accurate relative trajectory between the two (or more) vehicles.

            We do this sort of thing real-time on spacecraft, like the space station. It should be possible to get down to centimeter-level accuracy for a relative trajectory between multiple vehicles without much expense.

            Maybe lawyers will start having to hire guidance and navigation folks like me for their cases!Report

            • Ryan in reply to boegiboe says:

              It isn’t only cars that we’re worried about, unfortunately. We’d also need to compare the recorded time in the black boxes to the timing systems in traffic signals, which there is no reason to believe are coordinated. Indeed, right now this data isn’t even always recorded, so we’d still be relying upon eyewitness testimony to tell us whether the signal was yellow or red.

              But note that two of my examples there don’t actually have anything to do with the relative position or velocity of the cars. The issue of right of way and signal color are independent of either of those things and will depend on human judgment and testimony.

              Besides, what you’re describing would not be cheap, which is actually a decent reason to be skeptical of the whole thing. You’re basically talking about the creation of an entire new expert discipline in accident cases, one which both sides would feel almost obligated to consult. This could easily add several thousand dollars to the cost of litigating every single auto accident, even the ones where there’s only a few thousand dollars at issue.Report

              • Kim in reply to Ryan says:

                rofl. in pittsburgh, people break right of way all the time. It’s called the Pittsburgh Left, and it’s something that people just understand around here. Maybe we ought to just have left turn lights… but that would take more work than just making a Midwestern Consensus about what People ought to do.Report

              • Burt Likko in reply to Kim says:

                I’ve seen it in Boston — light turns green, and the first guy turns left, boom-right-now, before the opposing traffic proceeds straight through the intersection.

                Insane.Report

              • Ryan in reply to Kim says:

                Exactly what I’m talking about. But a black box isn’t going to help you establish who had the right of way any more than plain-old witness testimony will. Exactly what the driver was doing is usually less important than one might imagine, because the results are usually pretty obvious.Report

          • Burt Likko in reply to Ryan says:

            Velocity at time of impact would be a readily-ascertainable datum to gather from a black box. From my taste of your line of work, I gathered the suspicion that vehicles both plaintiff and defendant were moving faster than anyone would admit (people are bad at estimating speed too); the accident reconstruction guys would usually say so in cases that had enough economic stakes to justify them. A plaintiff would probably like to have that information handy to justify her claims of closed-head brain trauma etc.Report

            • Ryan in reply to Burt Likko says:

              Or not, because, as you point out, the odds are pretty decent that if plaintiff vehicle was moving, it was going too fast as well, which would lead to a proportional reduction of fault.

              Plaintiff attorneys generally like painting with broad strokes, hoping that if they can make things sound bad enough, the jury will be sympathetic enough to give them some money. Defense attorneys generally like hard data, as this either gives them the justification to settle more quickly, or gives them the information to win in court.Report

    • Michael Drew in reply to Ryan Davidson says:

      Interestingly, Richard Posner argues in this book that we drastically under-account for the negative value of extremely unlikely but extremely highly destructive possible events from a strictly rational point of view. Perhaps not from an actuarial perspective, because these are the types of events that would essentially void essentially all explicit material contracts in the world. But that doesn’t mean that in overall risk-assessment terms, Posner’s point isn’t valid. Tod can probably fill us in more on how such extreme low-probability, high-damage events are treated in his profession.

      I should add that this doesn’t negate your view that we still regard low-probability, medium-to-high damage events as greater risks than they in fact are to the myriad high-probability, low-damage events and conditions we court routinely every day.Report

      • Tod Kelly in reply to Michael Drew says:

        Ooo, cool book! Thanks, Michael!

        Regarding this question: ” how such extreme low-probability, high-damage events are treated in his profession” The answer is: like the boogey man. What’s hard for most people to understand – even me – is the clockwork precision that most predictions of financial insurance losses come in correct. Seriously, it’s scary and sometimes creeps me out. We used to manage a really small self insured group that handled workers comp claims, that only had about $3 million in premium. The underwriters would shoot for a very exact loss ratio, and it was almost always on damn dot. If it was off 5% everyone would freak out and rebuild the model to improve it.

