The thread on ‘evil drug companies’ got me thinking about the legal system, and how it’s falling down when trying to assign guilt in highly technical cases. I’m wondering if it isn’t time to modify the system.
It just doesn’t make a lot of sense that a jury of lay people should pass judgement on someone who may have been negligent in a highly technical area.
For example, let’s say I’m a nuclear engineer, and there’s an accident at the plant. I’m charged with negligence, and have to defend myself against a jury.
Now, the original concept of a jury was a group of your ‘peers’ who could reasonably put themselves in your shoes and come to a decision about your judgement. Does that really hold when the nature of your error was, say, neglecting to consider the Bayesian analysis of a statistical evaluation of risk? What the hell do they know about it, and how can they be expected to know if what you did was reasonable?
In these cases, the plaintiff and defense have to hire ‘experts’ who will give testimony to the jury to help it make its decision. But doesn’t that just shift the decision onto one ‘expert’?
These cases often come down to which side could buy the best or most well-spoken ‘experts’. If your expert does a better job than mine, I lose.
Top-rank experts can be hideously expensive, which means if you’re a person of average means being sued by a large corporation, you’re toast. On the other hand, large corporations often get off the hook for negligent actions because they can afford to buy better experts than the prosecution.
Even more troublesome - the legal system is supposed to be non-biased, but the very nature of expert testimony introduces tremendous bias. First, if 99 'experts say my client is guilty but one says he’s not, I’m going to hire the one who says he’s not. So I’m not presenting a representative viewpoint to the jury. Second, just the fact that I’m paying the guy means he’s biased towards giving testimony I want.
I’m thinking that perhaps a person or corporation charged with negligence like this would have the option of a “technical trial” in front of a jury comprised of experts randomly selected from his field. That jury would become part of the process.
From there, I’m not sure. Perhaps this jury actually makes the determination of guilt, or perhaps it just writes a report that is available to the regular jury.
Comments? Has the legal system been outpaced by the the rapid pace of science and technology?
i notice that you’re posting from Canada. we don’t use juries in civil suits nearly as much as in the United States, for some of the reasons you suggest. highly technical cases are more likely to be tried by judge alone, on the theory that a judge is more able to assess expert testimony, plus can reserve judgment for several moths if necessary to go over all of the materials.
with respect to your example of picking the 100th and only expert who supports your position, trust me that if I’m the lawyer on the other side and have done my homework, I’ll make sure that the court knows that you’ve chosen the one in one hundred to support your position, and attack him accordingly. that’s the purpose of the adversarial system.
Recently (well, ok, it began a couple years ago), a federal judge hearing some of the cases against breast implant manufacturers essentially decided the same thing – it would just be one expert against another, with a non-technical jury left to flip a coin.
So he formed a panel of objective scientists to review the evidence and data. They took a while (which many courts don’t like to do), but they came back and said there was essentially no evidence to back up the claimants.
This is the way such trials should work. Otherwise, the court can too easily out-pace the science.
Modify it how? I by no means think the system is perfect, but I’m not sure how it could be improved in this respect.
Why not? The question presented to the jury is not a high-tech one. The question is: was this person negligent in the performance of his or her professional duties? Negligence is usually fairly easily explained to lay-people: it consists of duty, breach of that duty, the causation of damages by that breach of duty, and the actual existence of damages. If any one element does not exist (duty, breach, cause, damages), then there’s no negligence.
The way that expert testimony comes into it in high-tech cases is that different professions have different “standards of care” that they must take in pursuing their profession. In other words, what the duty a particular person must professionally meet depends on what the “standard of care” is for his or her profession. Expert testimony is usually offered to explain that professional standard of care, and then to offer an opinion on whether the person in question violated that duty by failing to meet that standard of care.
Actually, I think it does. There is a specific duty of care for doctors, lawyers, plumbers, teachers, engineers – are each of them entitled to a jury of their professional peers? The courts have repeatedly held that they are not. Why? Because 12 doctors sitting in judgment of a doctor (and thinking, there but for the grace of God go I) may not judge the doctor impartially. You seem to be arguing that there are technical jobs or processes that are SO complicated they are beyond the grasp of the average person. I don’t think this is so. My experience has been that complex concepts must be carefully broken down for laypeople and that the precise nature of the negligence (or the defense) must be explained in the simplest possible terms. Experts are often of very little help in this regard; they tend to talk (as we all do) with the assumption that the people they are addressing know the basics of the field already – an assumption it’s never safe to make with a jury. In other words, they talk over the jury’s collective head anyway.
