[rayiner]

So during the Obama era, rules were put in place to force colleges to deal with sexual assault accusations. The basic idea is sound: the university needs to do something when one student accuses someone else. Applying a criminal standard of proof is incredibly unfair. In the case of a false negative, that results in some student being forced to continue attending classes with her rapist, or to give up educational opportunities by leaving the school herself. Unlike in the criminal context, a false negative (incorrect acquittal) has a direct prejudicial impact on the victim. A rape victim whose rapist doesn’t get convicted may suffer a moral loss, but it’s not like conviction can undo the rape. Accordingly, many women don’t even press charges. But in a college context, the victim is seeking something that can actually help. Not punishment of the accused, but freedom to continue her education without the threat posed by the accused.

Once you realize that false negatives are as harmful as false positives, then it makes sense to use a civil standard instead of putting a thumb on the scale for the accused. Statistically, sexual assault is far more common than false accusations. Applying a criminal standard results in a lot of social harm through false negatives that isn’t outweighed by avoiding false positives.

Where I think Title IX goes off the rails is treating these cases as being about punishment. Civil standards aren’t for punishment, they’re for resolving disputes. In this context, it’s not about deciding whether the accused is guilty of rape, but about deciding: “which student should be the one to leave?” I think Title IX resolutions should be secret and the only remedy should be expulsion or some sort of internal restraining order.

[MaxBarraclough]

Wasn't planning on batting politics here on HN, but - I'm inclined to agree with asfd... who has apparently deleted their comment. How annoying.

> The basic idea is sound: the university needs to do something when one student accuses someone else.

No, that isn't sound. It's a criminal matter. You don't inform your professors, you don't inform your employer, you don't inform your gardener, you inform the authorities. We're not talking about plagiarism here, we're talking about a criminal offence.

As a Brit, the idea that a college should have some sort of mini criminal justice system, seems utterly absurd. The only institution with its own criminal justice system, is the military, and they take it every bit as seriously as the civilian world.

> Applying a criminal standard of proof is incredibly unfair.

We're talking about adults dealing with crime. How is the criminal justice system any more unfair simply because both the victim and the suspect are both students at the same institution?

> Unlike in the criminal context, a false negative (incorrect acquittal) has a direct prejudicial impact on the victim.

That's a nasty attribute of rape, and of various other crimes, not of colleges.

Aren't most rapes committed by someone the victim already knew? You seem to be treating the college case (where both are students of the same institution) as categorically different from a case where the suspect is a friend (well, 'friend') or colleague of the victim.

I don't see it. I certainly don't see it as a justification for lowering the burden of proof, or for letting the suspect escape ordinary criminal-justice proceedings.

> But in a college context, the victim is seeking something that can actually help. Not punishment of the accused, but freedom to continue her education without the threat posed by the accused.

Again I don't see a category difference here. If the suspect is a colleague, or a member of a social circle shared with the victim, the same issue arises.

> Once you realize that false negatives are as harmful as false positives

Of course false negatives are harmful, it would be absurd to deny that, and no-one is doing so. That is the case for all serious crimes. It remains necessary to insist on a high standard of proof, and on presumed innocence.

> sexual assault is far more common than false accusations

You seem to be assuming that this ratio is set in stone. If you make it easier to make a false accusation 'stick', we would expect them to happen more often.

> In this context, it’s not about deciding whether the accused is guilty of rape

But it is. The consequences of the accused being found guilty can be severe, no? If a rape has occurred, a crime has been committed. Dealing with crimes is not within the purview, or the competences, of a college.

[rayiner]

> No, that isn't sound. It's a criminal matter.

What distinguishes criminal matters from civil matters is the remedy sought. For example, OJ Simpson was acquitted of murder, applying criminal standards of proof. But he was found liable for wrongful death, and ordered to pay damages to the families, applying civil standards of proof.

Criminal law and civil law address different concerns, and the same conduct can raise both criminal and civil issues. Criminal law vindicates the government's interest in retribution, deterrence, or rehabilitating an offender. It does not generally address the victim's personal rights. Thus, the government can prosecute when the victim does not want to, and vice versa.

In a campus rape case there is a criminal aspect. But there is also a civil aspect. Students have a civil legal entitlement to pursue their education at schools receiving government funding free from harassment. Forcing a student to either continue attending classes with her rapist, or to leave the school is undoubtedly an infringement of that legal entitlement. That civil concern is wholly distinct from the criminal question of punishing the rapist.

> That's a nasty attribute of rape, and of various other crimes, not of colleges.