        But those very, very low frequency/very, very high severity losses are impossible to predict as well -and those are just the one’s you had the foresight to predict in the first place. (And nobody predicts all the bad things.)

        Usually great reliance is paid on reinsurers.Report

        • DensityDuck in reply to Tod Kelly says:

          Maybe the only good thing about major disasters is that they hit everyone at once, so you can always say “well at least I didn’t do anything uniquely wrong”.Report

        • Kim in reply to Tod Kelly says:

          >those very, very low frequency/very, very high severity losses >are impossible to predict as well.
          Depends. I firmly believe that people think about airplane crashes a lot more than car crashes — but that’s mostly a result of propaganda. I DON’T think most people consider the possibility of losing their house because BoA forgot to pay their taxes, and the sheriff shows up and forecloses on the house. Ain’t Escrow Fun?

          In fact, I believe that Democrats tend to see more of the high damage, low risk side of things, and Republicans tend to see more of the low damage, high risk side of things. Or maybe they both see them, but only one bothers them. Because corporations (oligopolies) tend to cause high damage, low risk (open jaw, anyone?), and government tends to cause low damage, high risk (waiting thirty minutes in the DOT, for something that could be done online in five).Report

      • Patrick Cahalan in reply to Michael Drew says:

        The Posner book is excellent, btw.Report

  3. DarrenG says:

    Because in our oddly bi-chromatic world of political tribalism, those for the death penalty are usually against abortion and vise versa.

    Not true according to most readily-available polling data.

    More importantly, I think you missed the main, and to my mind, most important objection to mandatory black box telemetry in cars — lack of trust for the institutions that would have access to that data.

    I think many of us are rightly skeptical of what insurers and various government agencies would do with this data, and the potential costs and risks there outweigh the likely safety benefits. And as to the likely safety benefits, many of us also remember the dire predictions from insurers over the elimination of the national 55mph speed limit that turned out to be entirely specious.

    In both telemetry data cases I think you’ve missed on the comprehensive analysis, and made the mistake of assuming the only cost is from the required equipment and installation when the total potential cost burden is much higher and more complex.Report

    • Tod Kelly in reply to DarrenG says:

      @DarrenG: “In both telemetry data cases I think you’ve missed on the comprehensive analysis,”

      And I think you might have missed the point of the post. Again, I am not advocating either for or against these potential mandates. Rather, I was trying to argue that although we all tend to think we value human life and people on the other side of the fence don’t, the truth is a great deal trickier. The issue for me, if you will, is not Public Policy so much as Fundamental Attribution Error.

      Still, since you’ve taken the time to respond, let me do likewise:

      Regarding this: “Not true according to most readily-available polling data.” OK. That does not mix with my experience, but that after all is anecdotal. (Plus, it’s anecdotal in the Pacific NW, which means even less. We’re a bit off, culturally.)

      Regarding this: ” I think you missed the main, and to my mind, most important objection to mandatory black box telemetry in cars — lack of trust for the institutions that would have access to that data.” This would carry more weight with me were people willing to answer differently when given circumstances that don’t require data to exchange hands – such as the built-in breathalyzer. This is not to say that those Big Brother fears are unfounded and that giving the government the ability to tap into that data a bad idea… it’s just that doesn’t seem to be the reptilian reason people respond the way they do.

      Regarding this: “many of us are rightly skeptical of what insurers and various government agencies would do with this data” I get what the objection for government agencies would be, but what is the objection for insurance companies? Again, I am not advocating this policy, but whether or not you speed or run traffic lights is legitimate underwriting information, is it not? This is not a “challenging you” question, I am really curious as to what your objections is.

