Feminist Philosophers

News feminist philosophers can use

O thanks, AAUP July 30, 2011

Filed under: academia — annejjacobson @ 7:43 pm

From the Chron of HE:

July 28, 2011, 5:13 pm

The American Association of University Professors is urging the Education Department to reconsider recommending as the standard of evidence “more likely than not” in campus sexual-harassment cases. To uphold academic freedom, the AAUP’s burden of proof is “clear and convincing evidence”; the newly recommended lower standard would erode due-process protections, Gregory F. Scholtz, associate secretary and director of the association’s Department of Academic Freedom, Tenure, and Governance, wrote to Russlyn H. Ali, the Education Department’s assistant secretary for civil rights. AAUP joins the Foundation for Individual Rights in Education, a free-speech group, in opposing the new guidelines. Other groups, including the American Association of University Women, have applauded them.

As the AAUP points out in the letter the first link in the quote goes to, their standards were established in the 1940′s. The Department of Education’s change would bring the standard of evidence in line with other civil rights laws.

 

58 Responses to “O thanks, AAUP”

  1. yuiop Says:

    Yeah, why should people accused of sexual harassment have the same rights as people accused of other offenses? Everyone knows that no one has ever lied about being sexually harassed.

  2. annejjacobson Says:

    Yulop, with the revised standards, the accused would have the same rights as defendants in other civil rights cases. This means that victims of sexual harassment have to meet the kind of standards required for, e.g., charges of discrimination, and no more.

  3. Jamie Dreier Says:

    I don’t understand the quoted material about “other civil rights laws.” Anne (or whoever does understand it), can you explain what it means?

    The sexual harassment standard of proof that the AAUP is talking about is a standard used on campus, at a campus hearing, say in a hearing to break a professor’s tenure. What does the DoE mean by “the standard of proof established for violations of the civil rights laws”? Is Ali (the Sec. of Ed.) talking about a standard used in a criminal trial? Or does she suppose that campus committees will be ruling on violations of civil rights laws?

    It seems to me that “more likely than not” is a terribly weak standard in this context. But I may be misunderstanding what the sides are talking about.

  4. yuiop Says:

    There are significant differences between the case of sexual harrassment and civil rights discrimination that make the proposed lowering of evidential standards, I think, outrageous.

    Take a case of someone who claims that she was not hired by a company because of her sex. If I am judging this case there are many factors I can consider. For example, did a similarly or less qualified male get the job instead of her? Is there a pattern of women being passed up for this kind of job? etc. So in this case if, after considering all these factors, it looks like the relevant factor was her sex, then I agree that the company should have to pay.

    But this is very different than a case of alleged sexual harrassment. Suppose there is a fresh-out-of-graduate-school assistant professor with no history of sexual harrassment or any kind of sexist behaviour, and a female undergraduate claims that he inappropriately touched her during his office hours, which he denies. Since, I imagine it is fairly rare that a woman makes purely fabricated claims of sexual harrasment, I contend that the mere fact that this woman has come forward claiming to have been sexually harrassed makes it more likely than not that she was in fact sexually harrassed. And in most such cases there is no other evidence to be found by which to judge one way or another – it’s his word against hers. But I think it’s clearly ridiculous to say that the woman’s claim alone should be evidence enough to conclude that the professor is guilty. That would mean that all it would take to essentially sink a person’s career is one disgruntled female undergraduate who claims to have been sexually harrassed.

  5. annejjacobson Says:

    Yulop, there are a number of factors determining how laws are interpreted and applied. To put it a bit bluntly, what you or I think, based on our assumption about false allegations, is just irrelevant. The law is supposed to work in the real world, which contains details we non-experts may not know about.

    Plaintiffs in civil rights cases typically have to meet a preponderance of evidence burden and not at beyond a reasonable doubt level. These cases include allegations of violations of voting rights, the rights of disabled people, rights to non- discrimination, etc. I think a central justification for the lower standard is this: the requirement of proof beyond a reasonable doubt is extremely hard to meet in such cases and having it is a deterrant to making a complaint. The result is that the laws fail to protect rights. That could not be more serious.

    This argument can be seen in the reasoning around Stanford’s embracing of the lower standards. The incidence of sexual harassment on campus tends to be quite high, with the result that many women regard their campuses as hostile environments where it is not safe to be alone. That is horrible.

    I don’t know for sure whether any violations of civil rights are criminal, as opposed to civil violations. I think the AAUP is talking about procedures within a university and I know the dept of ed is, so the question of civil v. Criminal doesn’t really arise in this issue.

    Jaime, I think taking a case to court is always taking it outside the university. In that case, everyone is talking about intra- university cases, not court cases.

  6. annejjacobson Says:

    For Stanford, see here:

    http://www.stanforddaily.com/2011/04/12/stanford-lowers-standard-of-proof-for-sexual-assault/

    Also, for the record, I would be hesitant to recommend anyone file a harassment complaint today AND even after a discussion with a vice-president and the head of the aao-eeoc, I decided not to on my own behalf. Mine all concerned a fairly recent and awful experience, one that demonstrated that women can be vulnerable long after their youthful days.

  7. yuiop Says:

    I’m not particularly concerned with the particular legal facts about how civil v. criminal cases, etc. are dealt with in this country. I was trying to make a moral argument which went something like this

    It would be unjust if the mere allegation of sexual harassment (without clear evidence) were enough to seriously harm a person.
    The effect of the proposed policy would be just that.
    Therefore, the policy would be unjust.

    I don’t dispute that it’s horrible to be sexually harassed. But it’s also horrible, say, to be the victim of homicide, but surely none of us would propose lowering the standard for a homicide conviction to “more likely than not.” In that case it would be *incredibly* difficult to defend yourself against a charge of homicide Likewise, the proposed policy would make it virtually impossible to defend yourself against a charge of sexual harassment.

  8. annejjacobson Says:

    But yulop, your argument depends on lots of factual assumptions. I have no idea why you think likelihood would be settled in the real world by one person’s testimony, but that’s not the way these things work – unlike your imaginary case that you started with. The argument is not about merely possible worlds.

  9. yuiop Says:

    Here’s what I was arguing, to be more clear:

    1. In many sexual harassment cases, we have no evidence by which to judge other than the testimony of the accuser versus that of the accused.

    2. It is relatively rare for someone to make false allegations of sexual harassment.

    3. Therefore, by (2), in cases where we only have two opposing testimony, it is rational to believe that it is more likely than not the the accuser is telling the truth.

    4. The proposed policy would make proof that the harassment is more likely than not to have occurred sufficient for discipline.

    5. Therefore, in cases where there is only two disputing testimonies, the new policy would make it such that the mere allegation would be sufficient to lead the accused to be “convicted”.

    6. But that is unjust (an allegation alone should not suffice to convict someone).

    7. Therefore, the policy is unjust.

  10. Jamie Dreier Says:

    The standard in all civil cases is Preponderance of the Evidence. That’s not special to civil rights law. Historically, the reason for the low standard is that in a civil suit the two parties are symmetrically placed; they are disputing ownership of a piece of land, say. Since they are symmetrically placed, the burden of proof must also be symmetrical. Preponderance of the Evidence is the only standard that induces a symmetrical burden. (That was a long-winded way of saying something fairly obvious.)

