EDWARD C. PRADO, Circuit Judge:

Plaintiff-Appellee Eddie Wooten filed suit against Defendant-Appellant McDonald Transit Associates, Inc. under the Age Discrimination in Employment Act (ADEA), 29 U.S.C. §§ 621-634, alleging discrimination and retaliation. McDonald Transit never answered or defended the suit. The clerk entered default against McDonald Transit, and, after holding a hearing on damages in which Wooten provided live testimony, the district court entered default judgment for Wooten. McDonald Transit filed a motion to set aside the default judgment, which the district court denied. Wooten's complaint contained very few factual allegations, but his testimony at the damages hearing provided evidence on the elements of his claim that were absent from his pleadings. Accordingly, we confront the question we left open in Nishimatsu Construction Co., Ltd. v. Houston National Bank, 515 F.2d 1200 (5th Cir.1975), and conclude that evidence adduced at a default-judgment "prove-up" hearing cannot cure a deficient complaint. We therefore vacate the district court's entry of default judgment and remand the case to the district court.

I. BACKGROUND

On June 22, 2012, Wooten sued McDonald Transit in federal court, alleging discrimination on the basis of his age and retaliation after he made a claim of age discrimination. In his complaint, Wooten alleged that he was a former employee of McDonald Transit, where he had worked from 1999 until May 1, 2011. At the time he was fired, he worked as a Class B Mechanic. He further alleged:

Wooten's complaint contained no additional factual allegations.

The district clerk issued a summons the same day that Wooten filed his complaint. On July 18, 2012, Wooten returned the summons with an affidavit of service indicating that service had been executed on July 5 on McDonald Transit's president and registered agent, Robert T. Babbitt, by certified mail, return receipt requested. But the return receipt indicated that process had in fact been served on Brenda Roden, another McDonald Transit officer. After McDonald Transit failed to appear, plead, or otherwise defend Wooten's suit, the district clerk entered default against McDonald Transit on October 30, 2012, and Wooten moved for a default judgment.

The district court held a hearing on the motion in December 2012, but took issue with the fact that Roden, not Babbitt, had been served. The court adjourned the hearing so that Wooten could attempt proper service on Babbitt again. The district clerk issued new summons, and Wooten returned with a new affidavit of service indicating service had been executed by personal delivery on Babbitt on January 17, 2013. Again McDonald Transit failed to appear, answer, or defend; again the district clerk entered default; and again Wooten moved for a default judgment.

The district court held a hearing on whether to enter default judgment on June 7. At that hearing, which the court expressly designated "a hearing to prove up damages for a default judgment," Wooten provided testimony that elaborated on his factual allegations. He testified that he was born in January 1956, making him fifty-four years old at the time he made his claim to the EEOC. He explained that during his tenure at McDonald Transit, he had been promoted from the position of Class B Mechanic to the position of Shop Foreman. He also described his retaliation claim in greater detail: he stated that he was demoted from Shop Foreman, lowering his pay by $2 an hour. He was given menial work, and his hours were changed. He was denied opportunities for additional job-related certification. He stated that he "never got [written] up, [and was] never reprimanded." The district court entered a default judgment that same day.

McDonald Transit filed a motion to set aside the default judgment on June 18. In an affidavit accompanying the motion, Babbitt averred that he was never served with process, that he had not learned of the suit naming McDonald Transit as a defendant until June 11, and that he retained counsel to challenge the default judgment soon afterward.

McDonald Transit challenged the suit on numerous grounds under Federal Rules of Civil Procedure 55(c) and 60(b). In particular, McDonald Transit invoked Rule 60(b)(1) (mistake, inadvertence, surprise, or excusable neglect); (b)(3) (fraud, misrepresentation, or misconduct by an opposing party); (b)(4) (the judgment is void); and (b)(6) (any other reason that justifies relief). In asking the court to set aside the default judgment under Rule 60(b)(1), McDonald Transit claimed it had a few defenses to Wooten's suit — namely, McDonald Transit was not Wooten's employer; Wooten had failed to obtain a right-to-sue letter before suing McDonald Transit; and Wooten had failed to file suit within the required time from the issuance of a right to sue letter. McDonald Transit also asserted that it was not properly served and therefore had not willfully disregarded its duty to respond. In response, Wooten argued that McDonald Transit had failed to offer any explanation for its default, failed to produce sufficient evidence of a meritorious defense, and relied on Babbitt's uncorroborated and self-serving statements.

