GWIN, District Judge.
Plaintiffs, Thomas Cicero and his wife Marlene Cicero, appeal from a district court order granting summary judgment to Defendant Borg-Warner Automotive Automatic Transmissions Systems Corp. (“Borg-Warner”). The district court gave Borg-Warner judgment on Thomas Cicero’s age discrimination claim and dismissed [582] Plaintiff Marlene Cicero’s loss of consortium claim.
In this case the plaintiffs sued for age discrimination under Michigan’s Elliott— Larsen Civil Rights Act (“Elliott-Larsen Act”) and the common law of Michigan. Cicero said that Borg-Warner, his former employer, discriminated against him based on age when it fired him. Finding that Cicero failed to establish a prima facie case of age discrimination because he did not show he was qualified for his position, the district court granted Borg-Warner’s summary judgment motion.1 The district court dismissed Plaintiff Marlene Cicero’s loss of consortium claim because it was derivative of Thomas Cicero’s claim. The plaintiffs appeal from the district court’s decision on both claims.
I. Overview of Appeal
This case comes from Borg-Warner’s firing of Plaintiff Thomas Cicero from his job as human resources manager, In giving Borg-Warner summary judgment, the district court misread the requirements for establishing a prima facie case of discrimination set forth in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973).
In deciding that the district court erroneously granted summary judgment, we first examine whether Cicero showed all of the elements necessary to establish a pri-ma facie case of discrimination under the McDonnell Douglas analysis. The district court found that Cicero did not show that he was qualified, an element of the prima facie case. We find the district court erred when it made this ruling.
After finding that Cicero was qualified, we next look to whether Borg-Warner comes forward with a nondiscriminatory justification for firing Cicero. Because Borg-Warner does come forward with a nondiscriminatory justification, we then consider if Cicero shows evidence sufficient to make out an issue that Borg-Warner’s justification is a pretext.
II. Background
A. Procedural Background
Plaintiffs-Appellants Thomas and Marlene Cicero sued Defendants Appellees Borg-Warner Automotive and Borg-Warner in the Circuit Court for the County of Wayne, Michigan. In their complaint, the Ciceros alleged that the appellees unlawfully fired Thomas Cicero because of his age. The plaintiffs brought their claim under Michigan’s Elliotb-Larsen Civil Rights Act and the common law of Michigan. Mich. Comp. Laws §§ 37.2101 et seq. (2001). Marlene Cicero filed a claim for a loss of consortium. The Ciceros asked for compensatory and other damages resulting from Thomas Cicero’s termination.
Claiming diversity jurisdiction under 28 U.S.C. § 1332(a)(l)(l) (2001), the defendants removed the case to the U.S. District Court for the Eastern District of Michigan. The defendants then filed a motion for summary judgment on the plaintiffs’ age discrimination and loss of consortium claims.
In deciding Borg-Warner’s motion for summary judgment, the district court used the familiar McDonnell DouglasBurdine tripartite test. See McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973), later clarified by Tex. Dep’t of Cmty. Affairs v. Burdine, 450 U.S. 248, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981). Under the first stage [583] of that test, the plaintiff must establish a prima facie case of discrimination. See McDonnell Douglas, 411 U.S. at 802, 93 S.Ct. at 1817. To establish a prima facie case of discrimination, a plaintiff must show that (1) he is a member of a protected class, (2) he was qualified for his job and did it satisfactorily, (3) despite his qualifications and performance, he suffered an adverse employment action, and (4) that he was replaced by a person outside the protected class or was treated less favorably than a similarly situated individual outside his protected class. See id.; see also Johnson v. Univ. of Cincinnati, 215 F.3d 561, 572-73 (6th Cir.2000); Mitchell v. Toledo Hosp., 964 F.2d 577, 582 (6th Cir.1992).
The district court granted the defendants’ motion for summary judgment. In giving Borg-Warner judgment, the district court found that Cicero did not make out a prima facie case because he did not show he was qualified for his position.
The district court declined to decide whether Cicero’s replacement was substantially younger, an alternative for showing Borg-Warner replaced him with a person outside the protected class. In dicta, the district court then discussed the remaining stages of the McDonnell Douglas test. If Cicero had established a prima facie case, the district court found Borg-Warner had given a legitimate, nondiscriminatory reason for Cicero’s discharge. The district court then reasoned that Cicero did not show Borg-Warner’s reason for his firing was pretextual. Plaintiffs-Appellants appealed the decision to this Court.
