CARNES, Circuit Judge, concurring specially: I concur in the holding that Title IX does not create a cause of action against public school boards or officials for failure to prevent or remedy student-student sexual harassment. In my view, that holding is correct for essentially those reasons stated in Parts I, II, III A, and IV of Judge Tjoflat’s opinion, and I join those parts of it, which constitute the opinion of the Court.

However, for the reasons explained below, I do not join Parts III B and C of Judge Tjoflat’s opinion, which express only his own views.1

I.

The “Hobson’s choice” or “whipsaw liability” discussion in Part III B of the opinion is based upon a fundamentally erroneous premise. If school officials could be sued for failing to prevent or remedy student-student sexual harassment, that part of the opinion says, the potential liability would amount to a financial incentive to punish the accused harassers, which would or could render school officials impermissibly biased and require recusal.

Of course, a student does have a property interest in a public education which is protected by the Due Process Clause of the Fourteenth Amendment.2 And, due process does require that a Parts I, II, III A, and IV of Judge Tjoflat’s opinion constitute the opinion of the Court, because those parts are joined by six of the ten judges participating in this decision. By contrast, none of the other nine judges participating in this decision have joined Parts III B and C of that opinion.

The nature and extent of the protection afforded the property interest in a public education, the due process requirements decision depriving the student of that property interest be made by someone who does not have a pecuniary interest in having the student suspended or expelled. To take an extreme example, regardless of any other process afforded, due process would be violated if a principal took a bribe from the complaining student’s parents in return for suspending or expelling the alleged wrongdoer. But it is an entirely different matter to suggest, as Part III B of the opinion does, that a school official’s potential liability to the complaining student if that official fails to take legally required action amounts to a “financial incentive” which renders that official “impermissibly biased” and requires recusal from deciding what action, if any, is required in the

attendant to its loss, depends upon the severity of the loss. In Goss v. Lopez, 419 U.S. 565, 95 S. Ct. 729 (1975), the Supreme Court held that, with any suspension of up to ten days, all the Due Process Clause requires is for the student to “be given oral or written notice of the charges against him and, if he denies them, an explanation of the evidence the authorities have and an opportunity to present his side of the story.” 419 U.S. at 581, 95 S. Ct. at 740; accord Arnold v. Board of Educ., 880 F.2d 305, 318 (11th Cir. 1989). The Supreme Court said in Goss that “[i]n the great majority of cases the disciplinarian may informally discuss the alleged misconduct with the student minutes after it has occurred,” and “[w]e hold only that, in being given an opportunity to explain his version of the facts at this discussion, the student first be told what he is accused of doing and what the basis of the accusation is.” 419 U.S. at 582, 95 S. Ct. at 740. The Court has since explained that all Goss requires before a suspension is an “informal give and take” in order to provide the student “the opportunity to characterize his conduct and put it in what he deems the proper context.” Board of Curators v. Horowitz, 435 U.S. 78, 86, 98 S. Ct. 948, 953 (1978) (quoting Goss, 419 U.S. at 584, 95 S. Ct. at 741); accord, e.g., C.B. v. Driscoll, 82 F.3d 383, 386 (11th Cir. 1996) (“The dictates of Goss are clear and extremely limited.”). These “rudimentary precautions,” to use the description from Goss itself, 419 U.S. at 581, 95 S. Ct. at 740, are a far cry from a due process tribunal hearing attendant to some property interest deprivations.

circumstances. As authority for that novel proposition, the opinion cites only Gibson v. Berryhill, 411 U.S. 564, 579, 93 S. Ct. 1689, 1698 (1973). The Gibson decision provides no support for the proposition, because it does not hold, or even imply, that an official’s potential liability for failing to properly exercise decisionmaking authority constitutes a “financial incentive” which renders the official “impermissibly biased.”

Gibson involved a state optometry board composed exclusively of private practitioners who were in competition with corporate employee optometrists. Those board members had a substantial pecuniary interest in excluding from the market corporate employee optometrists, who accounted for nearly half of all the practicing optometrists in the state. The Supreme Court affirmed the district court’s holding that the private practitioner’s pecuniary interest in eliminating competition disqualified them from deciding whether the practice of optometry by corporate employees as such constituted unprofessional conduct justifying license revocation.

See 411 U.S. at 578-79, 93 S. Ct. at 1698. That holding does not support the proposition that any time an official can be sued for failing to respond properly to a complaint that official is disqualified from making a decision about how to respond to the complaint.

If that suggested proposition were the law of this circuit -- and thankfully it is not -- no school official could ever discipline a student for any alleged misconduct as a result of another student’s complaint without violating the due process rights of the disciplined student. The reason such an imposition of discipline would violate due process is that such an official would always have a financial incentive, under that view, to believe the complaint in order to avoid a lawsuit filed by the complainant. The ramifications of such a rule would extend to discipline for any type of misconduct, because there is no principled basis on which a distinction can be drawn between discipline following a complaint about sexual harassment and that following a complaint about any other type of misconduct.