The district court denied McDonald Transit's motion to set aside the default judgment. Based on evidence of service of process to Babbitt and Roden (who the court had learned was a vice president of McDonald Transit), the court inferred that McDonald Transit had knowingly and intentionally failed to answer or otherwise defend against the complaint. The court further rejected McDonald Transit's claim to raise meritorious defenses on the grounds that the "record is far from conclusive" and these defenses were effectively waived by failing to answer the complaint.

McDonald Transit timely appealed both the default judgment and the order denying its motion to set aside the default judgment.

II. JURISDICTION AND STANDARD OF REVIEW

Wooten sued McDonald Transit for violations of federal law under the ADEA; accordingly, the district court had subject-matter jurisdiction under 28 U.S.C. § 1331. We have jurisdiction under 28 U.S.C. § 1291.

A district court must exercise its discretion in determining whether it should enter a default judgment under Rule 55(b)(2). Mason v. Lister, 562 F.2d 343, 345 (5th Cir.1977). We review the entry of a default judgment for abuse of discretion. U.S. for the Use of M-CO Constr., Inc. v. Shipco Gen., Inc., 814 F.2d 1011, 1013 (5th Cir.1987). Rule 55(c) provides that a district court "may set aside an entry of default for good cause" and "may set aside a default judgment under Rule 60(b)." We correspondingly review the district court's refusal to set aside a default judgment for abuse of discretion. Lacy v. Sitel Corp., 227 F.3d 290, 291-92 (5th Cir.2000); Shipco, 814 F.2d at 1013. "Any factual determinations underlying that decision are reviewed for clear error." Lacy, 227 F.3d at 292.

Yet, we undertake this review with a grain of salt. "`Because of the seriousness of a default judgment, and although the standard of review is abuse of discretion, even a slight abuse of discretion may justify reversal.'" In re Chinese-Manufactured Drywall Prods. Liab. Litig., 742 F.3d 576, 594 (5th Cir.2014) (quoting Lacy, 227 F.3d at 292). Review of a default judgment puts competing policy interests at play. On one hand, "[w]e have adopted a policy in favor of resolving cases on their merits and against the use of default judgments." Id. On the other, this policy is "counterbalanced by considerations of social goals, justice and expediency, a weighing process [that] lies largely within the domain of the trial judge's discretion." Id. (alteration in original) (quoting Rogers v. Hartford Life & Accident Ins. Co., 167 F.3d 933, 936 (5th Cir.1999)) (internal quotation marks omitted).

III. DISCUSSION

On appeal, McDonald Transit raises two principal issues: (1) whether the district court erred in entering a default judgment and (2) whether the district court erred in denying McDonald Transit's motion to set aside the judgment.1 Because we only need to answer the first question, we do not reach the second.

"A default judgment is unassailable on the merits but only so far as it is supported by well-pleaded allegations, assumed to be true." Nishimatsu Constr. Co. v. Hous. Nat'l Bank, 515 F.2d 1200, 1206 (5th Cir.1975) (citing Thomson v. Wooster, 114 U.S. 104, 113, 5 S.Ct. 788, 29 L.Ed. 105 (1885)). Put another way, "a defendant's default does not in itself warrant the court in entering a default judgment. There must be a sufficient basis in the pleadings for the judgment entered." Id. In sum, "[t]he defendant is not held to admit facts that are not well-pleaded or to admit conclusions of law.... On appeal, the defendant, although he may not challenge the sufficiency of the evidence, is entitled to contest the sufficiency of the complaint and its allegations to support the judgment." Id. In addition, a court "may conduct hearings ... when, to enter or effectuate judgment," it needs to, inter alia, "establish the truth of any allegation by evidence ... or ... investigate any other matter." Fed.R.Civ.P. 55(b)(2)(C); see also 10A Charles A. Wright et al., Federal Practice & Procedure § 2688 (3d ed. 1998) ("[W]hen it seems advantageous, a court may conduct a hearing to determine whether to enter a judgment by default.... [T]he court, in its discretion, may require some proof of the facts that must be established in order to determine liability.").