B. Factual Background
On August 1, 1994, Cicero began working for Federal Mogul as the human resources manager for its Precision Forged Products Division (“Forged Products Division”). In April 1995, Borg-Warner bought the Forged Products Division from Federal Mogul and kept Federal Mogul’s management staff, including Cicero. On November 20, 1997, Borg-Warner fired Cicero and several other members of the management staff. Cicero says that Borg-Warner fired him because of his age.
Borg-Warner says that it fired Cicero for reasons unrelated to his age. To scrutinize this argument, we first look to the factual background of Cicero’s employment. Because we review decisions of summary judgment de novo, and summary judgment requires that all reasonable inferences be drawn in the light most favorable to the nonmoving party, we construe the facts in the light most favorable to the Plaintiffs-Appellants.
III. Discussion
A. Standard of Review
We review a district court’s grant of summary judgment de novo. See Doren v. Battle Creek Health Sys., 187 F.3d 595, 597 (6th Cir.1999). To decide whether summary judgment is appropriate, this Court applies the same legal standards as the district court. See Cline v. Catholic Diocese of Toledo, 206 F.3d 651, 657 (6th Cir.2000).
Summary judgment is only appropriate when the evidence submitted shows “that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(c). Ultimately the Court must decide “whether the evidence presents sufficient disagreement to require submission to a jury or whether it is so one-sided that one party must prevail as a matter of law.” Terry Barr Sales Agency, Inc. v. All-Lock Co., 96 F.3d 174, 178 (6th Cir.1996) (internal quotation marks omitted).
[584] B. The McDonnell Douglas Test
Plaintiff Cicero brought this age discrimination claim under Michigan’s Elliott-Larson Civil Rights Act and under Michigan common law. The U.S. Supreme Court case McDonnell Douglas Corp. v. Green provides the standard by which to analyze a discrimination claim under the Elliott-Larsen Act. See, e.g., Tinker v. Sears, Roebuck & Co., 127 F.3d 519, 522 (6th Cir.1997) (applying the McDonnell Douglas elements to an age discrimination claim brought under the Elliott-Larsen Act); Lytle v. Malady, 458 Mich. 153, 172-73, 173 n. 19, 579 N.W.2d 906, 914-15, 915 n. 19 (1998) (noting the Supreme Court of Michigan’s adoption of the McDonnell Douglas test for age discrimination).
The McDonnell Douglas Court established a three-stage process for analyzing discrimination claims. First, the plaintiff must show a prima facie case of discrimination, which gives rise to a presumption of discrimination. Second, the burden shifts to the defendant to rebut the presumption by offering a legitimate, nondiscriminatory reason for the plaintiffs discharge. Third, the plaintiff must then show the defendant’s proffered reason for discharge is a pretext for discrimination. See McDonnell Douglas, 411 U.S. at 802-04, 93 S.Ct. 1817; Lytle, 458 Mich. at 172-74, 579 N.W.2d at 914-15.
The district court granted the defendants’ motion for summary judgment after it found that Cicero did not establish a prima facie case of age discrimination. Additionally, the district court noted that Cicero likely did not show that the defendants’ reason for firing him was a pretext for discrimination. We address these issues below.
1. The Prima Facie Case
In granting the defendants’ motion for summary judgment, the district court found that Cicero did not establish a prima facie case of discrimination. With that reasoning, the district court conflated the stages of the McDonnell Douglas analysis and- impermissibly increased Cicero’s pri-ma facie burden. The prima facie requirement for making a discrimination claim “is not onerous” and “poses a burden easily met.” Cline, 206 F.3d at 660 (internal quotation marks omitted). Cicero met this requirement.
To establish a prima facie case of age discrimination, the plaintiff must show that (1) he was a member of a protected class, (2) who suffered an adverse employment action, (3) was qualified for the position, and (4) was replaced by a younger person. See Tinker, 127 F.3d at 522 (citing Matras v. Amoco Oil Co., 424 Mich. 675, 683, 385 N.W.2d 586 (1986)); Town v. Mich. Bell Tel. Co., 455 Mich. 688, 695, 568 N.W.2d 64, 68 (1997); Lytle, 458 Mich, at 172, 579 N.W.2d at 914.