      Regarding this: “the dire predictions from insurers over the elimination of the national 55mph speed limit that turned out to be entirely specious.” I am unaware of what you are referring to. Might this have been a regional thing from your neck? If so I would be curious to see more about it. Underwriters I know don’t care much about speed limits. This is not because driving at 70 is or is not less safe than driving at 50, although it is. (The reason has less to do with speed than behavior. If you look at cars on semi-busy highways, people use the same amount of space between cars at 55 that they do at 75, even though they need more reaction time. That’s why most states gauge their over 55 limits with traffic volume.) Underwriters don’t care about losses they can predict, and if you change the interstate in your state to 85 they can predict very, very accurately what this will can in changes of incidents and adjust accordingly. Again, if you have seen insurers lobby for lower speed limits I would be fascinated; please link!

      Is it possible that you are thinking of the federal law to max limits at 55 30 years ago? If so, this was not a safety issue, but a fuel efficiency issue during the energy crisis (of that era). I honestly cannot speak as to its effectiveness, other than to irritate everyone.Report

      • Will Truman in reply to Tod Kelly says:

        Regarding this: “the dire predictions from insurers over the elimination of the national 55mph speed limit that turned out to be entirely specious.” I am unaware of what you are referring to. Might this have been a regional thing from your neck? If so I would be curious to see more about it.

        I would guess it’s a reference to the IIHS’s efforts. They’re constantly making noise about speed limits, red light cameras, and so on. They don’t donate the campaigns, but they file briefs, get petitions started, and so on.Report

      • DarrenG in reply to Tod Kelly says:

        I was trying to argue that although we all tend to think we value human life and people on the other side of the fence don’t, the truth is a great deal trickier.

        In that case I think you’re largely arguing against a straw man, as I think political partisans of all sorts would quickly concede that they are not in favor of banning or restricting every human activity that has a non-zero risk of mortality.

        what is the objection for insurance companies? Again, I am not advocating this policy, but whether or not you speed or run traffic lights is legitimate underwriting information, is it not?

        Two objections: One is that this data wouldn’t just be used for legitimate underwriting. Obvious concerns would be using the data to ratchet up rates unidirectionally regardless of risk of loss, using the data for marketing purposes, selling the data to third parties, and/or inadequately securing the data against hostile third parties.

        Another objection would be using the data with insufficient context. Sometimes the safest course of action is to accelerate out of trouble. Should my rates go up if I speed briefly to avoid an otherwise-unavoidable accident? What sort of data signature is evidence of unsafe speed, exactly, and how transparent is this analysis?

        Again, if you have seen insurers lobby for lower speed limits I would be fascinated; please link!

        Ask and ye shall be linked:

        http://www.cato.org/pub_display.php?pub_id=5443
        http://www.ibiblio.org/rdu/lyingwsc.htmlReport

        • Tod Kelly in reply to DarrenG says:

          “In that case I think you’re largely arguing against a straw man, as I think political partisans of all sorts would quickly concede that they are not in favor of banning or restricting every human activity that has a non-zero risk of mortality.”

          Dude, no offense, but did you even read my post? Where did I argue that?Report

          • DarrenG in reply to Tod Kelly says:

            I quoted the part of your reply where you restated your intent. To wit:

            I was trying to argue that although we all tend to think we value human life and people on the other side of the fence don’t, the truth is a great deal trickier.

            I don’t see anyone seriously arguing the universal proposition that “people on the other side of the fence don’t [value human life].” They may argue it in the specific context of abortion, health care, war, the death penalty or whatever, but when you extend it to arguments about auto safety you’ve stretched it past the breaking point, I think.Report

            • Tod Kelly in reply to DarrenG says:

              Darren, this post was written in response to arguments -here – where people on both sides of the political fence argued with one another that their side was the one that valued human life and was consistent with that belief, and that at the other not so much.

              The point I am trying to make is that we all have a tendency to believe this – that human life to us is valuable to the point of being sacred, but that we don’t always act that way when that life is separate enough from us to be off of our radar.