    Here’s what worries me: a hearing at a university, whether the complaint is sexual harassment or conflict of interest or whatever, is not symmetrical in the same way. It is a procedure for determining punitive action. If you are guilty, you are punished. If you are not guilty, the provost is not punished, the complainant is not punished, nobody is punished. It is therefore relevantly like a criminal and not a civil proceeding. The defendant is in jeopardy.

    On the question of justice, I tend to agree with yuiop. Think about what the Preponderance standard means. Just consider the defendants who will be convicted under the preponderance standard and not under a stricter standard. (Because if the question is whether to use Preponderance or a stricter standard, they are the only defendants who are relevant.) Almost half of these are innocent! Sexual harassment can be awful, but losing your tenured position because of a false accusation is also pretty awful. And I don’t mean just that it is an awful thing to happen; it is also an awful thing to do to someone, and when it is done the university will have done it. I guess what I’m saying is, rights are things it is important to protect, but even more important not to violate. At least, that’s my view about rights.

  11. annejjacobson Says:

    Jaime, your claim – Just consider the defendants who will be convicted under the preponderance standard and not under a stricter standard – looks like a prediction. I can’t see any reason for it.

    I honestly think that you are both considering possible worlds and are not showing anything about what happens in the actual world. There is a huge difference between the two approaches.

    In fact, in the actual world, women’s claims seem typically to be accorded low credence, and do not stand up against a man’s denial. That’s before the question of proof beyond a reasonable doubt is raised. She said/he said does NOT CUT IT in the actual world in her favor.

    I honestly think you all need some evidence for your claims to the contrary.

  12. yuiop Says:

    “I honestly think that you are both considering possible worlds and are not showing anything about what happens in the actual world. There is a huge difference between the two approaches.”

    I gave an argument that makes no appeal to possible worlds. If there’s a bad premise or inference, let me know what it is.

    “In fact, in the actual world, women’s claims seem typically to be accorded low credence, and do not stand up against a man’s denial. That’s before the question of proof beyond a reasonable doubt is raised. She said/he said does NOT CUT IT in the actual world in her favor.”

    First, I don’t believe this is true at all. I think people have a tendency to believe the accuser whether it’s a man or a woman. But even if you are right, that’s beside the point. I made a claim about what it is rational to believe and not what people in fact do believe. And so if the policy was applied correctly, it would have unjust consequences. Saying that a policy is acceptable because it probably wouldn’t be applied correctly doesn’t seem like a very compelling approach.

  13. Anonymous Says:

    It might be worth pointing out that “clear and convincing evidence” is a lower standard of evidence than “beyond a reasonable doubt” (though higher than “preponderance of evidence”). The AAUP is not asking that sexual harassment cases be proven beyond a reasonable doubt.

  14. annejjacobson Says:

    Yulop: you argument is invalid with a questionable or false premise. It is disputable whether it is rational to regard a charge as such to change the balance of evidence. Even if it were true, it does not follow that that will be how any case is in fact decided. I am not sure it even follows that if it were decided correctly, it would lead to unjust premises. My concern is that you have a hidden premise about what it is to decide a case correctly. If we put that premise in, then the argument may be valid, but that premise looks to be false.

    I don’t know if you know that “valid” is typically used as a technical term by philosophers. You might google “logically valid.”

    I think you have stated your view a number of times. I have tried to express my view of it several times in different ways. We need to stop repeating ourselves.

  15. yuiop Says:

    “It is disputable whether it is rational to regard a charge as such to change the balance of evidence”

    If 90% of all observed x’s are F, then it is rational to believe that the next x is more likely than not to be F. You might google “inductive logic”.

    “Even if it were true, it does not follow that that will be how any case is in fact decided.”

    Again, this is beside the point.

    “I am not sure it even follows that if it were decided correctly, it would lead to unjust premises.”

    Um, do you mean unjust consequences? Before I wasn’t talking about deciding the case correctly (if they decided the case correctly they wouldn’t be following this ridiculous policy), I was talking about applying the policy correctly. And to apply the policy correctly is to apply the policy rationally.

    “My concern is that you have a hidden premise about what it is to decide a case correctly.”

    Again, I wasn’t talking about deciding a case correctly, I was talking about applying a policy correctly. And yes, there is a hidden premise. Here it is: If S applies policy P correctly, S applies P rationally.

    “I don’t know if you know that “valid” is typically used as a technical term by philosophers. You might google “logically valid.””

    Um, okay. I never used that term, but yes, thanks for the valuable lesson. I’ll be sure to repeat it to students in my logic class.

    “I think you have stated your view a number of times. I have tried to express my view of it several times in different ways. We need to stop repeating ourselves.”

    I haven’t been repeating myself. I have been clarifying my position in hopes of getting a coherent response from you. But it has become increasingly clear that none is forthcoming.

  16. Bijan Parsia Says:

    @9, I don’t know if the fault is quite localizable to a premise, but in some presumptions made throughout.

    In particular, you make a presumption about how evidence is assessed in typical proceedings. Even under a “more likely than not” standard, there are typically formal guidelines controlling what counts as evidence and how it is assessed.

    Thus, for example, it is typical that the mere fact of a single accusation without any corroboration is not deemed sufficient to convict (a presumption of innocence still holds) under any standard of evidence. In other words, presumption of innocence adjusts the prior probabilities. Given the patterned nature of harrassment (i.e., typically there must be a pattern of conduct) as opposed to the incidental nature of, e.g., assault, it’s pretty hard on the more likely standard to convict someone.

    So, I tend to agree that you aren’t describing the actual world and the actual way these standards are interpreted and applied. You are using a standard of rationality which is not the only, in fact, applied. (In the sense that you are admitting evidence in ways that would not be admitted and neglecting key priors.)

  17. Jamie Dreier Says:

    Anne,

    I have said literally nothing about merely possible worlds, so I don’t understand that objection. I said nothing, either, about “he said/she said” cases, so I suppose there you might just be confusing me with yuiop. (By the way, you are regularly misspelling both of our names.) I did say some things about (actual) law, and some things about (actual) university hearings. Of course I might be wrong about those things.

    On my point about the innocence rate: I think you have misunderstood me. It is just a mathematical point.

    Look, by definition, the defendants in the class we’re talking about (not convictable under the strict standard, convictable under a laxer one) have a probability of guilt of just over ½. Therefore, in the long run, just under half of them will be innocent. To be more pedantic: their probability of guilt given the evidence of membership in this class is just over ½, so the expected fraction of them that are guilty, given that evidence, is just over ½. (This is just mathematics; the expected frequency is always the same as the probability.) So in the long run, just under half of them will be innocent. This is just the fact that the long term relative frequency of a property in a population approaches its probability. (Sorry if this last part was too pedantic; I am trying to strike a balance between too pedantic and insufficiently explicit, and I think last time I erred on the latter side.)

  18. Jender Says:

    I admit I have not read all of these comments, so please feel free to ignore. But from what I’ve seen so far it looks like you are all making good points but talking past each other:

    1. Jamie and Yulop are discussing what would happen if the standard were successfully applied– that is, cases were actually decided *in accord with* the standard. I think they’re right that this would lead to a lot of injustices, as they describe.

    2. Anne is discussing what would happen if the standard were put into use by people as they are now in our institutions as they are now. Her claim is that there are such serious barriers to successfully pressing a sexual harassment suit that *attempting* to use the standard would in fact result in something much closer to justice than we have now. I suspect she is right also, though I don’t actually know enough to be confident of this. But it is clear that there’s no conflict between Anne being right and Jamie and Yulop being right.