Cicero clearly satisfies the first two elements of the prima facie case. He was within the protected class, employees age forty and over. At the time that Borg-Warner fired him, Cicero was fifty-one. Furthermore, his termination is an adverse employment action.
The defendants said that Plaintiff Cicero failed to satisfy both the third and the fourth elements of the prima facie case, arguing that Cicero was not qualified for his position and that a “substantially younger” person did not replace him. The district court noted that courts should evaluate whether an employee’s replacement is “substantially younger” on a case-by-case basis, but the district court expressly declined to decide this issue. Instead, the district court based its finding that Cicero did not establish a prima facie case on Cicero’s failure to show that he was qualified for his position.
[585] We find that Cicero offered enough evidence on the third and fourth prongs of the prima facie case to survive summary-judgment. He offered sufficient evidence to show he was qualified for his position. Additionally, though the district court did not decide this issue, we determine that the age difference between Cicero and his replacement provides enough evidence to allow a reasonable jury to find that Cicero made out the prima facie case.
a. Qualified for the Position
“The court first determines if a plaintiff has put forth sufficient evidence for a reasonable jury to find [him] to have met the prima facie requirements, including whether [he] has met the legitimate expectations of [his] employer.” Cline, 206 F.3d at 661.
In Towm, the Supreme Court of Michigan set out the standard for determining whether an employee is qualified for his position. The Town court stated, “An employee is qualified if he was performing his job at a level that met his employer’s legitimate expectations.” Town, 455 Mich. at 699, 568 N.W.2d at 69. The court further states, “ ‘To establish that he was qualified a complainant must show that he was doing his job well enough to rule out the possibility that he was fired for inadequate job performance, absolute or relative.’ ” Id. at 699 n. 22, 568 N.W.2d at 70 n. 22 (quoting Menard v. First Sec. Servs. Corp., 848 F.2d 281, 285 (1st Cir.1988) (internal quotation marks omitted)).
A court must evaluate whether a plaintiff established his qualifications independent of the employer’s proffered nondiscriminatory reasons for discharge. See Cline, 206 F.3d at 660-61 (“[W]hen assessing whether a plaintiff has met her employer’s legitimate expectations at the pri-ma facie stage ... a court must examine plaintiffs’ evidence independent of the nondiscriminatory reason ‘produced’ by the defense as its reason for terminating plaintiff.”). In short, a court must be careful not to conflate the distinct stages of the McDonnell Douglas test. The district court here did just that.
As we describe below, the district court erred when it failed to judge Cicero’s qualifications independent of the reason that Borg-Warner now offers for firing Cicero. Rather, the district court judged Cicero’s qualifications by using Borg-Warner’s discharge justification as evidence that Cicero was not qualified. Instead of using the discharge justification to judge whether Cicero was qualified at the prima facie stage, the Court should only consider the employer’s proffered reason in the later stages of the McDonnell Douglas analysis.
When viewed independently of Borg-Warner’s proffered reason, Cicero offers sufficient evidence for a reasonable jury to find that he was qualified for his position. Before working for Borg-Warner, Cicero worked as the human resources manager of Federal Mogul’s Forged Products Division. He was responsible for human resources management at all three of Federal Mogul’s Forged Products Division plants.2 Larry Finnell, then the general manager for Federal Mogul, supervised Cicero. Throughout his employment with Federal Mogul, and later with Borg-Warner, Cicero reported to general manager Finnell.
Even before his employment with Federal Mogul, Cicero had an extensive background in human resources management. Before Federal Mogul hired him, the Detroit Newspaper Agency employed Cicero [586] as its director of labor relations. Before that position, Cicero worked for the Detroit Free Press. For the Detroit Free Press, Cicero began as personnel manager, advanced to labor relations manager, and eventually became the director of employee relations.
As described, Cicero worked for Federal Mogul immediately before working for Borg-Warner. During this employment, Federal Mogul gave Cicero good evaluations. For example, in December 1994, Federal Mogul awarded Cicero a $14,040 merit bonus, and general manager Finnell praised Cicero’s performance. In reviewing Cicero’s work at Federal Mogul, general manager Finnell stated: “Tom has been with us for only a few short months. During his short tenure, he has come up to speed quickly and supported our needs admirably. Tom is a welcomed addition to our team.”
While prior work history is not probative at the second and third stages of the McDonnell Douglas test, common sense dictates that it is relevant at the prima facie stage for determining whether an employee has at least the minimum attributes needed to perform the position. Employers regularly consider a potential employee’s prior work experience to decide whether that individual is qualified for a position.