              I used auto safety because there are things you can do that can quantifiably save lives that do not infringe on our freedoms in a gulag-extreme way, and we reject those things. I’m not even arguing from a holier-than-thou point of view, because I reject those things too.

              What I was saying to everyone here that was so sure they were arguing from a place of altruism while their opponents were arguing from a place of self centeredness was that how we choose to deal with human life is more complex than that.

              You may or may not agree with that sentiment, and you may not agree that I used the best vehicle (ha!) to prove my point (but as I say in the post, I chose it because this is what I do), but does what I am saying at least make sense?Report

              • DarrenG in reply to Tod Kelly says:

                I think it makes sense, I’m just not sure it’s as relevant to the prior discussion of “who’s really the most pro-life” as you intended.

                Then again, I skipped over most of that discussion as predictable boilerplate, so what the hell do I know?Report

              • Kim in reply to Tod Kelly says:

                I don’t mind breathalyzers. Then again, I realize how much of a danger I am just being on the road.

                I question how accurate the models would be, or how much the insurance companies would fairly tell people about their prospective insurance rates, if you use black boxes. If it’s a simple “If you constantly speed, here’s your $100 increase” that’s one thing. if it’s a simple metric like “every time you speed through a light, you pay $5 more next year” that’s another. What I don’t like, is the idea that people will walk into something they can’t afford, and couldn’t budget for.Report

        • Tod Kelly in reply to DarrenG says:

          “One is that this data wouldn’t just be used for legitimate underwriting.” This makes me ask, what is illegitimate underwriting? The point about marketing is an OK one, I guess. But I’m dubious that’s the reason people don’t want their insurers to know if they’re speeding or not.

          “Another objection would be using the data with insufficient context. Sometimes the safest course of action is to accelerate out of trouble. Should my rates go up if I speed briefly to avoid an otherwise-unavoidable accident? What sort of data signature is evidence of unsafe speed, exactly, and how transparent is this analysis?” This isn’t the way these systems work. Again, I think that you are correct that there are reasons to reject them, but effectiveness isn’t one of them. As I said above, employer fleets do see very, very dramatic reductions in both frequency and severity.

          About this:http://www.cato.org/pub_display.php?pub_id=5443
          http://www.ibiblio.org/rdu/lyingwsc.html

          Thanks! I appreciate it. I live for this stuff.Report

          • DarrenG in reply to Tod Kelly says:

            This makes me ask, what is illegitimate underwriting?

            I meant “as opposed to purposes other than underwriting,” although I suppose my first objection could also be framed as “illegitimate underwriting” (i.e. using black box data as a ratchet for steadily increasing, but never decreasing, rates, and increasing corporate profitability.)

            As I said above, employer fleets do see very, very dramatic reductions in both frequency and severity.

            I think a key difference is that company and trucking fleets come with the added disincentive of employment-related sanctions.

            Overall, though, I don’t think this is a simple matter of “this will make you safer, so you don’t *really* value human life if you don’t assent to this.” The American public has been sold enough snake oil under this slogan that we’re rightly suspicious whenever the government or a cartel of large corporations tells us “we need to closely monitor your activities for your own good.”Report

            • Tod Kelly in reply to DarrenG says:

              ” I don’t think this is a simple matter of “this will make you safer, so you don’t *really* value human life if you don’t assent to this.””

              Neither do I.Report

              • Tod Kelly in reply to Tod Kelly says:

                And, by the way these are fair points:

                “although I suppose my first objection could also be framed as “illegitimate underwriting” (i.e. using black box data as a ratchet for steadily increasing, but never decreasing, rates, and increasing corporate profitability.)

                As I said above, employer fleets do see very, very dramatic reductions in both frequency and severity.