  19. annejjacobson Says:

    Jamie, yulop, I do apologize for mispelling your names. That was rude, but not intentionally so.

    My point about possility is that we need to judge the law in relation to how it works in the real world.

    Relately, if we look at those found culpable because the preponderance of admissible evidence was against them, I don’t get the move to the mathematical laws, since the sample of the accused may be biased in termes of those who are guilty, or it could also be that a much higher burden could be met, but juries tend to be biased in some ways.

  20. annejjacobson Says:

    Jender, interesting and thanks. I think my concerns are a bit more directed. E.g., yulop has asserted that an accusation can be enough to meet the burden of proof. I do not think that is so.

  21. annejjacobson Says:

    Let me add that i’m not sure of the terms in which Jamie is casting his argument. So there seems to me quite a difference between probabilities concern with burden of proof and probabilities of their having done it. Part of the gap is created by rules of evidence. A tape of a beating, for example, may show who did what, bit it may be inadmissible. Excukpatory evidence can meet a similar fate.

    In any case, apparently most universities and all civil cases use the weaker evidence, so one thing would be to see what the practice looks like

  22. Bijan Parsia Says:

    “Jamie and Yulop are discussing what would happen if the standard were successfully applied– that is, cases were actually decided *in accord with* the standard. ”

    Actually, I don’t think that’s the case. Neither have provided any evidence that the technical meaning of the standard supports their reasoning.

    Indeed, it’s easy to form a reductio, Yulop. Let’s use the 90% true allegation figure. Is there a reasonable doubt there? Is it “rational” to doubt the allegation? What if it were 95%?

    “More likely than not” is just “preponderance of evidence. Which is the norm in civil cases, which are not normally decided on accusation alone. So what’s different here?

    (I’m fine with the idea that some other standard of evidence should apply, but then there should be some real distinction invoked. It is really bizarre that the greater reliability of an accuser would induce the need for a higher standard of evidence! Similarly, the loses in lots of civil trials are comparable to losing tenure (e.g., restitution can bankrupt you).)

    Prima facie, these arguments are either risible, or made in complete ignorance of easily obtainable facts. (E.g., go read Wikipedia.) This is even before we consider how standards get applied in practice.

  23. Jamie Dreier Says:

    Jender, very good point, thanks. I guess we should expect some systematic deviations in practice from any standard that would be implemented.
    Personally, I don’t know of any reason to think that a systematic bias in a university hearing would work in one direction (favoring accused, say) rather than the other — does anyone know how this works?

    Anne, I’m confused about this:

    So there seems to me quite a difference between probabilities concern with burden of proof and probabilities of their having done it.

    The probability in a standard of proof is the probability of… what? I thought it was the probability of their having done it. So I can’t see a contrast here.

    Also,

    Part of the gap is created by rules of evidence.

    Which rules of evidence? (As you rightly reminded us, we aren’t talking about a court of law.)

    Bijan,

    In other words, presumption of innocence adjusts the prior probabilities.

    In a civil case? I just want to be clear about what you’re claiming.Your claim is that in a civil trial, the presumption of innocence serves to adjust the prior probability that the defendant is guilty?

    Actually, I don’t think that’s the case. Neither have provided any evidence that the technical meaning of the standard supports their reasoning.

    Not sure what it is you want evidence for. The technical meaning of Preponderance of the Evidence is not in dispute, is it? We all agree: it means more likely than not. You said this yourself. So please say what you meant.

    Hm, I see that right at the end you say that my argument is “risible” (I think — you could be saying that either my argument is risible or I should read Wikipedia). I wish you would be a little more specific than that. I don’t know how to respond to such a broad insult.

  24. yuiop Says:

    “Indeed, it’s easy to form a reductio, Yulop. Let’s use the 90% true allegation figure. Is there a reasonable doubt there? Is it “rational” to doubt the allegation? What if it were 95%?”

    Have you read *any* of the discussion? The whole point is that you DON’T have to prove it beyond a reasonable doubt!

    You obviously have no idea what you’re talking about and have nothing to contribute. Thanks.

  25. yuiop Says:

    Baj Pars

    And yes 10% is a reasonable doubt.

    “More likely than not” is just “preponderance of evidence. Which is the norm in civil cases, which are not normally decided on accusation alone. So what’s different here?

    The old standard was “clear and convincing evidence”. According to the new standard you don’t need clear evidence. You just need to prove that it is more likely than not which is not the standard for a civil case.

    I would highly recommend you actually read the main post before making moronic objections.

  26. Bijan Parsia Says:

    Jamie,

    First, I apologize for the tart remarks. They are clearly generating unuseful heat. Thanks for being restrained.

    In any case, I’m was more referring to yuiop’s contention that adopting the preponderance of evidence standard leads to accusation == conviction. It is evidently not the case that this standard leads to plaintiffs winning in all other civil cases. Now, unless you belief that sexual harassment allegations are uniquely veridical, then the reasoning yuiop put forth would kick in quite a lot. It doesn’t seem to. So what’s the difference?

    This is what I mean by “technical detail”. Evidently there are a slew of considerations that control the application of the standard in actual hearings, including what evidence gets considered and how it is interpreted.

    Cheers,
    Bijan.

  27. Bijan Parsia Says:

    @24,25
    First sorry for mangling “yuiop”…I was going from a quote.

    “Have you read *any* of the discussion? The whole point is that you DON’T have to prove it beyond a reasonable doubt!

    You obviously have no idea what you’re talking about and have nothing to contribute. Thanks.”

    I have read the discussion. My point is a reductio. If your argument works just as well against stronger standards, then there is clearly something wrong with it. I suspect that either you are not considering how evidence is actually admitted or how it is interpreted.

    “The old standard was “clear and convincing evidence”. According to the new standard you don’t need clear evidence. You just need to prove that it is more likely than not which is not the standard for a civil case.”

    http://en.wikipedia.org/wiki/Legal_burden_of_proof#Preponderance_of_the_evidence

    “Preponderance of the evidence, also known as balance of probabilities is the standard required in most civil cases. This is also the standard of proof used in Grand Jury indictment proceedings, (which however unlike civil proceedings, are procedurally unrebuttable), and in family court determinations. The standard is met if the proposition is more likely to be true than not true.”

    So, no. Clear and convincing is a stronger standard.

    I made one tart comment at the end of my post, which I hereby retract. I hope you can rise above it and continue more fruitfully.

  28. Bijan Parsia Says:

    So, presuming more likely than not *is* the standard in civil cases (if anyone would care to rebut that, let’s see it), then the question of whether it is reasonable to use it as a standard in internal sexual harassment disciplinary remains open. If that standard combined with the rest of the procedure led to people being fired on an arbitrary accusation, then I think we all agree that that would be wrong.

    If it only did so de jure because de facto the application was effectively using a much higher standard of evidence (for example), then I think there’d still be two problems: 1), that the standard was inappropriate and 2) that we’d want to know if an appropriate standard was de facto being applied or whether the accussors are being treated fairly.

    If the standard of evidence plus procedures let to a similarly reliable outcome as in a well run civil court, then I can think of two types of objection: a), the sort of accusation was inherently more prone to abuse (thus we need to apply a more stringent filter on accepting them), or b) the consequent of a successful prosecution was great enough to warrant letting more guilty people go so as to restrict the number of innocents punished.