Beyond Cicero’s prior work experience, his performance at Borg-Warner shows that he met his employer’s legitimate expectations. After Borg-Warner purchased Federal Mogul in April 1995, Cicero became the division human resources manager for Borg-Warner. Cicero consistently received merit and Management Incentive Plan (“MIP”) bonuses from Borg-Warner. The purpose of the MIP bonus policy, as stated in the policy language itself, was to help “Borg-Warner Automotive in attracting and retaining ... individuals of outstanding competence” and to provide “performance incentives for officers, executives and other key employees of the Company and its subsidiaries.” Borg-Warner used merit increases to reward employee performance as well.
In December 1995, Borg-Warner gave Cicero a $15,840 MIP bonus. Borg-Warner also gave Cicero a 4.27 percent merit increase for the 1995 performance year. For the 1996 performance year, Borg-Warner again gave Cicero both the MIP bonus and a merit increase. Like the MIP bonus policy, the merit increase policy included guidelines explaining what the various increases meant. Under those guidelines, Cicero’s 3.7 percent increase for the 1996 performance year indicated Cicero’s performance “exceeded job requirements.”
Not only did the defendants continue to award Cicero MIP bonuses and merit increases while he was an employee, Borg-Warner management praised Cicero’s performance several times. For the 1995 performance year, general manager Finnell gave Cicero a positive performance evaluation. Although Finnell ranked Cicero lower in one category, he gave Cicero an overall score of “6.” This score suggests that Cicero exceeded job requirements.
In 1996, general manager Finnell again praised Cicero’s performance. In a June 27, 1996 letter to Cicero, Finnell said that he considered Cicero’s performance in the Gallipolis negotiations to be “outstanding” and worth “sineerest gratitude for a job well done.” Borg Warner president Robert Welding echoed Finnell’s sentiments in an email sent a day after the letter. Welding’s email to all those involved in the Gallipolis negotiations stated, “Nice work guys! I appreciate all the time and effort you put into this and of gaining the right outcome.”
[587] Cicero shows other evidence that he did his work satisfactorily. In October 1997, Borg-Warner circulated a memo describing the Romulus Connecting Rod Business Critical Employee Retention Plan. The memo said the plan’s purpose was to encourage employees to maintain employment during the period in which the business was for sale. The plan included a list of the employees Borg-Warner wanted to retain during the anticipated sale. It identified Cicero as a key member of the management staff whom the company wanted to keep through the transition period.
The Borg-Warner management consistently awarded bonuses to Cicero. Members of the management team regularly praised Cicero’s performance. Further, Borg-Warner targeted Cicero as a key member of the management staff that it sought to retain during the sale of the connecting rod business. This, coupled with the extensive experience Cicero had at the time he started with Federal Mogul and Borg-Warner, shows that Cicero offered more than enough evidence to allow a reasonable jury to conclude that he was qualified for his position.
“[I]f reasonable minds could differ as to whether a preponderance of the evidence establishes the facts of a prima facie case, then a question of fact does remain, which the trier of fact will be called upon to answer.” St. Mary’s Honor Ctr. v. Hicks, 509 U.S. 502, 510, 113 S.Ct. 2742, 2748, 125 L.Ed.2d 407 (1993) (applying the stages of the McDonnell Douglas test). Cicero offered evidence tending to show he was qualified for his position, a fact the defendants attempt to dispute. While president Welding, Geraldine Kinsella, vice president of human resources for Borg-Warner Automotive, and Borg-Warner human resources vice president Keith McLeod now profess to unhappiness with Cicero from “day one,” they did not contemporaneously document any of their complaints. They never gave Cicero any warning. While a jury may accept their belated, never-before-expressed complaints about Cicero’s work, consideration of these complaints is not appropriate at the prima facie stage of the McDonnell Douglas test.
This Court in Cline cautioned against improperly considering the employer’s proffered reason for termination as a “predicate for finding [the plaintiff] to have failed to make a prima facie case.” Cline, 206 F.3d at 660. To do so sets the burden far too high. “[A] court must examine plaintiffs evidence independent of the nondiscriminatory reason ‘produced’ by the defense as its reason for terminating plaintiff.” Id. at 661. Here, the district court analyzed the defendants’ proffered evidence of Cicero’s poor performance as an element of the prima facie case, a fact shown by the district court’s own language. When discussing whether the defendants would have been able to rebut a prima facie case, the second stage of the McDonnell Douglas test, the district court noted, “As stated in Section I [discussing the prima facie case], defendants have offered ample justification for Mr. Cicero’s dismissal.”