Nor is there any principled basis by which such an automatic disqualification rule could be confined to school settings. It would also apply outside the Title IX context; for example, in jail and prison settings. If one prisoner complains to a jailer or warden about what some other prisoner has done to him, under Judge Tjoflat’s view that official will have a financial interest in avoiding a lawsuit from the complaining prisoner (alleging deliberate indifference), and such an interest disqualifies the official from making any disciplinary decision about the complaint.

So, not only would the disqualification rule be automatic, it also would be universal. No one would be able to decide any disciplinary matters in schools, in prisons, or in any other setting within the purview of the Due Process Clause. All federal, state, or local officials called upon to decide what to do in response to one person’s complaint about another would have a financial incentive to avoid a lawsuit, which would disqualify them

from making a decision. That cannot be the law, and it is not the law.

Judge Tjoflat’s response to having these flaws in his reasoning pointed out is contained in footnote 21 of his opinion, which will reward close scrutiny. First, that footnote assures us that we should not worry about the far-reaching ramifications of the suggestion that potential liability equals disqualifying bias, because this Court is holding that school officials have no liability under Title IX for student-student sexual harassment.

Apparently forgotten is the assurance, in Part IV of the opinion, that “Georgia tort law may indeed provide redress” for the very same conduct. If a school official’s potential liability for not acting properly is a disqualifying financial interest, it matters not whether that potential liability is posed by Title IX or by state tort law. The opinion does not, and logically cannot, suggest otherwise. Instead, it adopts a head-in-the-sand approach which ignores everything but Title IX, as though that were the only potential source of liability for school officials who are called upon to decide what to do about student-student sexual harassment complaints.

With its head comfortably in the sand, the opinion also ignores entirely the obvious implications of its proposition for student-student disputes involving allegations of misbehavior other than sexual harassment. Part of the quotidian business of teachers and principals is resolving disputes in which one student alleges another has threatened, hit, stolen from, or otherwise mistreated

him or her. Some of those disputes pose potential liability for the teacher or principal who fails to act. For example, a school official who fails to take appropriate action to protect a student from a threatened thrashing at the hands of another student may have to answer in a state court tort action. Under the reasoning contained in Part III B of the opinion, that potential liability would prevent any school official from deciding what to do about such a complaint, because that official’s potential liability to the complaining student would amount to a disqualifying financial bias. A careful reading of the opinion reveals that it fails to explain why that result would not necessarily follow from its suggested reasoning.

As to settings outside the school context, footnote 21 of the opinion offers two responses to this criticism. First, it simply denies SS “We suggest nothing of the kind” SS that its proposition about potential liability equaling disqualifying bias would have any application outside the schoolhouse. That ipse dixit assertion has as little reasoning behind it as the proposition itself. The opinion fails to offer any reason why the automatic bias theory it suggests would not apply in non-school contexts, because there is no reason. The right to an unbiased decision maker is a rudiment of due process, which is as applicable outside schools as within them.

Apparently realizing that the ipse dixit approach will not shield the naked illogic of its position from view, the opinion attempts to camoflauge the problem with talk of immunity. “Don’t worry,” we are told, officials in non-school settings have “immunity from suit” which removes any potential liability for failing to decide for the complaining party, and any financial incentive to favor that party disappears along with the potential liability. The thinnest stripe of the attempted camouflage is the opinion’s reference to judicial immunity. We are not talking about judges. We are talking about the myriad of federal, state, and local non-judicial officials who are regularly called upon to decide what to do in response to one person’s complaint about another. Jailers, wardens, and other corrections officials are but a few examples. These people are not judges. They do not enjoy judicial immunity.

Even so, the opinion says, there is qualified immunity. There are three problems with the assertion that the availability of qualified immunity distinguishes non-school officials from school officials by removing any threat of lawsuit by a complaining party dissatisfied with an official’s resolution of a complaint outside the school setting. First, qualified immunity is not absolute.

Second, qualified immunity does not shield officials from liability grounded on state law. Third, and most obviously, the doctrine of qualified immunity is the same for school officials as for non- school officials. If that doctrine shields non-school officials from threat of lawsuit sufficiently to remove any disqualifying financial incentive to decide for a complainant, it does exactly the same for school officials. Thus, with its talk of qualified immunity, Part III B of the opinion has succeeded in reaching

around and biting itself in the back. If what the plurality opinion says about the due process implications of qualified immunity is true, then the opinion has disproven the very proposition it is seeking to defend.

II.

Part III C of Judge Tjoflat’s opinion attempts to establish that student-student sexual harassment is such a widespread and extensive problem that a different holding in this case would impose massive liability upon school officials and boards. In its words, agreeing with appellant’s theory of liability would give rise to “thousands of lawsuits.” Tjoflat Opinion at n.25. The factual premise of that reasoning is based entirely upon one survey report. See American Ass’n of Univ. Women Educ. Found., Hostile Hallways: The AAUW Survey on Sexual Harassment in American Schools (1993) (hereinafter “AAUW Survey Report”).

The AAUW Survey Report was not the subject of an evidentiary hearing in the district court, nor has it been examined in a hearing in any other court insofar as we know. Neither party to this appeal even mentioned the survey in the briefs; it was discussed only in one amicus brief. In general, we should be reluctant to incorporate into our reasoning the results of a survey that has not been examined critically or tested in a trial or evidentiary hearing, the time-honored and proven methods our system of justice uses to determine material facts.