No, it's a distinction between civil and criminal aspects of a course of conduct.

Say you hit me with a car because you were driving drunk. There is a criminal aspect to the case (driving drunk is a crime), but there is also a civil aspect to the case (hitting someone with a car due to negligence is a tort giving rise to damages liability). You being erroneously acquitted of drunk driving under a criminal standard of proof doesn't hurt me. It's a moral loss, nothing more. But you being erroneously found not liable for negligent driving does hurt me. It means I'm forced to bear my medical costs, and have been erroneously denied compensation. The false negative (an wrongfully injured person being forced to bear their own medical costs) is just as bad as the false positive (a wrongfully accused person being forced to pay someone else's medical costs). So in the civil context, we do not apply the maxim of "better to let ten guilty men go free than convict one innocent man." Because it's not better to find 10 people not liable for drunk driving when they did so than to find one person liable for drunk driving when he did not. That results in 10 victims who are wrongfully denied compensation for their injuries. So that's not the standard we use. Instead, we weigh credibility under a "more likely than not standard."

The same is true for sexual assault and sexual harassment in the workplace or the school. Applying criminal standards of proof in that context, what you're saying is: "it's better to force 10 people to continue to attend classes/go to work with their harassers/rapists than to force 1 person to quit their job/school due to a false accusation." But that doesn't make any sense. In both cases (false positives and false negatives), someone is wrongfully denied educational opportunities. Saying that it's okay to have 10 of one outcome to avoid 1 of the other outcome basically just says that false accusations are somehow worse than sexual assault.

[wildmusings]

I disagree. The Title IX kangaroo courts were a disgrace to America. The accused had no right to an attorney, no right to remain silent, no right to face their accuser. They faced panels, of professors, college administrators, and naive fellow students, that were politically hostile to young men and that accepted allegations as necessarily true. Even in civil cases, where there is a preponderance standard, the burden of persuasion is on the plaintiff. Here, the burden was effectively on the defendant.

The consequences of a finding against the accused was ruinous to that person's career, and wasteful of their previous investment in their diploma. The consequences of the process itself, where the accused was subject to treatment that would grossly violate the Bill of Rights if it were the government's doing, was likely to prejudice any criminal prosecutions or real civil cases. And yet it was the government mandating this treatment.

This is different than an employment situation, where the at-will relationship is understood to govern and employees aren't generally seen as entitled to their job or any due process for ending it. Besides occupying a special place in American society as a necessary rite of passage, a university is charging students tens of thousands of dollars per year to attend. To expel a student without adequate due process, wasting their previous investment in their diploma, is unconscionable.

In many of these cases where students were expelled, the only evidence was the allegation itself. Is that the world you want to live in, where a young person's life can be ruined by a single malicious accusation?

[rayiner]

I agree that the university processes are problematic. But the problem isn't the standard of proof. Universities were treating the Title IX process as something with criminal consequences, e.g. metting out punishments for the sexual assault itself. But if you apply civil standards of proof, you can't mete out punishments. You can only adjudicate rights: "which one of us gets to stay?" Title IX proceedings should be directed at nothing more than adjudicating that issue. There should be no public show, there should be a sealed record, etc.

> Is that the world you want to live in, where a young person's life can be ruined by a single malicious accusation?

The alternative is the world we live in, where a different kind of malicious act (a sexual assault), can force a young person to leave school, or impair her education because she's forced to continue to attend classes with someone who attacked her. For too long, the reaction to accusations of sexual assault was to shift the cost of reduced educational and career opportunities to the victims. The victims were the ones shuffled to different classes, or to different groups in the company, or encouraged to leave the school/company entirely. That's not the world I want to live in either.

You can't ignore the cost of false negatives in order to avoid false positives. In the false positive scenario, a young person is forced to leave school due to a malicious act (a false accusation). In the false negative scenario, a young person is forced to leave school, or continue to go to school with someone who attacked her, due to a malicious act (a sexual assault). Both of those outcomes are bad, and if you're being rational, you're trying to minimize SocialCost(false_positives) + SocialCost(false_negatives).

Setting the burden of proof high to minimize false positives does not minimize the total social cost. Indeed, in a world where sexual assault is far less common than false accusations thereof, it's completely irrational to set the standard in a way that minimizes the cost from false positives at the expense of increasing the cost from false negatives.

[MaxBarraclough]

> if you apply civil standards of proof, you can't mete out punishments. You can only adjudicate rights: "which one of us gets to stay?"