The evidence Cicero offers is sufficient to satisfy his prima facie burden of proving his qualifications. If the parties dispute the facts establishing the prima facie case, however, a question of fact remains for the fact finder to decide. See Hicks, 509 U.S. at 510, 113 S.Ct. 2742; Cline, 206 F.3d at 662-663, 663 n. 7 (applying the reasoning of Hicks).
The defendants dispute Cicero’s qualifications, saying his poor performance made him unqualified. None of the arguments the defendants raise, however, are appropriately considered at the prima facie stage. Under the McDonnell Douglas test, “ ‘concerns about [a plaintiffs] perfor-[588] manee are more appropriately raised as part of the second and third steps of the ... scheme.’ ” Town, 455 Mich. at 708, 568 N.W.2d at 73 (quoting Clark v. Coats & Clark, Inc., 990 F.2d 1217, 1227 (11th Cir.1993)); see also Cline, 206 F.3d at 664 (noting with approval cases in which courts treat evidence of poor work performance as evidence of an employer’s legitimate, nondiscriminatory justification and not as prima facie evidence of qualification). Because we find more than sufficient evidence showing that Cicero was qualified, we find the district court erred.
b. Replaced by a Younger Person
As applied to age discrimination claims, the McDonnell Douglas test requires that an employee’s replacement be a younger person. The district court expressly declined to address the issue of whether Cicero’s replacement satisfied this element of the prima facie case. While the defendants assert that an employee’s replacement must be “substantially younger,” we have not adopted a bright-line rule defining what such a standard entails.
“On a motion for summary judgment, a district court considers whether there is sufficient evidence to create a genuine dispute at each stage of the McDonnell Douglas inquiry.” Cline, 206 F.3d at 661.
Here, Cicero was fifty-one at the time of his firing. His replacement was forty-three. A question of fact remains for a fact finder to resolve whether, under the circumstances of the present case, the seven and one-half year age difference satisfies the fourth prong of Cicero’s prima facie case. See, e.g., Hicks, 509 U.S. at 509-10, 509 n. 3, 113 S.Ct. 2742 (noting that if “reasonable minds could differ,” a fact finder “will be called upon” to decide whether “a preponderance of the evidence establishes the facts of a prima facie case”).
2. Rebuttal
A presumption of discrimination arises when a plaintiff establishes a prima facie case under the first stage of the McDonnell Douglas test. The second stage of the McDonnell Douglas test allows the defendant to rebut that presumption by offering proof of a legitimate, nondiscriminatory reason for termination. If the defendant offers such proof, the presumption falls away. Although the district court granted the defendant’s motion for summary judgment because it held Cicero failed to establish a prima facie case of discrimination, it noted that if Cicero had established such a case, the defendants would have rebutted it. We agree that Borg-Warner offered a legitimate, nondiscriminatory reason for Cicero’s firing.
The defendants argued that Cicero’s performance dissatisfied Borg-Warner management. Welding, Kinsella, and McLeod all claim dissatisfaction with Cicero from early in his employment. They offer examples of how he did not meet their legitimate expectations. They further note that we should discredit general manager Finnell’s positive evaluation of Cicero because Borg-Warner also fired Finnell.
The defendants offered affidavits and deposition testimony to show that Cicero’s performance dissatisfied various members of Borg-Warner management. They claim such dissatisfaction caused them to fire Cicero. Terminating an employee because he fails, to perform satisfactorily is a legitimate and nondiscriminatory reason to end his employment. Therefore, the defendants successfully offered a nondiscriminatory justification to rebut the presumption of discrimination that the pri-ma facie case raised.
[589]*5893. Pretext
The district court stated that even if the plaintiffs established a prima facie case, the plaintiffs would be unable to defeat the defendants’ proffered reason for termination. We find that the district court erred in reaching such a conclusion. We find that Cicero offered enough evidence to create a genuine issue of material fact that the defendants’ reason for firing him was a pretext for discrimination.