Neither party disputes that entry of default was appropriate. The parties disagree about (1) the ADEA standard that governs Wooten's claim; (2) whether the district court can consider evidence presented at the hearing in addition to the allegations in supporting default judgment; and (3) the sufficiency of Wooten's allegations.

A. ADEA Framework

McDonald Transit asserts that Wooten's complaint insufficiently alleged the essential elements of his prima facie retaliation claim under the ADEA — in particular, membership in a protected class and qualification. Wooten contests both elements.

The ADEA makes it unlawful "for an employer to discriminate against any of his employees ... because such individual ... has opposed any practice made unlawful by this section, or because such individual ... has made a charge, testified, assisted, or participated in any manner in an investigation, proceeding, or litigation under this Act." 29 U.S.C. § 623(d). The framework for establishing a prima facie retaliation claim under the ADEA is straightforward. A plaintiff "must show (1) that he engaged in a protected activity, (2) that there was an adverse employment action, and (3) that a causal link existed between the protected activity and the adverse employment action." Holtzclaw v. DSC Commc'ns Corp., 255 F.3d 254, 259 (5th Cir.2001). Holtzclaw unequivocally added a fourth element to the claim — a plaintiff who sought re-employment under the ADEA must prove as a part of his prima facie case that he was qualified for his position. Id.2

Contrary to McDonald Transit's position, neither § 623(d) nor Holtzclaw requires that Wooten prove he was a member of a class protected by the ADEA discrimination provisions. Compare 29 U.S.C. § 631 (defining class of individuals covered by age-discrimination provisions), with id. § 623(d) (permitting retaliation provision to apply to "any" employee). Therefore, the only elements that Wooten must sufficiently allege are (1) protected activity, (2) adverse employment action, (3) causal link, and (4) qualification.

B. Assessing the Entry of Default Judgment Based on Allegations and Testimony

We begin by determining whether Wooten's complaint, either standing alone or considered together with his testimony at the hearing, adequately states a claim upon which default judgment could properly be entered. We conclude that Wooten's complaint is impermissibly bare, but if viewed in combination with his live testimony, it provides a sufficient basis to support the default judgment against McDonald Transit. Correspondingly, we must address the question on which we reserved judgment in Nishimatsu: May fatally defective pleadings be corrected by proof taken at a default-judgment hearing? 515 F.2d at 1206 n. 5. We answer this matter of first impression in the negative and therefore conclude that the district court erred in entering default judgment on Wooten's deficient complaint.

1. Wooten's Prima Facie Case

Despite announcing that a default judgment must be "supported by well-pleaded allegations" and must have "a sufficient basis in the pleadings," the Nishimatsu court did not elaborate on these requirements. See 515 F.2d at 1206. Nothing in the record or the parties' briefs discusses how to determine what is "well-pleaded" or "sufficient," and we have found no guidance in our own cases. The Second Circuit has said that "so long as the facts as painted by the complaint `might [...] have been the case' they may not now be successfully controverted [following entry of default]." Trans World Airlines, Inc. v. Hughes, 449 F.2d 51, 64 (2d Cir.1971) (first alteration in original), rev'd on other grounds sub nom. Hughes Tool Co. v. Trans World Airlines, Inc., 409 U.S. 363, 93 S.Ct. 647, 34 L.Ed.2d 577 (1973). In comparison, the Ninth Circuit has held that factual allegations are not well-pleaded when they "parrot the language" of the statute creating liability. DirecTV, Inc. v. Hoa Huynh, 503 F.3d 847, 854 (9th Cir. 2007).