And there is some truth to this. But the expert is not allowed to just sit on the stand and pontificate about his or her opinion. He or she is subject to cross-examination and must defend that opinion during that process. Any decent defense attorney will shred an expert how opinion is not truly defensible. Moreover, attorneys in the U.S. are allowed to elicit from the expert the amount of times that he or she has testified; either for plaintiffs or defendants; and for how much money. Juries are quick to spot “prostitutes” – experts whose opinions can be bought and will change depending on whether they’re testifying for the plaintiff or the defense.
I don’t think this follows. Expertise is based on a person’s qualifications in their professional field – not their time on the witness stand. Therefore, a good expert (a plumber, for example) may be one who has been in the business for thirty years, but who may not have ever testified before. A professional expert (one who makes the court rounds) is often not a good witness, because the opposing side will elicit that what he or she really does is testify at trials, not work in his or her field. Plus, they come across as “prostitutes.”
The legal system, in every respect, is not supposed to be non-biased. The plaintiff’s case is the best one possible for the plaintiff and the defendnat’s case is the best one possible for the defendant; the non-biased parties are the jury and the judge. If a jury is presented with duelling experts, they are expected to determine who they believed, and why. Will they sometimes believe the guy who is slicker and better-spoken? Sure. But I’m not sure what the answer to that problem would be.
I, as the opposing attorney, will put on three experts who say that your expert is full of shit. I will also poke enormous holes in his opinion on cross, possibly making him look like a total idiot if his opinion is truly indefensible. Paid testimony does not necessarily indicate bias; an expert, having no relation to the case, may reasonably expect to be compensated for their time. But professional integrity (in EVERY profession) will generally keep people from testifying in court to an opinion they don’t truly hold, just to get a check. People who are willing to do so almost inevitably come across as what they really are – prostitutes.
So would doctors be judged by doctors, and lawyers by lawyers? I don’t think that would be “a jury of your peers.” I think it would be a jury that might well be partial, as opposed to impartial, for the reasons given above.
DAVID B. says:
I disagree. It is not the function of a panel, impartial or not, to determine whether or not there is “evidence to back up the claimants.” That’s the precise function of the jury. In so doing (if that is in fact what happened), the judge abdicated his responsibility in favor of the panel. It’s a fine line that we’re talking about but, in my opinion, an important one: the judge and/or jury should have resources available to help them understand the evidence presented to them, but the function of weighing that evidence ought to remain the province of the trier of fact – a jury, if there is one, the judge if there is not.
Hello all. De-lurking for the first time here…be gentle.
I, until recently, worked at the Association for the Advancement of Science, and recalled that they are actually very busy on this very issue. Thought it might be helpful to provide this URL to help the discussion along:
Jodi: The reason I brought the whole thing up is because it doesn’t look like the legal system is working well at all in these areas. There have been a number of lawsuits that were won or lost based on ‘junk science’. Especially in areas concerning public safety.
What made me think about this was the Ford Pinto case, in which a Ford statistical risk analysis document was used by the plaintiff and jury to put the boots to them. I don’t want to argue the merits of Ford’s case, but these risk analyses have been cropping up a lot as plaintiff’s tools, and juries seem to accept them as a priori evidence of wrongdoing. To me, this indicates a breakdown between the way a typical jury member sees issues of safety versus, say a safety engineer’s view.
Another example would be a medical study which shows a causal relationship within 2 standard deviations of error. This may be a reasonable standard for one study, but due to Bayes Theorem it’s a terrible standard for evaluating a whole pile of studies, or studies which are purporting to analyze a rare defect.
A two-standard deviation acceptance rate means that 5% of all studies done will have incorrect conclusions. A 5% error rate is fine when you start your own study, but when you are looking at historical groups of studies and ‘data mining’ for studies which prove your case, you’ll find lots of them that do simply because of random factors. But haul four or 5 of these into court that ‘prove’ your case, each of which claimed to have an error rate of less than 5%, and the jury is yours.
