dissenting in part:
I join the majority’s decision to affirm summary judgment in favor of the defendants on Ray Gant’s racial harassment claim because I do not believe the record contains sufficient evidence that the defendants acted with deliberate indifference. I also join the majority in affirming summary judgment in favor of Grace Candido on the kindergarten transfer claim, because she was not ultimately responsible for the decision to transfer Ray. I cannot agree, however, that summary judgment was appropriate with respect to the role of Patricia Cronin and, by extension, Dr. Ci-rasuolo and the school board in transferring Ray Gant from first grade to kindergarten. Therefore, I respectfully dissent insofar as the majority affirms summary judgment in favor of Mrs. Cronin, Dr. Cirasuolo, and the school board defendants on the kindergarten transfer claim.
The majority apparently agrees that, given the “arguably unusual” nature of the transfer decision in this case, Ray succeeded in demonstrating pretext. See ante at 150. In my view, the majority’s characterization does not go far enough. I consider the treatment this lone black child encountered during his brief time in Cook Hill’s first grade to have been not merely “arguably unusual” or “indisputably discretionary,” but unprecedented and contrary to the school’s established policies. Viewed in the light most favorable to the plaintiff, the record indicates that every other Cook Hill student having academic difficulty received some form of transitional help, such as compensatory education, testing, or transitional classes. It further indicates that the school undertook these measures in consultation with the child’s classroom teacher and with the consent of his parents. In Ray’s case, however, Mrs. Cronin adhered to none of those procedures. Rather, she decided after nine days to transfer him into a grade he had already completed, over the objections of his parents and without consulting his teacher. Moreover, as the majority notes, Mrs. Cronin’s own statements contradict one another as to why she undertook this kindergarten transfer rather than moving Ray to the “transitional class” between first grade and kindergarten. See ante at 149-50. Given these circumstances, a jury would be entirely justified in finding Mrs. Cronin’s explanations for the transfer to be pretextual. See, e.g., Stern v. Trustees of Columbia Univ., 131 F.3d 305, 313 (2d Cir.1997) (noting that departure from procedures may suggest pretext for discrimination); EEOC v. Ethan Allen, Inc., 44 F.3d 116, 120 (2d Cir.1994) (holding that conflicting explanations serve as evidence of pretext).
I also take issue with the majority’s conclusion that although a jury might infer pretext from this evidence, Ray nonetheless failed to satisfy his burden of showing race discrimination. In Fisher v. Vassar College, 114 F.3d 1332 (2d Cir.1997) (in banc), we stated that a finding of pretext does not relieve the plaintiff in a discrimination case of his burden of demonstrating that “ ‘discrimination was the real reason’ ” for the adverse action. Id. at 1339 (quoting St. Mary’s Honor Ctr. v. Hicks, 509 U.S. 502, 515, 113 S.Ct. 2742, 125 L.Ed.2d 407 (1993)). We emphasized, however, that “a prima facie case and a finding of pretext may in some eases powerfully show discrimination.” Id. at 1338. In my view, this is such a case.
*152Ray presented significant evidence in support of his prima facie case, including the unprecedented nature of his transfer and the disparate treatment of similarly situated white students. In her deposition, for example, Mrs. Candido testified that two other students in Ray’s first grade class at Cook Hill were receiving “compensatory education” to remedy their difficulties in reading, but that this option was never offered to Ray. The majority apparently considers this evidence insufficient to demonstrate that similarly situated white students were treated differently because Ray did not prove that he and the two other students were similar in any respect other than their common need for remedial reading instruction. In my view, however, this similarity is more than enough to satisfy the plaintiffs evidentiary burden on summary judgment. Although it may be the case that Ray’s troubles were more serious than those of the other students, and that compensatory education thus would not have addressed his difficulties adequately, the burden of demonstrating such facts lies with the defendants and not with Ray. It is not the plaintiffs burden on summary judgment, when all inferences must be drawn in his favor, to show that he resembles members of the comparison group in every conceivable respect. Rather, once the plaintiff has adduced evidence to support a reasonable inference of similarity, as Ray did here, it is the defendant’s burden to show that the plaintiff is in fact dissimilar in some relevant way. The defendants here failed to make this showing, leaving Ray with an unrebutted prima facie case on the issue of disparate treatment.