    I presume the former is not the issue, at least for yuiop, since their argument relied on the high reliability of accussors in this case. The severity of consequence can be divided into, roughly, loss of job and loss of reputation (and thus, perhaps, future jobs, plus shunning, emotional distress).

    Note that it is very unclear that the typical penalty for a first time offender is that severe for a wide class of actions. If the actual consequence is a note in your file, a warning, and some sensitivity training, well, that isn’t the end of your life or career.

  29. Jender Says:

    There are a lot of violations of our Be Nice rule. Please start following it. Comments which don’t will be deleted.

  30. Jamie Dreier Says:

    I think it would be counterproductive for me to step in the middle of the dispute between yuiop and others.

    But in jender’s irenic spirit, here are a couple of relevant thoughts.

    First, it is very difficult to give a good example of a case in which it is clear what the probability of guilt is in light of the evidence, while keeping to at least moderately realistic examples (so as not to run afoul of the Merely Possible Worlds worry). Judy Thomson has a few splendid examples in her two papers on “statistical evidence”, actually, although one has to have some tolerance for idealization. So an example in which we just stipulate that the probability (witnessed by the relative frequency) of guilt given accusation is such-and-such has to be taken as illustrative, rather than dispositive.

    Second, Preponderance is the only standard that is quantifiable. We know the other standards are much higher, but we have no idea how much higher (or even how to go about answering the question how much higher).

  31. Bijan Parsia Says:

    Jamie,

    Just out of curiosity, do you agree that the strength of the standard is not the only factor involved in assessing whether it is appropriate? E.g., how and what evidence is admitted, standard rules of interpretation, rebuttal opportunities, possible consequences, etc.?

  32. annejjacobson Says:

    I’ve just finally been able to see the difference between yulop and yuiop. I again apologize, but I think it is more the ipad’s fault.

    I probably can’t engage any more today, since I am expecting agroup of philoaophers to arrive for tea. I will hope it is resolved when I get back to it. You will have gotten that my underlying concern is about whether we can decide the issue in advance of having more empirical data. This stems from a general worry that philosophers may (sometimes) think empirically comex isses can be decided by reason and common sense. For example, one relevant question concerns what happens to a complaint that is simply not answered. Does the defendent become probably at fault? If not, then a complaint by itself does not meet the burden of proof. And so on.

  33. Jamie Dreier Says:

    Bijan, certainly those other things are important! I’m not sure I understood the question. (How would the question of whether, e.g., hearsay evidence is admissible, be germane to the question of the appropriateness of the standard of proof?)

  34. Bijan Parsia Says:

    Jamie,

    I’m going back to your original concern:

    “It seems to me that “more likely than not” is a terribly weak standard in this context. But I may be misunderstanding what the sides are talking about.”

    What makes it a terribly weak standard? For yuiop, it seems to be because it allows “accusation as proof”. But that would only hold if accusation was per se admissible. If you have a fairly stringent pre-filter, then your standard is not the sole determinant.

    “On the question of justice, I tend to agree with yuiop. Think about what the Preponderance standard means. Just consider the defendants who will be convicted under the preponderance standard and not under a stricter standard. (Because if the question is whether to use Preponderance or a stricter standard, they are the only defendants who are relevant.) Almost half of these are innocent!”

    First, this isn’t quite right, yes? Just because we only require a preponderance standard doesn’t mean that most cases will be met with just that level. Furthermore, just because the mobilized permissible evidence supports only a 50% confidence doesn’t mean that the *outcome* (or the all things considered evidence) will be that weak.

    “Sexual harassment can be awful, but losing your tenured position because of a false accusation is also pretty awful. And I don’t mean just that it is an awful thing to happen; it is also an awful thing to do to someone, and when it is done the university will have done it. I guess what I’m saying is, rights are things it is important to protect, but even more important not to violate. At least, that’s my view about rights.”

    This presumes that from “convicted” we get to (a fairly maximal) sentence. But that’s not clear. As I wrote a bit earlier, if the consequence for a first offense under preponderance of evidence is a note in a file, plus a warning, this is annoying (and if applied to an innocent, wrong! clearly!), but it’s not so very dire. Again, comparable harms can happen in a civil trial, right? Are you against preponderance of evidence there as well?

  35. Jamie Dreier Says:

    Bisan,

    1.
    First, this isn’t quite right, yes? Just because we only require a preponderance standard doesn’t mean that most cases will be met with just that level.

    Right, that itself doesn’t imply that most cases will be met with just that level. But are you forgetting the sentence that you just quoted? We are only considering the defendants who will be convicted under the preponderance standard, and would not be convicted under a stricter standard.

    2.
    Furthermore, just because the mobilized permissible evidence supports only a 50% confidence doesn’t mean that the *outcome* (or the all things considered evidence) will be that weak.

    Sorry, I’m afraid I don’t understand what you mean.

    3.
    This presumes that from “convicted” we get to (a fairly maximal) sentence. But that’s not clear.

    I don’t think it presumes any such thing. Could you explain how you infer such a presumption?

    4.
    As I wrote a bit earlier, if the consequence for a first offense under preponderance of evidence is a note in a file, plus a warning, this is annoying (and if applied to an innocent, wrong! clearly!), but it’s not so very dire. Again, comparable harms can happen in a civil trial, right? Are you against preponderance of evidence there as well?

    No, I am not against preponderance in a civil trial. But I’m pretty sure I explained why (here, if the anchor tag works in these comments, which I doubt). I’ll try again.

    In a civil trial (at least in many typical ones) the parties are symmetrically positioned. If the court finds for the plaintiff incorrectly, the defendant is deprived of something that is properly hers. But if the court finds for the defendant incorrectly, the plaintiff is deprived of something (often the same thing) that is properly his. There is no way, therefore, to arrange things so that we err on the side of caution, which is on the side of not violating the rights of the parties.

  36. Bijan Parsia Says:

    “Bisan,”
    “Bijan”! If I mispelt your name too, let me know directly :)

    Re: 1.
    “Right, that itself doesn’t imply that most cases will be met with just that level. But are you forgetting the sentence that you just quoted? We are only considering the defendants who will be convicted under the preponderance standard, and would not be convicted under a stricter standard.”

    I wasn’t forgetting. I presume that if we were to assign a probability, the stricter standards would be something like 90, 95, and 99%. So there’s a huge gap.

    Re:2 (does blockquote work in these comments?!)

    “”Furthermore, just because the mobilized permissible evidence supports only a 50% confidence doesn’t mean that the *outcome* (or the all things considered evidence) will be that weak.”

    Sorry, I’m afraid I don’t understand what you mean.””

    Consider criminal cases. To indite you use a lower standard (preponderance) and then a higher standard (reasonable doubt). You can reverse that and use a higher standard then a lower standard. Similarly, you can assign certain weights to certain evidence, e.g., fact of accusation counts for at most 20%.

    3.
    “”This presumes that from “convicted” we get to (a fairly maximal) sentence. But that’s not clear.”

    I don’t think it presumes any such thing. Could you explain how you infer such a presumption?”

    In general, the greater the possible harm due to punishment, the greater the required standard of evidence, yes? So, again, if the worst that a first time defendant gets is a caution, a note in a sealed file, and cannot grade the student’s work, then (to me, at least) a preponderance standard seems quite reasonable. If the consequence is immediate termination for cause, then much less so. You yourself mobilized loss of tenure, which I agree is fairly serious (but is is less serious that being required to pay out your life savings?)