                I think a key difference is that company and trucking fleets come with the added disincentive of employment-related sanctions.”Report

      • Kim in reply to Tod Kelly says:

        Tod,
        Legit: whether you routinely speed through traffic lights.
        Not Legit: whether you have a red car.
        Let’s guess which one auto insurance currently uses, shall we? 😉

        Most Democrats, who tend to be “pro legal abortion” are also pro death penalty. Because liberals aren’t half the democratic party. Me? I’m Liberal, but I’m kinda agnostic towards the death penalty. To me, it’s a “shrug” question — arguable, certainly, but I’d rather work on reducing prison populations in general first.Report

  4. Burt Likko says:

    How difficult would it be to engineer an auto breathalyzer to make it cheat-proof? And what about the question of emergency overrides?Report

    • Tod Kelly in reply to Burt Likko says:

      As I’ve said above, I had not meant this to be a public policy argument so much as a Fundamental Attribution Error argument; I’m not advocating these mandates.

      That being said, if you’re looking for an actual answer, regarding the cheating: I do not believe you can have any risk management system that can’t be cheated. Most risk management is less about eliminating problems – because you can’t, any more than you can insure against any outcome. It’s about reducing the number of occurrences that instigate problems. So if you have a breathalyzer system and it stops 9 out of 10 dangerously intoxicated people from driving, but doesn’t catch the cheat (until they crash, I suppose), in our industry that’s chalked up as a win.

      The question of emergency override is an interesting question though. I can’t say as I know enough about the breathalyzer systems to say whether or not they do have one, or how they might work.Report

    • DensityDuck in reply to Burt Likko says:

      The kind of people who think that breathalyzer interlocks are a good idea think that if there’s alcohol in your system then you shouldn’t be driving end of sentence full stop PERIOD. They don’t care if your wife is giving birth while your house burns down. That just serves you right for drinking.Report

      • Jesse Ewiak in reply to DensityDuck says:

        If your house is burning down while your wife is giving birth, I’ve hoped you’ve called 911 and someone sober can drive your wife to the hospital in a nice big truck with colored lights.Report

      • For the circumstances you describe (house burning, wife giving birth, E.T. going home), I’d support an emergency override. But outside according-to-Hoyle emergencies, I believe people shouldn’t drive even after having one drink: not so much because one drink necessarily makes that much difference, but because “one drink” makes it easier for someone to drink two drinks, or more, and still choose to drive.

        By the way, I’m not saying the state need ban driving by people with any alcohol whatsoever on their breaths. I do think it would be good if people followed that rule. I also think it would be good if people were perfect, too.Report

  5. Christopher Carr says:

    Tod, you’d love Japan. Zero tolerance for drunk driving: no rum cakes or you go to jail. But there’s widespread public transportation and infrastructure, everywhere is urban and well-lit, and a special profession called “daikousan” will actually drive your car home for less than the cost of a regular taxi ride if you suddenly decide to get wasted.Report

  6. Renee says:

    I think what Ryan Davidson wrote is right: we are really bad at thinking low probability events and there is a good reason for it!
    In 2010, there were 1.09 traffic fatalities for every 100 million miles traveled. So when I embark on a 100 mile journey (a not insubstantial trip), my odds of being involved in an accident are 0.00000109. Let’s (ridiculously) hypothesize that “speeding” would double my chances of an accident. I’m now up to 0.00000218. How hard is it to tell the two apart? Suppose I have a 1-million sided die. It either shows “1” with probability 1 in a million or 2 in a million (all other numbers have equal prob). How many rolls would it take for me to have 95% certainty which case I have? (Using a one tailed test, normal approximation to the binomial) Well over 14 million rolls! Probabilistically, I would have to travel over 1,400 million miles (assuming I could continue after a fatality) to know the difference between the two distributions to 5% uncertainty. So although the risk of fatality had doubled, I won’t really be able to tell in my lifetime of driving.
    But I am 100% certain that my privacy/security/etc has been violated if a black box is in use.
    Like you, I don’t write any of this to advocate or denigrate the black box policy – I merely offer it to justify why people may trade off an extremely unlikely tragedy with an assured inconvenience from their individual perspective.Report

    • Renee in reply to Renee says:

      Oops – I tried to whip out the maths and made an approximation (normal) when I probably shouldn’t (p very close to 0) and just plugged into an equation. In any case, my blunder lowballed the number of rolls (100 miles traveled). In wikipedia I trust, (http://en.wikipedia.org/wiki/Sample_size_determination), and I get that you need something on the order of 10^12 rolls in order to get 95% confidence to detect a probability difference of 1 in a million. The conclusion holds: it is impossible for an individual to be able to tell the difference in these probabilities.Report

  7. James K says:

    I’m always gratified to hear someone talk about risk who actually understand it, and what you identify is a truth every economist knows, but few will utter in public forums. And that is that people put a much lower value on human life they they claim to.Report

    • Ryan in reply to James K says:

      Well… sort of. I think that a lot of people are instinctively and probably unconsciously making a decision between intentional and unintentional acts.

      Abortion and the death penalty represent some combination of policy decisions and deliberate intentional acts by individuals. There’s arguably a difference between that and deaths by vehicular accident, which are usually the result of negligence at the worst, or natural disaster, which is pretty much just random. We can, probably, completely eliminate abortions or executions from our society by saying we aren’t going to do those things anymore. But even if we used black boxes and breathalysers, people would still die in car accidents, simply because people inadvertently make mistakes.

      Insurers make this same kind of distinction. There is, with a few limited exceptions, no liability coverage for intentional acts committed or authorized by the insured. We’ll insure you against liability arising out of the operation of a car, but we won’t insure you against liability arising out of knocking over the liquor store on the corner.

      So as much as I agree that Tod is correct about the way people tend to evaluate risk, I think me might be missing something, i.e. the fact that whether a given cause of death is intentional or unintentional does make a difference, and that it makes sense for people to be more concerned about intentional acts than unintentional acts.Report

  8. Jason Kuznicki says:

    I suspect you would appreciate Robin Hanson’s take on September 11. It’s the only other bit of 9/11 writing I envy, and it’s in most senses superior to mine.Report

    • James K in reply to Jason Kuznicki says:

      It really is an incredible piece.Report

    • Jason – Thanks for the link. This piece makes me remember a lot of people I knew on the left in the 80s. As many idiots on the other side of the fence were saying AIDS victims got what they deserved, I remember a lot of lefties saying making gush embarrassingly gushing about how having AIDS was such a noble way to die.

      I probably don’t have to point out that these lefties were predominantly straights who didn’t know a lot of people who were gay or dying of AIDS.Report

    • I should also say thanks for getting what I was trying to say. I have been surprised over the past 24 hours how many commenters are arguing with me about my desire to increase federal manufacturing regulations. It’s a funny thing, this new bloggy habit of mine.Report

  9. NoPublic says:

    But the free market has already solved this for you. You can voluntarily have one of these devices installed in your car and get lower insurance rates. Viola!

    I suspect that a great deal of this (particularly in the US) is tied to the “90% of people think they’re above average drivers” phenomenon. They think they are more capable and thus the laws are an inconvenience to them and they don’t want to be penalized for breaking them. I’ve known people who drink and drive the same way “I’m fine, I’ve only had three beers in the last hour”.Report

  10. Will Truman says:

    For what it’s worth, my objections to a lot of the speeding and red light camera things really are cause-specific. In other words, if you combined red light cameras with longer yellow lights, I’m game. Instead, there are reports of tinkering with the yellows in the opposite direction. There are a lot of other things they can do to stop incidental lawbreaking and the pursuit of revenue-enhancing alternatives makes me wary.

    If it ever came to be that they were serious about actual public safety, I genuinely would be more open to everything from red light cameras to speeding cameras and so on. Especially if we throw in a conversation about what reasonable speed limits should be (and on this, I confess, my views do accept fatalities in exchange for increased efficiency). It’s the “gotcha” aspect of it that annoys me. Get rid of that and my perspective changes.

    I will confess, however, that a lot of my comrades really just want to speed and run yellow lights. Just as much as municipalities want to enhance revenues.Report