Moreover, whether a factual allegation is well-pleaded arguably follows the familiar analysis used to evaluate motions to dismiss under Rule 12(b)(6). In that context, the Supreme Court has reiterated that "[t]hreadbare recitals of the elements of a cause of action, supported by mere conclusory statements," are not sufficient. Ashcroft v. Iqbal, 556 U.S. 662, 678, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009) (citing Bell Atl. Corp. v. Twombly, 550 U.S. 544, 555, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007)). See generally 2 James Wm. Moore et al., Moore's Federal Practice § 8.04[1][e]-[f] (3d ed.2014).

Wooten's complaint contains the following factual allegations. (1) Wooten is a former employee of McDonald Transit. (2) Wooten was employed by McDonald Transit from 1999 until May 1, 2011. At the time he was fired, Wooten was a Class B mechanic earning $19.50 per hour, plus benefits. (3) In October 2010, Wooten filed an age-discrimination claim with the EEOC, after which McDonald Transit "discriminated and retaliated against [Wooten], and created a hostile work environment, until such time that [Wooten] was constructively discharged on or about May 1, 2011." (4) McDonald Transit's unlawful conduct has caused Wooten harm, including damages in the form of lost wages and benefits, mental anguish, and non-economic damages.

We hold that Wooten's complaint, standing alone, fails to meet either the Rule 12(b)(6) "plausibility" standard or the broadly similar standards announced by the Second and Ninth Circuits. His few factual allegations are inextricably bound up with legal conclusions (e.g., "discriminated and retaliated" and "created a hostile work environment" leading to "constructive[] discharge"). Read in its entirety, the complaint merely "parrot[s] the language" of the ADEA, Hoa Huynh, 503 F.3d at 854, and comprises a "[t]hreadbare recital[] of the elements of a cause of action, supported by mere conclusory statements," Iqbal, 556 U.S. at 678, 129 S.Ct. 1937. Indeed, given the lack of detail in the complaint, we cannot even speculate as to whether the facts "might [...] have been the case," Hughes, 449 F.2d at 64 (alteration in original). Furthermore, the complaint makes — at best — only an indirect, inferential allegation of a causal link and fails to address altogether the qualification element required by Holtzclaw.

In view of the above, we conclude that Wooten's complaint is not "well-pleaded" for default-judgment purposes.3 However, we also conclude that Wooten's live testimony provides sufficient evidence of each of the elements of his ADEA cause of action to support the entry of default. We discuss each pertinent element in turn.4

a. Adverse Employment Action

In his testimony, Wooten described a variety of concrete actions (changes in hours, reduced responsibilities, demotion, preclusion from job-related certification) that together might amount to an adverse employment action. To establish that he suffered an adverse employment action, Wooten must show that "a reasonable employee would have found the challenged action materially adverse, which in this context means it well might have dissuaded a reasonable worker from making or supporting a charge of discrimination." Burlington N. & Santa Fe Ry. Co. v. White, 548 U.S. 53, 68, 126 S.Ct. 2405, 165 L.Ed.2d 345 (2006) (citation and internal quotation marks omitted). Wooten's undisputed testimony could support a finding that these actions were materially adverse.

b. Causal Link

McDonald Transit contends that nothing in the complaint nor in Wooten's testimony shows a causal link between the October 2010 claim and the subsequent alleged adverse employment action. Wooten argues that merely showing a tight sequence of events — as his hearing testimony does — can be enough to make a prima facie showing of a causal link.