    4.
    “In a civil trial (at least in many typical ones) the parties are symmetrically positioned. If the court finds for the plaintiff incorrectly, the defendant is deprived of something that is properly hers. But if the court finds for the defendant incorrectly, the plaintiff is deprived of something (often the same thing) that is properly his. There is no way, therefore, to arrange things so that we err on the side of caution, which is on the side of not violating the rights of the parties.”

    Civil trials often involve punitive damages. How are they symmetric? Indeed, why wouldn’t the professor losing their job just be a case of punitive damages?

  37. Jamie Dreier Says:

    Bijan (sorry about the earlier typo!),

    1.
    I wasn’t forgetting. I presume that if we were to assign a probability, the stricter standards would be something like 90, 95, and 99%. So there’s a huge gap.

    I presume that we could set the stricter standard pretty much anywhere we wanted. Of course, it’s a bit silly to pretend that we are setting a precise probability number, but insofar as probability numbers represent what we would really be doing, we are not limited to particular alternatives like .5, .9, .95. In my opinion, if we want to know whether a Preponderance standard is too low, we should look at the marginal case, so to speak. Pretend that we are choosing between a standard of .5 and a standard of .51. That’s the interesting test.

    Re:2 (does blockquote work in these comments?!)

    Sadly, no, but anchor apparently does.

    Consider criminal cases. To indite you use a lower standard (preponderance) and then a higher standard (reasonable doubt). You can reverse that and use a higher standard then a lower standard. Similarly, you can assign certain weights to certain evidence, e.g., fact of accusation counts for at most 20%.

    I still don’t follow. What are these points supposed to be arguments for? Maybe I already agree with it.

    3.
    In general, the greater the possible harm due to punishment, the greater the required standard of evidence, yes?

    Well, not in general, no! A judge doesn’t instruct the jury to lower their standard of proof because the prosecutor is only looking for a conviction on a charge with a short sentence. But anyway, I don’t think anything I said implied a presumption about this one way or the other.

    So, again, if the worst that a first time defendant gets is a caution, a note in a sealed file, and cannot grade the student’s work, then (to me, at least) a preponderance standard seems quite reasonable. If the consequence is immediate termination for cause, then much less so. You yourself mobilized loss of tenure, which I agree is fairly serious (but is is less serious that being required to pay out your life savings?)

    So, some of the injustices will be much worse than others – I certainly agree with that. But what I sqwanted you to explain is in what way anything I said presumes otherwise. (I do not agree that Preponderance is reasonable whenever the punishment is small, by the way, but I do think this is a separate issue.)

    4.
    Civil trials often involve punitive damages. How are they symmetric? Indeed, why wouldn’t the professor losing their job just be a case of punitive damages?

    Hm, it wouldn’t be ‘damages’ of any kind, would it? But this just seems like a semantic issue, which is not very interesting.
    When the complainant and defendant are symmetrically positioned, so that errors in favor of one violate the rights of the other, then I see no coherent alternative to Preponderance. (If you disagree, say so.)
    In a hearing to dismiss or otherwise discipline a professor, the parties are not so placed. The administration violates the rights of the professor if it fires him for something she didn’t do. It violates nobody’s rights if it fails to fire him for something he did do. (Do you agree with this?)

  38. Jamie Dreier Says:

    I left an italics tag open, sorry. If this doesn’t fix it, can one of your FBers do it?

  39. Jamie,

    I’d just like to hop in with a question about a point you raised:

    “Look, by definition, the defendants in the class we’re talking about (not convictable under the strict standard, convictable under a laxer one) have a probability of guilt of just over ½. Therefore, in the long run, just under half of them will be innocent. To be more pedantic: their probability of guilt given the evidence of membership in this class is just over ½, so the expected fraction of them that are guilty, given that evidence, is just over ½. (This is just mathematics; the expected frequency is always the same as the probability.) So in the long run, just under half of them will be innocent. This is just the fact that the long term relative frequency of a property in a population approaches its probability.”

    It’s not clear to me that “the defendants in the class we’re talking about (not convictable under the strict standard, convictable under a laxer one) have a probability of guilt of just over ½.” The only conclusion we can make is that the expected frequency of innocence will be less than .5 and greater than .1. In order to say any more, we must have some idea of the _distribution_ of guilt, given the evidence. For example, suppose a uniform prior and totally non-informative evidence, i.e. P(G|E) = P(G). You can show that the probability that an arbitrary accused is guilty, given 0.5 < P(G) <0.9, is a healthy 0.7. Depending on what you believe the prior probability of guilt looks like, and how informative the evidence can be, you can concoct examples for which the probability of guilt for defendants at the margin is arbitrarily close to the upper bound.

    In general, a random real variable X with support [a,b] need only satisfy E(X) \in [a,b].

    Cheers,
    Patrick.

  40. Bijan Parsia Says:

    I yield the probability discussion to Patrick, who’s being way clearer than I am.

    I also will focus just on 4 (the symmetry argument), since that seems key. Since we both agree that the preponderance standard is sometimes reasonable (e.g., in extent civil cases), I trust that we agree that to be inappropriate for campas hearings that there has to be some significant difference between the two.

    Jamie, your proposed difference is the symmetry of right violation. I’ll note that you phrase it as a sufficient condition: “When the complainant and defendant are symmetrically positioned, so that errors in favor of one violate the rights of the other, then I see no coherent alternative to Preponderance.” So, *if* there is symmetrical rights violation, then preponderance (for lack of alternative). This doesn’t establish that preponderance is inappropriate in other cases, but I hope its ok to presume that you would argue that since right violations are serious, we ought to take significant efforts to avoid them.

    I’ll reiterate my attack on your rationale for preponderance in the civil case. I think we have to be a bit careful because it’s not hard to cast these cases as symmetrical between either the university and the professor or the student and the university. E.g., if the student’s contract with the university includes a harassment free environment, or the employment contract between the professor and the university includes harassment free behavior, then there is a contract violation if there was sexual harassment. Both parties are entitled to their contracted provisions. So, symmetry? (I presume the only possible consequences are essentially contractual or reputational. The university cannot impose fines or jail time.) (Note that I don’t hold that symmetry requires preponderance.)

    However, I think that the fact of punitive damages is sufficient to show that (pure) symmetry does not exist in the civil case. (I’ll use the definitions from http://en.wikipedia.org/wiki/Damages just so that we have specific meanings. So, “damages is an award of money to be paid to a person as compensation for loss or injury” and “Generally, punitive damages…are not awarded in order to compensate the plaintiff, but in order to reform or deter the defendant and similar persons from pursuing a course of action such as that which damaged the plaintiff.”

    So, do you agree that 1) wrongly awarded punitive damages violate the plaintiffs rights (by wrongful deprivation of property)? and 2) that punitive damages asymmetrically affect the plaintiff?

    So, I don’t see that symmetry in rights violating outcomes distinguishes civil cases from all others so to force preponderance there. (Note an out: Awarding of punitive damages is generally held to be fairly restrictive, e.g., bounded by the 5th and 14th amendment in the US (see http://en.wikipedia.org/wiki/Punitive_damages#United_States ). But I don’t yet see that a higher standard of *evidence* is required.)

    (None of this establishes that a preponderance standard is appropriate here! It just attempts to show that it is not prima facie unjust.)

  41. Jamie Dreier Says:

    Hi Patrick,

    Yes, quite right.
    Sorry, I didn’t make this clear enough. I did say this:

    In my opinion, if we want to know whether a Preponderance standard is too low, we should look at the marginal case, so to speak. Pretend that we are choosing between a standard of .5 and a standard of .51.