No. The jury determines innocence or guilt. The judge was using a scientific panel as a gatekeeper to determine what witnesses should be allowed. And I think it was precisely what was necessary. As it was, a company was bankrupted because of unscientific claims about its products (breast implants) and because juries decided the bogus “experts” were more believable than the real ones (or just decided to sock it to the big evil companies who obviously hurt these poor innocent women).
I disagree – I think he upheld his responsibility to have the fairest trial possible.
I guess I was presenting my opinion, in a way, with my previous posted link as I think that taking expert selection out of the hands of the lawyers and into the hands of an “impartial” judge is, in theory, a good plan. To allow for a “peer review” of expert witnesses takes it one step better IMHO.
Of course, this is not a perfect solution, as only “accepted” science would make the cut, and as impartial as scientists claim to be, they are human and would tend to select those individuals who toe the proverbial line. One would think that the law and science would be a perfect marriage, as both purport to deal only with facts. In practice, however, I think both provide an atmosphere where going “out on a limb” is quickly quashed, even when there is evidence or data to back it up.
Perhaps what would be ideal, would be for both sides to allow their chosen experts to collect the data that they feel best supports their case, but for presentation of data to allow a court appointed expert to impartially present that data, with no mention of their own opinion. This would probably open a whole other can of worms, but would at least get past the problem of juries being swayed by a charismatic expert as opposed to the validity of the data presented.
“Junk science” is a problem within the legal arena, but I don’t think it’s a problem of the legal arena – it’s a problem of science. What I mean is that, ideally, only accepted scientific theory should be presented in a court room, for the obvious reason that WAGs do influence the jury, who might take an unproven hypothesis to have the same value as accepted theory, so long as it’s presented in suitably scientific techno-babble. If a scientific theory cannot be proven to be of value in determining the questions before the legal tribunal, then it shouldn’t be admitted. The best example of this, IMO, isn’t a “hard-science” one, but a phsychological one – recovered memories. As a theory, I think the validity of “recovered memory” has been, or is in the process of being, debunked. But dozens of people have gone to prison during the short period of time in which it was (wrongly, IMO) allowed in the court room as “accepted science.” The flip side of this, of course, is complaints that the law is too slow in accepting the validity of new scientific theories or processes, such as DNA evidence. Well, this is why.
Ah, risk-benefit analyses. I think there’s a good argument that RBAs are almost always prejudicial. Why? Because the average person does not like to see a company toting up the costs of human life – even one – versus company profits. Is it an accepted statistical means of determining the whether a particular course of action (a recall or a repair) is justified in terms of money? Yes. But it remains really, really unsavory, for obvious reasons. The problem is not that juries do not understand the workings of RBAs, but that they do not like them. Does that mean that Ford is entitled to a jury of people who are comfortable with the cold calculation of RBAs? IMO, no. You are entitled to a jury of your peers, and the fact is that most of your peers hate RBAs. If you think they are acceptable anyway, and it comes up at trial, you and your lawyers must do your best to carefully explain how and why RBAs are used. But juries generally understand RBAs – that’s the problem.
And, as opposing counsel, this is what you need to get before the court. If the counsel I am up against is using faulty theory to make their point, I am going to object to it – prior to trial, if possible, in an attempt to have it excluded. If I can’t get it excluded, I will explain to the jury (through my own expert, if possible) why extrapolating the results of one study through a number of studies is not good science. Again, expert testimony is not offered in a vacuum. It can and should be subject to cross-examination and counter testimony by other expert witnesses.
DAVID B. says:
And, by so doing, it took upon itself the duty of weighing the sufficiency of the evidence and usurped the province of the jury. By acting as a “gatekeeper” and evaluating what evidence the jury will or will not get to see, the judge is controlling the evidence upon which the jury will make it’s determination of guilt or innocence. That is simply inappropriate. If the evidence is compentent, relevant, and material, it should be presented to the jury. If it is not – if it is irrelevant, for example, because it is cumulative of other evidence already presented – then it should be excluded by the judge in his role of controlling the judicial proceeding – not in a role of evaluating the evidence, which is manifestly beyond the duty of the judge in the context of a jury trial.