The Supreme Court of Michigan and most federal courts of appeals follow the “intermediate position” for determining “the proper summary judgment disposition standard for employment discrimination claims under Michigan’s Civil Rights Act.” Lytle, 458 Mich, at 175, 579 N.W.2d at 915-16. To raise a genuine issue of fact as to pretext and defeat a summary judgment motion under this position, Cicero must show one of the following: “(1) that the proffered reason[ ] had no basis in fact, (2) that the proffered reason[ ] did not actually motivate the action, or (8) that the proffered reason[ was] insufficient to motivate the action.” Nelson v. Gen. Elec. Co., 2 Fed.Appx. 425, 430, 2001 WL 69201 (6th Cir.2001) (unpublished); Tinker, 127 F.3d at 523; Manzer v. Diamond Shamrock Chems. Co., 29 F.3d 1078, 1084 (6th Cir.1994).
Under the first and third methods of showing pretext, the fact finder may infer discrimination from the circumstances. See Kline v. Tenn. Valley Auth., 128 F.3d 337, 346 (6th Cir.1997) (“[W]hen a plaintiff proves that the defendant’s proffered reasons either have no basis in fact or are insufficient to motivate discharge, a permissive inference of discrimination arises.”); Manzer, 29 F.3d at 1084. Under the second method, Cicero may not rely exclusively on his prima facie evidence, but instead must introduce some further evidence of discrimination.3 See Reeves v. Sanderson Plumbing Prods., Inc., 530 U.S. 133, 143, 120 S.Ct. 2097, 147 L.Ed.2d 105 (2000) (allowing the fact finder to consider the plaintiffs prima facie evidence when evaluating if the defendant’s proffered reason was a pretext); Kline, 128 F.3d at 346 (noting that “when the reasons offered by the defendant did not actually motivate the discharge,” the plaintiff must introduce “additional evidence of discrimination”); Manzer, 29 F.3d at 1084 (noting that the plaintiff “may not simply rely on his prima facie evidence”).
Under any of the three methods used to show pretext, Cicero offers sufficient evidence to permit a reasonable fact finder to conclude that age discrimination motivated Borg-Warner’s decision to fire Cicero.
Borg-Warner says it fired Cicero for poor work performance. Cicero offers evidence that raises an issue regarding the factual basis of the defendants’ proffered reason. The evidence of Cicero’s qualifications, coupled with a lack of contemporaneous criticism of his performance and Borg-Warner’s continued grant of bonuses, could allow a fact finder to find that Borg-Warner’s proffered reason either had no basis in fact or that it was insufficient to motivate the discharge decision.
In April 1995, Borg-Warner bought the Forged Products Division from Federal [590] Mogul. After the purchase, Borg-Warner kept all of Federal Mogul’s Forged Products Division employees, including the management staff. Cicero became the division human resources manager for Borg-Warner. In that position, he reported directly to general manager Finnell and indirectly to McLeod, vice president of human resources for Borg-Warner.
Borg-Warner says it was justified in firing Cicero because he inadequately coordinated Forged Products Division benefits with Borg-Warner benefits. After the acquisition, Borg-Warner told Cicero to integrate Borg-Warner policies into the new organization. Cicero’s key objectives were to coordinate Borg-Warner’s pension plan, post-retirement benefits, employee insurance benefits, salaried bonus plan, non-hourly salary plan, and automobile allowance plan. Borg-Warner argues that, in part, it fired Cicero because he failed to carry out this integration adequately.
As another justification for firing Cicero, Borg-Warner says Cicero hired outside legal counsel without prior approval. During Cicero’s employment, Borg-Warner gave Cicero a memorandum explaining the policies of the Borg-Warner corporate legal department regarding the use of outside counsel. The June 23, 1995 memo recommended using the Borg-Warner legal department rather than outside counsel.
Borg-Warner argued that Cicero’s failure to follow Borg-Warner’s policy regarding outside counsel justified its decision to fire him. In August 1995, Borg-Warner was sued. Cicero hired outside counsel to represent the company. Though the Borg-Warner legal department allowed him to do so for that case, it stated that it would monitor the proceedings. A year later, in August 1996, the legal department sent out another memo mandating the notification and use of Borg-Warner legal counsel in certain situations. In October 1997, Cicero again hired outside counsel to represent the company in an action seeking injunctive relief, this time with neither prior nor subsequent approval.