So you're fine with expelling a student (and likely ruining his life) on the basis of the findings of this 'court', so long as we don't admit that it's effectively a punishment?

> You can't ignore the cost of false negatives in order to avoid false positives.

So you disagree with Blackstone's "It is better that ten guilty persons escape than that one innocent suffer", I take it?

Erring on the side of 'not guilty' is a cornerstone of civilised justice systems. It isn't 'ignoring' anything.

> if you're being rational, you're trying to minimize SocialCost(false_positives) + SocialCost(false_negatives).

No. Again I refer you to Blackstone's formulation. Your utility function fails to capture the whole picture.

> Setting the burden of proof high to minimize false positives does not minimize the total social cost.

Yes it does. Blackstone was right.

What you are suggesting undermines the legitimacy of the justice system, which brings great costs. It also makes false accusations more likely, as well as more harmful.

It also assumes that a bad outcome is no worse for having been committed by the state, which is not so; it is far worse that an injustice be committed by the state than by ordinary citizens.

[wildmusings]

I think you need to consider the second order effects. If we create a world where people have the power to destroy others with a mere accusation, that will change how people behave. It will create an atmosphere of vindictiveness, fear, and distrust.

Having to deal with people who have wronged you, but who you can't prove have wronged you, is just an unfortunate reality of the world. That's how all people lived before the rise of big cities (if they lived in a society with due process), and it's still how millions of people live today in America. If you live in a small town and you get robbed, or mugged, or raped by someone, but the government can't prove it in court, guess what? You can't do anything about the fact that they frequent the same supermarket and same bar, send their kids to the same school, and so forth. In the social contract, this is the cost we pay for the security of our own rights should we ever be accused of wrongdoing. (And of course, a university is not a government or a town, but for the reasons I lay out in the previous comment, I think they should be treated as almost being such. And the comparison informs our intuition as to which social cost is better; it turns out we've already done such analysis and decided, in one similar sphere.)

For obvious reasons, there's no solid numbers on the false accusation rate, but the FBI has found that 8% of forcible rape reports were found to be false by investigation. And remember, those are the cases that the accusers decided to take to police. If you were the type inclined to make false allegations, you would be further encouraged by the lower burden of proof in university tribunals.

When they opened the East German archives, they found that ordinary people had turned the Stasi into their vehicle for personal vendettas. Got into an argument with your neighbor? Scorned by a lover? Stick the Stasi on them. And this was very widespread. A system created for one purpose, to punish wrongdoers of a certain kind, ended up being used to settle personal feuds. Such is the nature of man. Give people incentives to do bad things, and they will do them.

>But if you apply civil standards of proof, you can't mete out punishments. You can only adjudicate rights: "which one of us gets to stay?"

Potayto potahto. Getting expelled from your university for sexual assault is the end of you. Whether you call it a punishment or 'adjudication of rights' is for the law professors.

(And small point: the way you're framing it is a false dilemma, since there is a third option, 'both', which I discuss above.)

[rayiner]

> I think you need to consider the second order effects. If we create a world where people have the power to destroy others with a mere accusation, that will change how people behave. It will create an atmosphere of vindictiveness, fear, and distrust.

But Obama didn't propose that accusations alone should be able to get someone expelled. He proposed applying the same preponderance standard that applies, and works just fine, in civil litigation.

> Having to deal with people who have wronged you, but who you can't prove have wronged you, is just an unfortunate reality of the world.

First, that's not true. We have both formal and informal ways of dealing with people who have wronged you, who you can't have convicted. For example, you can get a restraining order in Maryland using the same "preponderance of the evidence" standard Obama proposed for Title IX proceedings. You can sue someone for fraud or breach of contract for screwing you on a business deal, under the same preponderance standard. You can also use informal social mechanisms. Entrepreneurs live and die by their reputations, for example.

Second, this begs the question. I pointed out that in the case of both false negatives and false positives, someone is denied the opportunity to pursue their education or their employment. My thesis is that we should not treat one of those bad outcomes as being better than the other. Your response boils down to "false positives are really bad, but false negatives are just something you have to deal with." But why?

> Give people incentives to do bad things, and they will do them.

Incentives are created in both directions. How many careers do you allow Harvey Weinstein to wreck in order to avoid erroneously wrecking the career of a film producer who is innocent? That's the calculus you refuse to grapple with.

[asfdsfggtfd]

.

[danharaj]

> In the case of a false negative, that results in some student being forced to continue attending classes with her rapist, or to give up educational opportunities by leaving the school herself.

You're excused.