The “fairest trial possible” is obtained by allowing our accusatory system to function as designed – one side putting on its best case through direct testimony (including experts) and evidence; the other side putting on its best case through cross examination and evidence. I do not agree that the “fairest trial possible” is obtained by taking the duties of the jury and vesting them in the judge, or in some panel called by the judge. If the jury requires assistance in evaluating the evidence put before it, fine; but the fundamental task of determining the weight and sufficiency of the evidence provided should not be taken from it.
I agree with you completely – and am trying to figure out how what you said is any different than what I related in regards to the breast implant case I mentioned. Maybe I did a poor job of describing it, but that case seems to fall into your description here.
You’ll only get 100% agreement from me on this one. But, again, how should the judges have determined that this stuff was crap and disallowed testimony on it? Should they have used objective panels to investigate it? Or some other method?
I have to admit, I’m confused. As I asked earlier in this message, how is this different from what you described?
But what if it is not competent because it is the result of poor science? Is the judge supposed to make that decision by himself, or isn’t he better off asking people in that field to take a hard look at the evidence and then advise him?
This seems to contradict what you’ve said above. If one side’s “scientific expert” is simply full of shit, why should he even be allowed to muddy the waters for the non-technical jurors?
A judge usually “investigates” an evidentiary ruling by demanding that the parties justify the science they are attempting to use, either orally, in court, or through briefing. Usually it would be through briefing, because any attorney worth his or her salt who is facing dubious scientific opinion from an expert will move in limine (ie, before trial) to have that opinion excluded. The judge then reads and researches the briefs, and is also free to research the issue independently (through his or her law clerk, almost certainly) and reach a decision. If the judge feels that he or she needs to consult an expert before making that evidentiary determination, he or she may certainly do so.
There may be no difference, except confusion on my part about what you meant. Your initial post seemed to advocate allowing the judge to evaluate the weight and sufficiency of the evidence. (Judge convenes a panel; panel decides there is no evidence; that’s how complex trials should work.) That’s what I disagree with. A judge may make preliminary evidentiary determinations, and in that function certainly may exclude “junk science,” but neither the judge nor any panel should be weighing the evidence – that’s what the jury is for. This is straying away from the OP, by the way, which asked what should be done if the evidence is too complex for a lay-person jury – and evidence may certainly be competent, relevant, and admissible, and yet still very complicated.
If the evidence is not admissible as “likely to assist the trier of fact” (competent), then it should stay out. That determination is for the judge – not the jury – and I have no problem with the judge turning to field experts to make that determination. This is different, however, from taking competent, relevant, admissible evidence away from a jury and presenting it to someone else for consideration simply because the evidence is complicated. That’s what I disagree with.
He shouldn’t be. It is the burden of the party electing to bring the expert to disclose (prior to trial) his or her qualifications and the substance of his or her anticipated testimony. That gives the opposing side the opportunity to challenge the expert on the basis that either the qualifications or the theory is suspect. That evidentiary determination is a preliminary matter and should be made by the judge. But there is a difference between an expert who is “full of shit” (who should not be allowed to testify before the jury) and an expert who is presenting complicated theory (who should). At this point, happily, I don’t think we’re even disagreeing.
If anyone is interested, for an interesting (and short) discussion of the Sixth Amendment right to trial by jury (and how the issue of “compexity” is starting to impact that), I’d recommend In Our Defense by Caroline Kennedy and . . . shoot, I can never remember the other woman’s name. But the book is a great thumb-nail sketch of the Bill of Rights and how it is applied and challenged in modern society.
See, my point is that in some cases the jury simply isn’t technically capable of doing that. Sometimes scientific or mathematical conclusions can be so counter-intuitive, and run so strongly against ‘common sense’, that it’s impossible for people to grasp unless they have specific training in the field.
I have spent a lot of time trying to teach probability theory to people without math backgrounds, and I can tell you it’s like running into a brick wall. You can spout all the fancy high-falutin’ math you want, but if it runs counter to a person’s experience they flat-out won’t believe you. In areas involving random behaviour, the human brain is hugely biased, because it tries to find cause-and-effect where it simply doesn’t exist.