In Clark County School District v. Breeden, 532 U.S. 268, 121 S.Ct. 1508, 149 L.Ed.2d 509 (2001), the Supreme Court made clear that a narrow band of retaliation claims can establish causation by the "very close" temporal proximity alone. Id. at 273, 121 S.Ct. 1508; accord Strong v. Univ. Healthcare Sys., L.L.C., 482 F.3d 802, 808 (5th Cir.2007). Breeden cites cases in which courts found a three-or four-month lapse in time insufficiently close to establish a causal link where the proof was based on temporal proximity alone. 532 U.S. at 273-74, 121 S.Ct. 1508. Here, though, the delay between filing and adverse treatment is far shorter. Wooten claims that the adverse treatment he received lasted for about six months. Since there was only a seven-and-a-half-month window between the date Wooten filed his EEOC claim and the date he resigned, the adverse treatment must have begun within two months of his filing. Therefore, Breeden suggests that Wooten's testimony establishes a sufficient causal connection to support the default judgment.

c. Qualification

Finally, Wooten's testimony evinces that he was qualified for his job at the time of the adverse employment action. Although Holtzclaw made qualification for the job an additional requirement, it did not explain how to evaluate that requirement. See 255 F.3d at 260. Nevertheless, a review of our ADEA jurisprudence indicates that "qualified" has a broadly colloquial meaning in this context; it refers to objective job qualifications (e.g., training, experience, and physical capacity), not "essential functions" or any other term of art associated with the term's counterpart in the Americans with Disabilities Act.

Bienkowski v. American Airlines, Inc., 851 F.2d 1503 (5th Cir.1988), appears to be our sole case expressly defining "qualification" for ADEA purposes.5 In Bienkowski, we rejected the proposition that an employee is not "qualified" under the ADEA if his supervisors are dissatisfied with his work (i.e., if the employee does not "me[e]t his employer's legitimate expectations"). 851 F.2d at 1505. Instead, we concluded that "a plaintiff challenging his termination or demotion [under the ADEA] can ordinarily establish a prima facie case of age discrimination by showing that he continued to possess the necessary qualifications for his job at the time of the adverse action." Id. at 1506. We then explained: "By this we mean that plaintiff had not suffered physical disability or loss of a necessary professional license or some other occurrence that rendered him unfit for the position for which he was hired." Id. at 1506 n. 3.

The cases in this circuit following Bienkowski have accordingly focused on objective job qualifications when assessing the "qualified" element of ADEA discrimination and retaliation claims. See, e.g., Berquist v. Wash. Mut. Bank, 500 F.3d 344, 349-50 (5th Cir.2007) (concluding that the plaintiff had made out a prima facie case of ADEA discrimination by showing that he "possessed the same job qualifications when [his employer] terminated him as when [it] assigned him to [his last] position," as evidenced by the employee meeting "the objective criteria listed in a job posting" and holding a similar job title for two years); Fullen v. Galveston Indep. Sch. Dist., 564 F.Supp.2d 719, 730-31, 732-34 (S.D.Tex.2008) (reasoning that the plaintiff could not establish a prima facie case of ADEA retaliation because "he was not objectively qualified to continue serving" in his position due to his failure to complete required training); cf. Tratree v. BP N. Am. Pipelines, Inc., 277 Fed. Appx. 390, 395 (5th Cir.2008) (per curiam) (determining that the plaintiff had failed to state a claim for failure to promote under the ADEA because "by his own admissions he had not completed the on-the-job training necessary to become eligible" for the position he sought, thereby precluding a showing of qualification).

Likewise, this Court has found a relatively spare ADEA complaint sufficient to survive a Rule 12(b)(6) motion to dismiss when it contains factual allegations evidencing the plaintiff's experience, promotions, and commendations. See Leal v. McHugh, 731 F.3d 405, 413 (5th Cir.2013). In Leal, we held that the federal-employee plaintiffs' ADEA-discrimination claims could withstand a motion to dismiss despite their barebones complaint. Id. The complaint never explicitly asserted that the plaintiffs were "qualified" for the positions they sought. First Amended Complaint at 1-2, Leal v. McHugh, No. 2:11-cv-00249, 2011 WL 6372820 (S.D.Tex. Dec. 20, 2011), ECF No. 11. But it indicated that one plaintiff had over twenty years of experience in similar positions and was recognized as "the most productive employee" in his division, while the other plaintiff had earned a promotion and had received top ratings on nearly every performance evaluation for nineteen years. Leal, 731 F.3d at 408. Acknowledging that the complaint contained "few facts," we nonetheless held that the plaintiffs successfully alleged qualification and that "the[ir] admittedly bare allegations sufficiently state[d] a plausible claim for age discrimination to survive a motion to dismiss." Id. at 413.