Borg-Warner also argued that it fired Cicero, in part, because of his handling of labor negotiations at the Gallipolis plant. In spring 1996, Cicero took part in the Gallipolis plant labor negotiations. Though Kinsella now says that Cicero’s performance dissatisfied her, at the time Borg-Warner praised his work on the negotiations. A June 27, 1996 letter from general manager Finnell to Cicero said, “[Tjhank you for what I consider to be an outstanding job in bringing the Gallipolis negotiations to a fruitful conclusion.... Would you please let me ... express my sincerest gratitude for a job well done.” The president of Borg-Warner gave Cicero similar praise for his work. On June 28, 1996, Borg-Warner president Welding sent Cicero and those involved in the Gal-lipolis negotiations an email praising their “nice work.”
While Borg-Warner said it fired Cicero for inadequacies, it gave him many bonuses during the same period. Borg-Warner rewarded the performance of its employees with two types of performance-based bonuses: MIP bonuses and merit increases. The defendants deny that they used the bonus plans to reward individual performance. Instead, the defendants say the bonus plans rewarded team or unit performance for reaching certain financial goals. The terms of the bonus plans suggest otherwise. The bonus plans say that the bonuses reward individual performance, and Borg-Warner could deny participants the MIP bonus if they “otherwise fail to perform satisfactorily.”4
[591] While Borg-Warner generally paid the MIP bonus to participants, it denied the bonus to an employee in December 1995. Because of poor performance, Borg-Warner gave John LeBlanc only 50 percent of his MIP bonus. With LeBlanc, Borg-Warner also warned him that it would fire him if his performance did not improve. Borg-Warner fired LeBlanc in April 1996.
While LeBlanc’s poor performance netted him only part of his MIP bonus, Borg-Warner paid Cicero his complete bonus. In a December 1995 memo, Gary Fukaya-ma5 authorized payment of a $15,840 MIP bonus to Cicero. In December 1995, Borg-Warner also gave Cicero a 4.27 percent merit increase. Borg-Warner paid the grant in a lump sum payment of $4000.
After the Gallipolis negotiations, Borg-Warner continued giving bonuses to Cicero. For the 1996 performance year, Borg-Warner again gave Cicero both the MIP bonus and a merit increase. Though in February 1997, Welding says that he knew Borg-Warner would fire the management team, including Cicero, because “they were not the kind of people Borg-Warner needed for the future,” Welding nevertheless authorized the payment of the bonuses to Cicero that same month.
In October 1997, a memo circulated describing the Romulus Connecting Rod Business Critical Employee Retention Plan. The memo said the plan’s purpose was to encourage employees to maintain employment during the period in which the business was for sale. The memo included Cicero as a key member of the management staff whom the company wanted to retain through the transition period.*
On November 20, 1997, Borg-Warner fired Cicero, Finnell, James Knaus, and Gerald Malen. Shortly after that in December 1997, Borg-Warner transferred Karen Schumann out of the Romulus plant. A few weeks later, Borg-Warner fired Nick Bruno, a plant manager. All of the employees that Borg-Warner discharged were age fifty-one or older.6 Borg-Warner replaced Cicero, Finnell, Knaus, and Malen with younger employees.7
Cicero offers evidence showing that Borg-Warner’s proffered reason for firing him was false. Borg-Warner says that it fired Cicero because of his poor performance, however, Borg-Warner shows no contemporaneous evidence that any failures by Cicero caused it concern. While Borg-Warner now says that Cicero failed in several ways throughout his employment, it never raised any serious complaints about his performance until after it [592] fired him. Until the time of his discharge, Borg-Warner continued to praise Cicero’s work and award him performance-based bonuses.
Not only does Cicero offer evidence showing that Borg-Warner’s reason for firing him was likely false, he also offers evidence that the falsity masked discrimi-natoryanimus. It was not until after Borg-Warner fired him that it raised serious complaints about his performance, and even then, it gave shifting justifications for his discharge.
“An employer’s changing rationale for making an adverse employment decision can be evidence of pretext.” Thurman v. Yellow Freight Sys., 90 F.3d 1160, 1167 (6th Cir.1996) am. on other grounds, 97 F.3d 833 (6th Cir.1997); see also Edwards v. U.S. Postal Serv., 909 F.2d 320, 324 (8th Cir.1990); Schmitz v. St. Regis Paper Co., 811 F.2d 131, 132-33 (2d Cir.1987). Shifting justifications over time calls the credibility of those justifications into question. By showing that the defendants’ justification for firing him changed over time, Cicero shows a genuine issue of fact that the defendants’ proffered reason was not only false, but that the falsity was a pretext for discrimination.