Let’s say that there is a .5% chance that a Chevy S-10 pickup truck will explode in a collision. Let’s say that that this is the norm for vehicles of that type. So if there are 1000 collisions involving Chevy S-10’s, you’d expect 5 of them to explode. But the variance might be ten times that amount. So, though pure variance we have a bad year one standard deviation off the mean, and twenty S-10’s explode in collisions. This makes the news, and now people start reporting every single instance of a an S-10 catching fire. So now we’re selecting data in a biased fashion to compare against other vehicles. When it all comes out in the wash, the prosecution shows that S-10’s have exploded 20 times more often than any other vehicle on the road. Then he produces a report from GM which shows that they KNEW the S-10 would explode X% of the time, but they did nothing because it was too expensive to change the design. Now let the defense put up their expert on statistics who drones on about standard deviation, sample sizes, etc. Good luck convincing a jury.
This exact scenario has led to several false judgements. One famous one was the ‘uncommanded acceleration’ of a certain Volvo model. Several people reported this, it made the news, and suddenly reports started coming in from all over the place. The defense argued that it was just bad luck, combined with biased data and false reporting, but they lost. Photos of burned bodies carry more weight than a diagram of a bell curve. As it turned out (after the guilty verdict), there was absolutely nothing wrong with the car, and never was. It was a pure statistical fluke.
The same thing happens in medical trials. The accuracy standard for most medical studies is 2 standard deviations. That’s a potential incorrect conclusion 5% of the time. That’s a reasonable number when you start a study - 95% certainty is pretty good. But now let’s say 100 such studies have been done in the past year. Just through randomness, 5 of them on average are going to have incorrect conclusions. But if the variance around that number is also high, you might have a year when 10 or 15 studies out of 100 show a problem which doesn’t exist. If you’re sued by a family of someone who died after taking your drugs, and the prosecution presents FIFTEEN studies showing a potential problem, I guarantee you’re going to have a hard time getting your statistical evidence accepted by the jury.
Not to be flippant, but my response to this, in a nutshell, is “tough break.” Citizens in our society are entitled to have their disputes heard by a jury of their peers. (Yes, I recognize this is a gross generalization, and doesn’t apply to suits in equity or suits in military courts or suits in city courts, but generally). It’s a right guaranteed by the Sixth Amendment. Under the Sixth Amendment, you are entitled to a jury of your peers – not a jury of extremely smart people or a jury of people in your field. Does that mean that occasionally something will go over the head of a jury? I’m sure the answer to that is yes. But I do not believe, as a general principle, that juries are as stupid as they are sometimes believed to be. A lay-person jury can grasp probability theory, but it must be reduced to it’s smallest parts and carefully and clearly explained. As I already said, sometimes experts are not good at explaining theory because they presume a basis of knowledge of their field (“well, everyone knows this so I won’t have to explain it”) that in fact the jury does not have. As for working against the inherent biases of a jury – which each person will bring, based upon their own personal experiences, even if they think they are unbiased – that’s just how our system works. Again, rick-benefit analyses (or cost-benefit analyses) are a good example of this. Juries get RBAs; they just don’t like them. A company may legitimately perform an RBA prior to determining whether or not to fix a faulty automobile latch, for example, and based on the RBA decide not to fix it. But a jury may legitimately determine that an RBA – even if properly done * and even if the jury understands it* – does not justify failing to fix the latch. This the jury may reasonably do.
No one is entitled to a jury of technical wizards, or a jury of their professional peers – which would present its own problems, in that professional peers are less likely to be unbiased or to be comfortable passing judgment on one-another. All we can do is do our best to explain the case to a jury of lay-people and have faith that they will understand it, if it is explained carefully enough, and reach the correct result. Will they always reach the correct result? No. But my experience is that they will do so more often than not.
Okay, so when you have two conflicting ‘experts’, how is the jury qualified to decide which one is right? If a jury is reduced to trying to make a technical decision it doesn’t understand, based on the testimony of ‘experts’, doesn’t this place the power in the hands of the expert rather than the jury?