    On the other hand, if we are choosing between Preponderance and some much higher standard and have no option of picking a standard in between, then as you say we won’t be able to know specifically what the rate of conviction of extra innocent defendants will be without knowing something about the distribution of cases over the various thresholds. (Is that clear?)

    Bijan,

    E.g., if the student’s contract with the university includes a harassment free environment, or the employment contract between the professor and the university includes harassment free behavior, then there is a contract violation if there was sexual harassment.

    Just looking at the latter case: that’s true, but a university failing to dismiss a professor even though (unbeknownst to the university) the professor has violated her contract is, obviously, not a violation of anyone’s rights. The former case is trickier, but it strikes me as unrealistic as a model of the legal relation between a student and a university.
    Try this as a comparison. The state is obligated to provide you with protection against, say, robbery, to the extent that it can do so. Still, if it adopts a Reasonable Doubt standard in its criminal trials, knowing full well that in doing so it will sometimes fail to convict robbers, it does not thereby violate your rights.

    So, do you agree that 1) wrongly awarded punitive damages violate the plaintiffs rights (by wrongful deprivation of property)? and 2) that punitive damages asymmetrically affect the plaintiff?

    Yes, assuming that by ‘plaintiff’ you meant ‘defendant’, in both occurrences.
    I think punitive damages are very dubious, actually. When a defendant is in jeopardy of punishment by the state, then I think morally speaking the defendant is entitled to the protections to which he would be entitled in a criminal proceeding.

    I must be looking at the dialectic differently from the way you’re looking at it. As far as I’m concerned, it’s quite possible (indeed it’s nearly certain!) that various aspects of trials are unjustified and unjustifiable. So if it turns out that there is no good justification for using Preponderance as the standard of proof in some civil trials, that will not surprise me at all. As I’ve said, the symmetry consideration does seem to me to be an excellent justification. In civil cases in which there is no symmetry (because of jeopardy of punitive action by the state), I’m perfectly willing to believe that there is no good justification at all.
    You seem to be taking it for granted that using Preponderance in civil trials is justified. If that’s right, I would like to know what the justification is.

  42. Bijan Parsia Says:

    Yes, please replace “plaintiff” with “defendant”. Brain fart!

    “Just looking at the latter case: that’s true, but a university failing to dismiss a professor even though (unbeknownst to the university) the professor has violated her contract is, obviously, not a violation of anyone’s rights.”

    Why isn’t that symmetrical? I.e., what right is being violated if the university violates their contract with the professor? I.e., is the harm a rights violation or a contract violation (at least, assuming there’s no other problem going on). I’m not saying its impossible for their to be asymmetry (e.g., I think its plausible that states don’t have *rights* vis-a-vis its citizens, but surely it can violate citizen rights; similarly, some employment practices rise to rights violation, e.g., discrimination; but lots of employment practices which lead to termination, for example, do not).

    “You seem to be taking it for granted that using Preponderance in civil trials is justified. If that’s right, I would like to know what the justification is.”

    Ok, part of the problem is that I’m still half-arguing against yuiop, who’s arguments relied heavily on this standard being somehow much less than in civil trials. The other half is that I do think preponderance in civil trials is at least prima facie justified. At least, I’m suspicious of differences in standards for (civil, or civil-like) sexual harassment cases and general civil trials.

    Now, you’ve articulated a principle (symmetry) that is (prima facie) non-arbitrary and seems to make a distinction. And you’re willing to embrace the consequence that punitive damages are problematic. Great! Your position is a tough nut to crack!

    My initial justification (i.e., what I would have said before this discussion) was the basic legal distinction between the relative nature of punishment in civil vs. criminal (in the latter, we have the possibility of the most serious forms of punishment, loss of life or liberty, and these require special protection) and the status of the plaintiff (the state vs. another individual). In many ordinary disputes, resolution by some balance of probabilities is common (I believe). Whether 50% probability is the exact right threshold is a bit tricky and, in general, I, at least, determine whether I should enact a stricter standard based on my perception of good will and the general reliability of outcome. Which leads to my other rationale: In practice, notwithstanding the tort reform movement, it seems that civil trials work fairly way. Indeed, with somewhat less overall abuse than, say, Texas capital trials. (This is an impression, of course.)

    However, if I were to use your symmetry principle, I would hold that a student (in particular) is owed by their professor the general respect for dignity as well as a special duty of care. Furthermore, other students have the rightful expectation that their dignity and care are respected. Sexual harassment violates both dignity and care, and restitution may range from an apology, to sensitivity training (i.e., an acknowledgement that the professor’s failure requires correction), to dismissal (i.e., an acknowledgement that the failure was egregious, malicious, esp. serious, or systematic). The lack of such acknowledgement is a further infringement on the student by the university.

    I would expect restraint to be exercised in the “sentencing” phase, per usual. So, a first time, minor offender convicted on a close to 50% level of evidence should receive a minor sentence. A 4 or 5 time serious offender with a 75% level of evidence would justify greater sanction.

  43. annejjacobson Says:

    I wonder if, when we consider whether a burden of proof is too high or too low, we are treating laws and legal recedents as different from what they are. For example, in some countries, such as the United States, areas of the law reflect the idea that it is better to protect all the innocent than to convict all the guilty. But we also have strict liability laws, when it is possible that one is held guilty for an action when no reasonable person could have foreseen was wrong, I think. The state’s interests can take precedent over what we might see as individual rights. Another kind of example, one’s children can be taken away on the preponderance of evidence, which has notoriously caused great pain to some innocent parents.

    Looking at what the dept of ed is saying, something like that seems to be in the background. That is, the state has a very strong obligation to provide women with equal educational opportunities. In fact, the recommendation is that the university intervene before culpability is established, and prevent routine contact. One interpretation of this is that the rights of the accuser are given precedence over some of the accused’s. If that’s going on, the idea that some innocent people will be found culpaple in not relevant in a legal assessment. Lots of laws may be like that.

    It is interesting that the AAUP did not, according to what I read, appeal to bad consequences. Rather, they appealed to the impotance of the rights. Actually, I may be misremembering …

  44. Jamie Dreier Says:

    Bijan,

    I’m a little lost in this first part, and inclined to just drop it, but I will give it one more shot.
    Me:
    Just looking at the latter case: that’s true, but a university failing to dismiss a professor even though (unbeknownst to the university) the professor has violated her contract is, obviously, not a violation of anyone’s rights.

    You:
    Why isn’t that symmetrical? I.e., what right is being violated if the university violates their contract with the professor? I.e., is the harm a rights violation or a contract violation (at least, assuming there’s no other problem going on).

    I think each party has a right that the contract be upheld. (Did you mean to dispute this?) This means that if the university violates the contract, it violates the professor’s rights. This is pretty intuitive, isn’t it? (You might want to say that contractual rights are different from other ones, which may be right — I can’t tell whether that’s your point. I’d say the same thing about strict liability.)

    Next:
    Ok, part of the problem is that I’m still half-arguing against yuiop, who’s arguments relied heavily on this standard being somehow much less than in civil trials.

    I think I missed that bit – where did yuiop say that?

    The other half is that I do think preponderance in civil trials is at least prima facie justified. At least, I’m suspicious of differences in standards for (civil, or civil-like) sexual harassment cases and general civil trials.