Beyond the general problems with using surveys in judicial decision making, there are specific reasons why employment of this particular survey for the purpose Judge Tjoflat uses it in Part III C of his opinion is ill-advised. That purpose, of course, is to show student-student sexual harassment is so rampant that if a cause of action existed for it the resulting flood of litigation would inundate our public school systems, or at least school officials would have a basis for fearing that result SS the basis being the survey.

The first reason we ought to be especially cautious about such a use of this particular survey is that its purported findings are, in the words of the sponsors of the survey: “startling,” and for some “the results will be surprising and shocking.” Id. at 2. The reason for such descriptions is that it is difficult to believe that 65 percent of all eighth through eleventh grade students have been sexually harassed by other students, and that half of all female and male students in those grades are self-professed sexual harassers. We ought to be reluctant to accept as fact, or assume that school officials would accept as fact, such “surprising and shocking” statistics based upon a single survey of only a tiny fraction of one percent of the total number of students in four grades.

Even a cursory look at the survey report gives more reason to be dubious about the opinion’s use of the report. The survey asked students how often “[d]uring your whole school life” has anyone “when you did not want them to” done any of the following things, and it then provided a list of behavior the survey defined as sexual harassment. See id. at 5. Some behavior on that list clearly constitutes sexually harassing behavior of the most serious type. But included in the list is other behavior that is less serious and far less likely to lead to complaints and litigation, which is what Judge Tjoflat uses the survey to predict (or posits that school boards will use it to predict). For example, included in the survey’s definitional list of sexual harassment was any instance in which another student: “Made sexual comments, jokes, gestures, or looks;” or “[s]pread sexual rumors about you;” or “[s]aid you were gay or lesbian.” Id. at 5. Remember that a single unwelcome instance of such activity, during the student’s entire school life, renders that student a victim of sexual harassment for purposes of the survey.

A student who has ever been looked at by another student in an unwelcome way perceived to be sexual is defined by the survey to be a sexual harassment victim. Any student ever called gay or lesbian is also a sexual harassment victim in the survey’s view. Any time unwelcome rumors are spread about a student having any type of sexual activity (presumably including kissing) with another student, those students are sexual harassment victims as the survey defines it. To take one final example of how the total incidence of “sexual harassment” reported overstates legally actionable incidents of sexual harassment, consider that the survey definition includes incidents in which someone “[f]lashed or ‘mooned’ you.”

Id. At 5. Suppose that a student at a school function (which the survey defines to include school sporting events and field trips) “moons” all the students in attendance, or all those from a rival school. A single episode of that misbehavior SS which is not nice and certainly should not occur, but has been known to happen SS could make sexual harassment victims, as the survey defines the term, out of scores or even hundreds of students. Yet, such an incident is extremely unlikely to result in litigation against the school.

It is also worthy of note that the survey asked students whether the behavior it defined as sexual harassment had happened to them “[d]uring your whole school life.” Id. at 5. Therefore, the 65 percent figure reflects those who have experienced that behavior at any time during any school year of their life. It does not purport to be annual data.

Finally, Part III C of the opinion fails to point out that the survey also asked the students if any of them who had been sexually harassed, as that term was defined in the survey, had told a teacher about the experience. Only 7 percent of the sixty-five percent had. See AAUW Survey Report at 14. Whatever the reasons for not reporting such behavior to a teacher, the failure to do so in all but the rarest instances has obvious implications for the existence of causes of action against schools or the likelihood of actual litigation.

The opinion attempts to deflect criticism about misuse of the survey by suggesting that while the opinion’s author does not necessarily think that the survey is a valid indicator of how much student-student sexual harassment occurs, school boards might think that the survey is and reject federal funding as a result of it.

With all due respect, there is no reason to believe that school boards would be less likely than federal judges to see the flaws in such an interpretation of the survey. School boards know more about what is going on in their schools than we do, and they can be expected to critically examine any survey before using it as a basis for turning down federal funding for their schools. Rather than hiding behind speculation about how school board officials might interpret the survey, the opinion ought to face up to the flaws in its suggested use of the survey.

Upon its release, the sponsors of the survey stated that they were “confident that the results of this survey will become a focal point on the agendas of policy makers, educators, and others concerned with the education of America’s children.” Id. at 21.

Their confidence about how the survey would be used might be undermined by Part III C of Judge Tjoflat’s opinion. More importantly, we are not policymakers. We do not have agendas. We ought to leave this survey to those who do.

III.

The parts of Judge Tjoflat’s opinion that neither I nor any other member of the Court except its author joins, Parts III B and C, are not necessary to the opinion’s essential reasoning or to the holding of this case. Neither the language of Title IX nor its legislative history indicates that Congress intended to saddle school boards and officials with liability for student-student sexual harassment, and school boards had no notice that such liability would result from accepting Title IX funds. For those reasons, I do join the holding of the Court and Parts I, II, III A, and IV of Judge Tjoflat’s opinion.