And this proves my point - if a jury doesn’t ‘like them’, it’s because juries don’t understand them. A risk-benefit analysis is part of good engineering. I’d argue that if Ford DIDN’T do one, it would be negligent. But juries don’t like them, because they don’t understand them. The fact that there is a major disconnect between accepted engineering practices and what juries ‘don’t like’ means there is a major problem.
The result is that companies that want to do proper engineering wind up having to act like criminals and burn their documents, lest they wind up in court with a ten billion dollar lawsuit against them because the non-engineer jury doesn’t like certain engineering practices.
Is ‘more often than not’ a particularly good standard to hold to the legal system? In criminal cases we demand evidence of guilt “Beyond a reasonable doubt”. In civil law it’s the preponderance of evidence, but would hope that juries are still correct in the vast majority of cases.
My reading of highly technical lawsuits would indicate to me that it’s almost a coin flip as to whether the jury will decide correctly or not. The jury is incompetant to weigh the actual evidence, so it has to resort to intangibles like which expert seemed more believable and distinguished, whether the poor widow was in the front row crying through the whole thing, etc.
No. You are assuming the jury does not understand the information put before it. I can only reiterate that my experience does not indicate this is often the case. If the jury is presented with testimony from two duelling experts, it must determine which was more credible and more persuasive, and why. I don’t mean who was slicker, I mean who made more sense in the context of the case. You appear to believe the juries must accept the testimony of experts as gospel, but they don’t; they may, can, and do evaluate testimony themselves and reject it if they don’t find it credible. That is why the question of the weight of the evidence is one for the jury.
I don’t believe this is correct. A jury may understand full well how something occurs without approving of it. Most juries are perfectly able to grasp the idea of weighing the cost to the company (in dollars) of fixing a defect versus the cost to the company (in dollars) of living with the consequences of leaving the defect in place – and knowing that, statistically, X number of people are likely to be injured or killed because of the defect. How is that analysis good engineering? Good engineering is fixing the defect. RBAs are good economics, but “good economics” is a weak defense when faced with a dead child, for example. You appear, at this point, to be arguing that because, in your opinion, RBAs are a legitimate way of allocating risk, the company should not bear the cost of assuming that risk in court. I don’t think this follows. Juries are perfectly free to see RBAs for precisely what they are – the valuing of human life in company dollars – and to dislike them for that reason.
Again, I disagree. A jury has the ability and the right to determine whether “acceptable engineering practices” are in fact acceptable in the context of society as a whole. Cloning may become acceptable in the context of engineering and/or medicine long before it becomes acceptable to society – and it is society at large you must deal with when dealing with a jury. As far as RBAs are concerned, juries tend to feel they are not acceptable – even if we consider RBAs to be a function of engineering instead of economics. To say that because someone doesn’t like something must mean they don’t understand it is the most dangerous form of patronization. And if there is one thing I’ve learned in my short career, it’s that you patronize a jury at your peril, and to your almost inevitable detriment.
If a company burns the documents tending to show why it followed a particular course of action, then they deserve what they get, and that includes million-dollar verdicts. People (and companies) who believe they are doing the right thing do not destroy the document trial that memorializes that.
Well, if you would like, change “more often than not” to “in the vast majority of cases.”
All I can say, again, is that this has not been my experience, and so I disagree with you. You have a striking lack of faith in the intelligence and abilities of your fellow members of society; suffice it to say that I do not share it. My opinion remains that the concerns you raise, while certainly having merit in some cases, do not justify overriding or overhauling the Sixth Amendment. And, for what it’s worth, the debate over the Sixth Amendment’s ratification raised just these sorts of issues – the concern that lay people (farmers and shopkeepers) would not be able to grasp the intricate issues raised in a court of law, which was the principal reason some framers did not want the right to trial by jury including in the Constitution. That, of course, was 220 years ago.
Wrong. Good engineering is minimizing risk while building a product that is useful, and can be sold at a profit.
A company that builds ladders KNOWS that for every 100,000 units they sell, X number of people will fall off of one and be seriously injured or killed. They also know of about a hundred ways in which that number could be reduced, but choose not to do it because it would make the ladder too expensive, too unwieldy, etc. (For example, every ladder could come with a safety harness. Every ladder could have a complex attach mechanism at the top for the 10% of people who use it to climb onto a roof with a Soffit. Every Ladder could have a counterweight system to make it more stable. Etc.) Ladder manufacturers have drawn a line at some point and said, “This extra small percentage of safety is simply not worth it,” and they are correct for having done so.