    Hmmmmm. I guess that’s a significant point of disagreement. A university charging and ‘trying’ someone for sexual harassment seems significantly different from the paradigmatic civil trial (say, a dispute over the ownership of some property). So I would expect there to be some important differences in the standards and procedures.
    I don’t think the difference in the seriousness of punishment between criminal and civil trials will justify an enormous difference in the standard of proof. After all, the punishment for a criminal could be very light – a fine, probation, a suspended sentence… And in a civil trial the consequences for a defendant (not a punishment, necessarily) could be enormous: she could lose custody of her child, for example. (But, Anne, that isn’t a punishment, and in such a case there is the ever-important symmetry, since if the court errs on the other side someone else’s rights are violated.)

    Which leads to my other rationale: In practice, notwithstanding the tort reform movement, it seems that civil trials work fairly way. Indeed, with somewhat less overall abuse than, say, Texas capital trials.

    But do you think that Texas capital trials would work better if the standard of proof were lowered? (That’s a rhetorical question!) So that’s a pretty dubious rationale.

    But this last part seems to be the key:
    However, if I were to use your symmetry principle, I would hold that a student (in particular) is owed by their professor the general respect for dignity as well as a special duty of care. Furthermore, other students have the rightful expectation that their dignity and care are respected. Sexual harassment violates both dignity and care, and restitution may range from an apology, to sensitivity training (i.e., an acknowledgement that the professor’s failure requires correction), to dismissal (i.e., an acknowledgement that the failure was egregious, malicious, esp. serious, or systematic). The lack of such acknowledgement is a further infringement on the student by the university.

    Everything up to the last sentence is plainly right, inarguably. I don’t agree with the last sentence.
    Above I offered this analogy:

    The state is obligated to provide you with protection against, say, robbery, to the extent that it can do so. Still, if it adopts a Reasonable Doubt standard in its criminal trials, knowing full well that in doing so it will sometimes fail to convict robbers, it does not thereby violate your rights.

    Did you not find that persuasive? Put it this way. If a citizen’s general right that the state (analogously, university) protect him makes every criminal trial symmetric with respect to rights, then the symmetry consideration is useless. I don’t think the antecedent is true, though, so I can still deny the consequent. How about you?

  45. Jamie Dreier Says:

    Argh!
    I wish there were a Review window for these comments.

    I stuck this sentence on as an afterthought:

    “I’d say the same thing about strict liability.”

    and put it in the wrong place. I meant to put it at the end of the “Hmmmm” paragraph, where I have a little response to Anne.

  46. annejjacobson Says:

    Jamie, WordPress definitely doesn’t do comment review; they refuse even to recognize the phrase and claim that no one have ever search “help” with it.

    Suppose we have a case where the state removes children from a couples’ home because there is a predonderance of evidence of mistreatment. What’s the symmetry in those sorts of cases. I haven’t followed the symmetry argument very well; were you claiming it holds in crimilnal cases, since this case looks like lots of those.

  47. Bijan Parsia Says:

    Jamie,

    Yes, this comments section is difficult for me to work :) I’ll adopt your quoting style, though.

    You:
    I think each party has a right that the contract be upheld. (Did you mean to dispute this?)

    I didn’t mean do nor did I (I don’t think). I read you as saying that there was no symmetrical right violation in the dismissal case, and I was trying to cast it as symmetrical.

    Or by symmetrical do you mean that the very same right is at stake? I.e., in the case of property, for me to own it is for you not to? But, that’s quite different from compensatory damages, in general. It’s not that when you negligently cut off my hand that the very same thing is restored (it can’t be!), or that you lose your hand too. I presume that only the possibility of rights violation (or other harm) exists on both sides, or perhaps, of like harms. Is this wrong?

    You:
    This means that if the university violates the contract, it violates the professor’s rights. This is pretty intuitive, isn’t it? (You might want to say that contractual rights are different from other ones, which may be right — I can’t tell whether that’s your point. I’d say the same thing about strict liability.)

    I agree with that, but then if the professor violates the contract, they violate the universities (at least contractual) right? So symmetry?

    You:
    I think I missed that bit – where did yuiop say that?

    yuiop:
    According to the new standard you don’t need clear evidence. You just need to prove that it is more likely than not which is not the standard for a civil case.

    You:
    Hmmmmm. I guess that’s a significant point of disagreement. A university charging and ‘trying’ someone for sexual harassment seems significantly different from the paradigmatic civil trial (say, a dispute over the ownership of some property)

    But torts are paradigmatic civil trials! Indeed, so are sexual harassment and discrimination cases. Paula Jones’ case against Clinton was a civil case.

    Now, of course, you are in a good position to consistently reject preponderance for torts (and moreso when there’s a risk of punitive damages), but that wouldn’t be that the cases are different. They seem largely the same.

    You:
    But do you think that Texas capital trials would work better if the standard of proof were lowered? (That’s a rhetorical question!) So that’s a pretty dubious rationale.

    My point was just that prima facie, preponderance as a standard of evidence in civil trials does not lead to wholesale abuse. It may be due to other compensating factors (like admissibility requirements; rebuttal; certain interpretative presumptions, etc.) such that it’s problematic to use preponderance *without* the remaining apparatus. Or it may be that the standard is not applied (I find that dubious). Or it could be that the balance of factors (e.g., empirical sum of total harms) is less with this standard (e.g., because most plaintiffs are not frivolous or false; false or frivolous plaintiffs are fairly easily detected; imposing a tighter standard on plaintiffs would make going to court prohibitive and thus make civil law completely ineffective) works out right.

    Obviously, this is just a prima facie rationale.

    And now, the key bit:

    Me:
    The lack of such acknowledgement is a further infringement on the student by the university.

    You:
    Everything up to the last sentence is plainly right, inarguably. I don’t agree with the last sentence.
    Above I offered this analogy:

    The state is obligated to provide you with protection against, say, robbery, to the extent that it can do so. Still, if it adopts a Reasonable Doubt standard in its criminal trials, knowing full well that in doing so it will sometimes fail to convict robbers, it does not thereby violate your rights.

    Did you not find that persuasive?

    It’s appealing, but no. Standard legal theory holds that the reason my right is not violated by the state by a failure to convict is that a crime is a violation of a duty toward the state, not to me. It’s primary duty to me, in this case, is to protect me (not to convict). If the protection is negligent (e.g., the police don’t respond to calls in my mostly black neighborhood) then the state may well have violated my (civil) rights, even if it successfully convicts.

    Since the state is both the plaintiff and the judge (!) and the penalties are in principle most grave (e.g., lost of life or liberty) and the resources of the state are vast in comparison to the individual, holding the state to a high standard is easily justified.

    In the case that the robber robbed me, they have violated my rights and if that violation is uncompensated for, then the violation remains. A civil trial *may* violate the defendants right (if they are innocent) but if it fails then my right remains violated (by the guilty party). Furthermore, in many cases (not always!) our resources are comparable (certainly in comparison with the state). Where resources are unequal, things get trickier.

    Why doesn’t this meet symmetry in your sense?

  48. Bijan Parsia Says:

    BTW, sorry to take over the comments section. Please let me know if it’s too much and I’ll go offline. I appreciate the hospitality.

  49. annejjacobson Says:

    Bijan, you are most welcome here. Please continue for as long as you wish.