Viewed in the light most favorable to the plaintiff, this prima facie evidence, combined with the showing of pretext, fully supports the inference that race discrimination played a role in Mrs. Cronin’s transfer decision. Unlike in Fisher, where the finding of pretext “suggested] numerous other possible unstated explanations no less likely than discrimination,” id. at 1345, a finding of pretext here would strongly suggest that race discrimination was one factor motivating the transfer to kindergarten. The defendants in this case, unlike the defendant in Fisher, had no apparent non-racial reason — such as “back-scratching, log-rolling, horse-trading, institutional politics, envy, nepotism, spite, or personal hostility,” id. at 1337 — to cover up their true rationale for transferring Ray. Cf. Fisher v. Vassar College, 70 F.3d 1420, 1448 (2d Cir.1995) (noting that the defendant’s real motivation for denying tenure may have been plaintiffs period of absence from academia, and that this motivation might have exposed the defendant to liability on other grounds), aff'd in banc, 114 F.3d 1332 (2d Cir.1997); cf. also Hollander v. American Cyanamid Co., 172 F.3d 192, 201-02 & n. 5 (2d Cir.1999) (noting that interpersonal friction, not age discrimination, may have been the real reason for plaintiffs termination). Only one circumstance in this case stands out as the likely reason for the discrepancy between the defendants’ treatment of other struggling students and their treatment of Ray: his race.
The majority maintains that this circumstantial evidence is not enough for Ray’s discrimination case to survive summary judgment. In so concluding, it seems to suggest that some additional, independent evidence of race discrimination — beyond the evidence adduced in the prima facie case and the showing of pretext — is necessary. However, “plaintiffs in discrimination suits often must rely on the cumulative weight of circumstantial evidence, since ‘[a defendant] who discriminates is unlikely to leave a “smoking gun Norton v. Sam’s Club, 145 F.3d 114, 119 (2d Cir.1998) (quoting Rosen v. Thornburgh, 928 F.2d 528, 533 (2d Cir.1991)). Indeed, a plaintiff like Ray Gant will seldom be able to offer direct evidence of discrimination on the part of educators.
In my view, however, the circumstantial evidence Ray produced in meeting his pri-ma facie and pretext burdens is easily *153sufficient to send the question of discrimination to the jury. Drawing all inferences in favor of the plaintiff, a jury might well conclude that Mrs. Cronin, confronted with a child whose academic difficulties were significant but not (at least in reading) necessarily worse than those of other stu-dénts, jumped to the conclusion after only nine days that the student would not succeed in the first grade or in a transitional class, and that demotion to kindergarten was the only viable option. Although the majority characterizes this decision as “amply supported by academic considerations,” ante at 149, many of those considerations arise in the form of after-acquired evidence — evidence that does nothing to show that Mrs. Cronin, at the time she made the transfer decision, had enough information about Ray’s capabilities to evaluate his placement adequately. Moreover, Mrs. Cronin made this decision despite having never taken such action with respect to a white child, despite the school’s history of giving white students compensatory help, and without directly consulting the classroom teacher, who had suggested to Ray’s mother that Ray receive such compensatory assistance and not that he be transferred to kindergarten. Mrs. Cronin also disregarded the recommendation from a prior teacher who, after spending nearly six months with Ray in her classroom, suggested not that he be moved back, but that he be evaluated by a child study team in order to determine what kind of compensatory help he required. Underlying this abrupt decision-making process, furthermore, was the crucial fact that Ray was the only black child in his classroom and one of the very few black students in the entire school.
Reading the record in this light, a jury reasonably could conclude that the school did not give the black student an equal chance to succeed (or fail). Contrary to the suggestion of the majority, such a conclusion would be based not “merely” on a finding that the events in this case were “unusual,” but on the inference, drawn from substantial circumstantial evidence, that they were tainted by race discrimination. In my opinion, Ray was entitled to an equal opportunity to learn, and — failing that — a full hearing in court. Accordingly, insofar as the majority affirms summary judgment in favor of Mrs. Cronin, Dr. Cirasuolo, and the school board defendants on Ray’s transfer claim, I dissent.