Good economics = good engineering, AND intelligent safety design, for the simple reason that there is a finite amount of resources that can be allocated for safety. Any time you have finite resources, you should do an analysis to determine where to use those resources to best advantage.
Let me give you an example: Side impact air bags appear to save lives. However, they are only available on a few cars. Why? Are all car companies negligent if they don’t use them? They aren’t universally used because they are expensive. If a car’s main selling point in its low sticker price, then putting a side-impact airbag in the car may kill it in the marketplace, and no one gets the benefit of the airbags ANYWAY. Of course, you can simply pass a law forcing all car manufacturers to use them (or use the legal system like a hammer, which is all too common). So all cars go up in price, and more people stick with their older cars, which are LESS safe than the new ones would have been without the airbags. The bottom line: The fatality rate goes up if you do not follow the recommendations of the Risk-benefit analysis, even if it tells you not to install those airbags.
Cars are the best example of this, because there’s hardly a single area that couldn’t be made better and safer, and often for very little expenditure. Fuel filler necks could be made out of heavier steel. Windshield wiper motors could be made to work faster. Taller tires can improve braking performance. Seatbelts could be made wider, or 3-point harnesses installed. The rollover structure can be made stronger. The door latch can be made to withstand more force.
EVERY ONE of these decisions involves a risk-benefit analysis. Each choice might only cost a couple of bucks and a few ounces of weight, but there are hundreds or thousands of those choices. If you did them all no one could afford the car and it would be a lumbering pig. So you do a risk-benefit analysis to try and figure out which ones are worth doing and which ones aren’t. But God help you if one of the rejected choices winds up in court.
It would appear that you don’t really understand them either. Engineering is all about making tradeoffs, including tradeoffs in matters of safety. There isn’t a single car on the road that’s as safe as it could be given unlimited resources, and there’s not a single car on the road that couldn’t be made somewhat safer by the expenditure of even a couple of dollars per vehicle. Risk-Benefit analysis is an attempt to allocate finite resources to where they do the most good, and this includes profitability (it HAS to - there’s no point in building the world’s safest car if no one will drive it).
And if you do a risk-benefit analyis properly, you will ALWAYS come to a point where you draw the line and say, “This much expenditure on safety, and not a penny more”. They are not an evidence of wrongdoing, they are evidence that a company considered safety factors in every area of the design and made the best decisions it could.
Now, a company *could be negligent if a Risk-Benefit analysis showed a serious defect and the company chose to ignore it. The problem with juries is that they see the RBA itself as prima facie (sp?) evidence of wrongdoing. They believe that all defects should be repaired if a company knows they exist, and only tend to let companies off the hook if they can show that the defect was unknown, and good engineering practices were used. Well, this just isn’t the real world. It’s a fantasy constructed by the media, helped along by plaintiff’s lawyers, and used on juries who lack the understanding of statistical methods to make a proper decision.
In this case, I can draw no other conclusion. If a Risk-Benefit analysis is a good and proper function of engineering science, then a jury must be ignorant of the facts if it doesn’t “like” them. And your arguments against them are exactly my point. The jury sees them as ‘trading dollars for lives’, and so do you. But every product on the market that carries any risk at all is a collection of tradeoffs between dollars and lives. EVERY one. And it has to be that way. The belief that products can be made perfectly safe is a denial of reality.
And if they can’t be made perfectly safe, the way to do the best you can is to consider and accept or reject features based on a Risk-Benefit analysis.
There’s a difference though, isn’t there Sam, in saying, “let’s not outfit this ladder with a harness because it’s too expensive, even though a harness could prevent foreseeable injuries,” and “let’s not recall this car because the potential damages in lawsuits are less than the cost of recall, even though the car has a pronounced tendency to catch fire when impacted from the rear”? There’s a degree of reasonableness involved that makes RBAs for ladders acceptable to most people and RBAs for cars with a life-threatening defect repugnant. I think that’s what jodih is saying–we can usually count on juries to exercise that degree of reason in distinguishing between the two.