  50. Gideon Says:

    Two small points about punitive damages, if that’s still relevant to the discussion: First, punitive damages are very rare in civil cases. They are awarded in something 1% of medical malpractice cases, 2-3% of negligence cases and so on. More importantly, many jurisdictions require clear and compelling evidence of malice or egregious misconduct — and not just preponderance of the evidence — if punitive damages are to be awarded.

    I should also say that having served on university committees where quasi-juridical decisions of this sort are made, there is a *world* of difference between a preponderance of the evidence standard and a clear and compelling evidence standard. I have seen many cases that would have gone against the faculty member on a preponderance of the evidence standard, but which went in her favor given the more demanding standard. (These were not sexual harassment cases.) So if the question is whether the lower evidential standard would make a difference in practice, the answer to my mind is: without a doubt.

    I agree with Jamie’s case for retaining the more demanding standard, but let me add some further reasons. First, the committees that make these decisions are staffed by amateurs. Committee members are not trained in gathering and evaluating information, and they typically rotate off the committee before they have had time to accumulate relevant practical wisdom. Second, faculty members accused of infractions are not represented by attorneys. The rules of evidence, and in many cases the procedural rules, are somewhat unclear. For these reasons and others, it is reasonable to think that these are highly unreliable procedures — significantly less reliable than real legal procedures with professionals involves on both sides. Moreover the sanctions involved are (in my experience) highly non-trivial. If an untenured member of the faculty is disciplined for an infraction, that has predictable and very serious consequences at tenure time. When the infraction is serious — misuse of university funds, for example —  the penalty for a ‘first offense’ can be severe: a year of separation from the university w/o pay, for example. Allowing a crew of amateurs to make consequential decisions of this sort on the basis of the preponderance of evidence is a recipe for gross injustice, in my opinion.

  51. Jamie Dreier Says:

    The issues that Gideon raises seem much more interesting (and less stale) than the ones we’ve been talking about, so let me add a couple if remarks on those issues.
    I’ve also been involved in the rules and procedures for cases like these, though now that I’ve thought it over I guess I really can’t say much of anything about any cases, damn. (I will say that I wasn’t a party or a witness.) But at Brown, too, there are no clear rules of evidence, no right to be represented by an attorney, no right to be confronted by accuser or to know the evidence against you. I spoke to the university counsel’s office at great length about some of these things. The counsel’s position was that it is important not to turn the hearings into legal proceedings, so she/they were against installing rules of evidence, for example.
    At Brown, there is a fact-finding office for cases of sexual harassment, staffed by professionals. There is nothing similar for any other punishable offense. A defendant could hire lawyers to help investigate, but could not be represented at a hearing.

    Let me add one more thing. Bijan, you said that something is dubious, but I could not quite tell whether what you found dubious was that Preponderance is actually used in practice in civil trials, or that it is not. I am quite sure that it is not typically use in practice. It is the standard that is articulated, but it is not followed. Some interesting and problematic examples are discussed in Tribe’s famous “Trial By Mathematics” paper.

  52. annejjacobson Says:

    Gideon, thanks for the interesting argument and references to experience. Unfortunately, you are also reminding me of further elements in such proceedings that might make one wonder whether such committees should have much power, and whether the upper administrators should be judging appeals, as they often do. But to restrict my attention to your argument: what does one say in the case of harassment? Clear and convincing is often extremely difficult there, perhaps especially when the accused is a colleague of the committee members. But if the accuser does not prevail, she may find her academic career is at an end. Indeed, even if she does prevail, she may “have” to leave the university; I am thinking here of an actual case.
    I think that in some universities, perhaps my own, the consequences of harassment (as opposed to rape, say) are pretty light (but quite tideous) for the first time. You end up having to take classes; perhaps something goes on one’s record, I’m less sure of that. Does your argument depend on the severity of the consequences?
    I am less clear, I realize, about assault and rape; I would have thought they should go to the police, and I think they do, again at my university.

  53. annejjacobson Says:

    Looking at 50 and 51, I should say that I think the first place where issues of sexual harassement are filed is the ASO/EEOC office at my university; I am wondering how wide-spread this is, and so whether we are talking about the D of Ed’s advice being largely about how that office should treat cases.

    I do know of cases of sexual harassment getting taken up by the faculty committee, but that may have just been about appeals against the administrations decisions, and so a process much further down the road.

  54. Bijan Parsia Says:

    I’m with Jamie that our debate is getting stale (though, frankly, I think Jamie has the better of it, overall), so I’m happy to bracket them.

    Gideon, I don’t think punitive damages need to be common to serve my earlier point. I did articulate that you might set different expectations for how punitive damages are awarded.

    I think all the concerns about lack of proper procedure/professionalism are relevant, but they seem to hold for more advanced standards as well. Indeed, don’t they tell both ways? Is there a reason to think that the university will side systematically with e.g., student complaintants over faculty? (Of course, this moves us a bit from what would happen if the standard were successfully applied.)

    I don’t doubt that lower standards of evidence would change outcomes (otherwise, there’d be no point). The question is how the balance of harm is changed overall (for me).

    But…aren’t quite a few university decisions made essentially by amateurs? Tenure? (Or, at least, aren’t loads of tenure procedures procedurally dubious, esp. given the great career harms involved?)

    I would definitely want sexual harassment cases to be handled by a professional office in accordance to clear procedures. In that context, is preponderance of evidence always an unjust standard? Is it less so if any individual finding (by preponderance) only triggers comparatively mild sanction?

    I’m happy with an across the board strengthening of standards. I’m much less happy with an idiosyncratic strengthening of standards in sexual harassment cases without a clear, non-sexist justification. In other words, comparable harms and comparable sanctions (with comparable background information, e.g., probability distributions) should be treated similar, ceteris paribus, yes?

  55. Jamie Dreier Says:

    Anne, at Brown sexual harassment cases are definitely handled by the EEO/AA (I don’t know what your ASO is) office in the first instance, where they could be mediated, resolved informally, or investigated. The determination (including penalty, if any) is made by the Provost, on the basis of a report by the EEO/AA office. Appeals could go to a faculty committee. Of course, whatever any university body decides, either side can still have recourse to the legal system.

    My understanding of the DoE letter is that it is about university sexual harassment policies, however the university chooses to set them up.

  56. Bijan Parsia Says:

    Sorry to butt back in, but Jamie wrote:
    “Let me add one more thing. Bijan, you said that something is dubious, but I could not quite tell whether what you found dubious was that Preponderance is actually used in practice in civil trials, or that it is not. I am quite sure that it is not typically use in practice. It is the standard that is articulated, but it is not followed. ”

    Fair enough. And that may be sufficient to establish that extending it to circumstances where it might be applied literally would have severe negative effects.

    This, however, reminded me that I was curious about the 50% standard and your symmetry argument. If we grant symmetry of possible violation in your strongest sense (i.e., such that preponderance is the only coherent option), what’s wrong with flipping a coin to decide a case where both sides are roughly comparable? I find this hard to accept, but I don’t see how to argue against it (assuming strong symmetry).

  57. Jamie Dreier Says:

    It’s a good question why coin-flipping is not an acceptable method. My personal view is that it is publicly unacceptable but not for any very good reason.

    David Lewis once suggested that criminal penalties could reasonably be made chancy, but my sense is that people who work in the criminal justice system think this would be completely nuts.

  58. Jamie,

    Quite!

    Thanks for the mention the paper by Tribe, too.

    Cheers.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
Follow

Get every new post delivered to your Inbox.

Join 1,312 other followers