concurring in the judgment.
Like the majority, I believe that Curry presented enough evidence to raise a material dispute as to whether similarly situated, non-African American employees received preferential treatment from Me-nard. I disagree, however, with the majority’s conclusion that the number of potential comparatives is necessarily limited to those individuals who had cash discrepancies during Michael Stanley’s approximately four-month tenure at the store. The majority rests its decision on the assumption that Stanley intended to enforce the progressive discipline policy more strictly while he was in charge. But that point is disputed. For instance, Curry presented testimony from assistant manager McDaniel (who started working at Menard one month after Stanley) that he never discussed the progressive discipline policy with Stanley or Horvath and could not recall any conversations he had with Horvath to the effect that the new manager wanted to “get tougher” on employees. Assistant office manager Ros-ner also testified that neither Stanley nor McDaniel ever spoke to him about the policy and that the policy was never discussed at any manager meeting. And, as the majority acknowledges, even while Stanley was in charge there were two non-African American employees, Mercu-rio and Venetico, who should have been disciplined but were not. In fact, the only evidence supporting the majority’s assumption is hearsay of questionable admissibility' — -Horvath’s testimony that Stanley told her the policy was to be rigidly enforced — and, even more dubious, Curry’s “admission” during her deposition that there was a “rumor going around” that Stanley was brought in to tighten up discipline. Curiously, Menard did not offer any testimony from Stanley himself (or for that matter from McDaniel, who was second-in-charge) that Stanley had implemented a policy of strict enforcement. Thus, in light of all the evidence of record and our obligation to view it in the light most favorable to Curry, I think that the majority unduly restricts the relevant time frame to the four-month period when Stanley was managing the store. I therefore respectfully disagree with that portion of the majority’s decision.
I offer two additional comments. First, regardless whether the number of other *481similarly situated individuals is 2 or 16 for purposes of the prima facie test, I note that nothing precludes Curry from offering evidence of all 16 at trial. As we have often stated, the burden-shifting approach of McDonnell Douglas applies only to pretrial proceedings and drops out once a case goes past the summary-judgment stage. E.g., Hamner v. St. Vincent Hosp. and Health Care Ctr., Inc., 224 F.3d 701, 705 n. 3 (7th Cir.2000); Hasham v. Cal. State Bd. of Equalization, 200 F.3d 1035, 1044 (7th Cir.2000); Diaz v. Fort Wayne Foundry Corp., 131 F.3d 711, 712 (7th Cir.1997).
Second, although I agree with the majority that Curry’s testimony regarding Horvath’s racial remarks was not direct proof of discrimination, that should not prevent Curry from introducing the same evidence to prove pretext should Menard renew its motion for summary judgment. Pafford v. Herman, 148 F.3d 658, 666 (7th Cir.1998); Huff v. UARCO, Inc., 122 F.3d 374, 385 (7th Cir.1997). Menard argues otherwise, citing cases such as Hong v. Children’s Memorial Hospital, 993 F.2d 1257 (7th Cir.1993), for the proposition that, before a plaintiff may rely on circumstantial evidence, she must show that there is a nexus between that evidence and the allegedly discriminatory decisions at issue. But those cases simply hold that such circumstantial evidence, zuithout more, is insufficient to give rise to an inference of discrimination. See, e.g., id. at 1266. And here, Curry’s circumstantial evidence was not offered alone but in conjunction with other proof of pretext. See Huff, 122 F.3d at 385 (“When a plaintiff uses the indirect method of proof, no one piece of evidence need support a finding of ... discrimination, but rather the court must take the facts as a whole.”).
Menard also contends that Horvath’s comments may not be considered because she “was not even present when Curry was terminated and there is no evidence whatsoever that either of the managers who were present (McDaniel or Rosner) had made any derogatory comments or acted in any way which would even suggest that their actions were motivated by racial animus.” But Menard does not dispute that Horvath personally imposed the written warning and the suspension on Curry (actions upon which the termination was based), and that Rosner, who discovered Curry’s third cash discrepancy, consulted with Horvath about the proper disciplinary action to be taken. Furthermore, McDaniel testified that he exercised no discretion in determining whether to discipline an employee for cash discrepancies; instead, he based his decisions solely on the writeups he would receive from Horvath, without investigating their validity or whether they were issued in a consistent manner. Horvath was therefore in effect the decision-maker in Curry’s termination, and so evidence of any racial animus she harbored may be considered in the pretext analysis. See Russell v. Bd. of Tr. of the Univ. of Ill. at Chi, 243 F.3d 336, 342 (7th Cir.2001) (“[A]ny improper motives [the plaintiffs supervisor] harbored had to be imputed to the other members of the disciplinary committee because of [that supervisor’s] extensive role in initiating and carrying out the disciplinary process.”); Hunt v. City of Markham, III., 219 F.3d 649, 652 (7th Cir.2000) (expressions of discriminatory feelings by “those who provide input into the [adverse employment] decision” are evidence of actionable discrimination).