At the time Borg-Warner management fired Cicero, Finnell, Knaus, and Malen, John Fiedler, president of Borg-Warner Automotive, said Borg-Warner was replacing them to provide the Forged Products Division with a “Borg-Warner team.” Fiedler never said Borg-Warner discharged Cicero for poor work performance.
In answering interrogatories, Borg-Warner for the first time said that it fired Cicero because of poor performance. At deposition, Welding testified that he fired the Forged Products division management team because Borg-Warner failed to sell the Romulus facility as planned. Welding defined the management team as those employees who reported directly to general manager Finnell and Finnell himself. Borg-Warner, however, did not fire all of the management team members. It also did not limit its firing only to the management team.8
During this Court’s consideration of the summary judgment motion, Borg-Warner again shifted its position. Borg-Warner then said that it did not fire the entire management team, only those members of the team who dealt directly with the company’s headquarters in Chicago.
While the Court does not question business decisions, the Court does question a defendant’s proffered justification when it shifts over time. When the justification for an adverse employment action changes during litigation, that inconsistency raises an issue whether the proffered reason truly motivated the defendants’ decision.
As further evidence that the defendants’ proffered reason for firing Cicero was a pretext and that the pretext masked a discriminatory motivation, Cicero offers statistical evidence about other discharged employees. Contrary to the district court’s analysis, Cicero does not offer the statistics as direct evidence of discrimination. Instead, Cicero offers the statistics as additional circumstantial evidence of discrimination. The cases that the district court relied upon regarding the plaintiffs’ statistical sample are inapposite and distin[593] guishable.9
This Court has previously noted that even a small statistical sample, though not as probative as it might otherwise be, can nevertheless serve as circumstantial evidence making discrimination more likely. See Scott v. Goodyear Tire & Rubber Co., 160 F.3d 1121, 1129 (6th Cir.1998); Kulling v. Grinders for Indus., Inc., 115 F.Supp.2d 828, 839 (E.D.Mich.2000) (noting the Sixth Circuit’s recognition that even small statistical samples can increase the likelihood that the “decisions to eliminate certain positions were based on age” (internal quotation marks and citations omitted)).
Cicero’s statistics offer some proof that the defendants had a pattern of hiring and firing that adversely affected older employees. This evidence coupled with the other circumstantial evidence that Cicero offers raises a triable issue of fact that the defendants’ proffered reason for dismissing Cicero was a pretext for discrimination. Therefore, the district court erred when granting the defendants’ summary judgment motion.
C. Loss of Consortium Claim
The district court summarily dismissed Marlene Cicero’s loss of consortium claim. The district court held Marlene Cicero’s claim to be derivative. After finding that Thomas Cicero’s claim failed, it gave Borg-Warner judgment on Marlene Cicero’s claim. Because we find that the plaintiffs’ age discrimination claim does not fail, we accordingly reverse the district court’s holding on the loss of consortium claim.
IV. Conclusion
Plaintiff Thomas Cicero showed sufficient evidence that he was qualified for his position and was performing satisfactorily to defeat summary judgment. He received merit increases, MIP bonuses, and written positive evaluations. The defendants counter Cicero’s argument that he was performing satisfactorily by offering affidavits and deposition testimony suggesting that from the beginning of his employment Cicero’s performance dissatisfied Borg-Warner management. The defendants, however, offer no written documentation of poor performance by Cicero. And Borg-Warner shows no contemporaneous warnings or significant criticism of Cicero’s work.
Cicero raises a genuine issue of material fact that the defendants’ reason for firing him is false and that the falsity masks a discriminatory intent. In his prima facie case, Cicero offers evidence showing that [594] he was doing his job consistent with his employer’s legitimate expectations and that Borg-Warner did not fire him for poor job performance. This evidence supports his argument that the defendants’ proffered reason for firing him was a pretext for discrimination. See Reeves, 530 U.S. at 143, 120 S.Ct. 2097 (allowing the use of prima facie evidence at the pretext stage).
Employing the proper summary judgment standard and resolving any factual disputes in the light most favorable to Cicero, the nonmoving party, we accordingly reverse the district court’s decision granting Borg-Warner’s summary judgment motion.