Here, Wooten testified about his twelve years of experience at McDonald Transit, his promotion to the position of Shop Foreman, and his positive — or at least neutral — work evaluations while in that position. There is no evidence that he "suffered physical disability or loss of a necessary professional license or some other occurrence that rendered him unfit for the position for which he was hired," Bienkowski, 851 F.2d at 1506 n. 3. To the contrary, his tenure, promotion, and clean performance record support the inference that he had the training and physical capacity required for the position he held. See Leal, 731 F.3d at 413. As a result, Wooten's testimony adequately establishes his qualification within the meaning of the ADEA.

Having determined that Wooten's testimony at the default-judgment prove-up hearing — and not his complaint, standing alone — presents a prima facie case of retaliation under the ADEA, we are now squarely presented with the question of whether such testimony can cure fatally deficient pleadings for the purpose of entering a default judgment.

2. The Effect of Hearing Testimony on a Deficient Complaint

Since reserving judgment in Nishimatsu on the "possibility that otherwise fatal defects in the pleadings might be corrected by proof taken by the court at a [default-judgment prove-up] hearing," 515 F.2d at 1206 n. 5, this Court has not subsequently confronted this issue. Upon consideration of the text of the Federal Rules, our precedents and those of our sister circuits, and the policies underlying our default-judgment jurisprudence, we hold that a defective complaint cannot be redeemed by evidence presented at a prove-up hearing and therefore cannot support a default judgment absent amendment of the pleadings.

We begin with the text of the rule governing default-judgment hearings — Federal Rule of Civil Procedure 55. Rule 55(b)(2) authorizes a court considering an application for default judgment to "conduct hearings ... when, to enter or effectuate judgment," it needs to, inter alia, "establish the truth of any allegation by evidence ... or ... investigate any other matter." Despite this expansive language, neither this circuit nor any other has squarely held that such a hearing would be appropriate to adduce facts necessary to state a claim that were absent from the pleading on which judgment was sought. Indeed, one would expect that if the text of the rule could be so construed, one of the courts of appeals would have endorsed this construction — and our Court in Nishimatsu would have had little reason to expressly avoid the question.6

As a matter of semantics, moreover, the text of the hearing provision presupposes valid allegations in the complaint. After all, to "establish the truth of any allegation," there must be an existing allegation to assess. And a court's authority to "investigate any other matter" in a default-judgment hearing is circumscribed by the stated purpose of the hearing — "to enter or effectuate judgment." As there can be no judgment absent competent pleadings, Nishimatsu, 515 F.2d at 1206, it strains the text of the rule to suppose that this investigatory power encompasses the adduction of facts necessary to render the pleadings competent in the first place.7

We next observe that while none of our sister circuits have addressed the precise issue before us, our precedents agree on the basic proposition that a default judgment must be founded on adequate pleadings. For example, in a case presenting facts broadly similar to those we confront today, the Ninth Circuit held that evidence presented at a default-judgment prove-up hearing could not cure a deficient complaint, at least where the defaulting party "had no opportunity to conduct discovery, and the court ... indicated that [the hearing] was limited to the issue of damages." Alan Neuman Prods., Inc. v. Albright, 862 F.2d 1388, 1393 (9th Cir.1988). This case provides limited guidance, as the opinion was conditioned in part on the RICO context in which it arose.8 Nevertheless, to the extent that the Ninth Circuit's ruling rests on the fundamental premise that the pleadings underlying a default judgment must state a cause of action, it finds ample support in the courts of appeal. See City of New York v. Mickalis Pawn Shop, LLC, 645 F.3d 114, 137 n. 23 (2d Cir.2011) ("Most of our sister circuits appear to have held expressly that a district court may not enter a default judgment unless the plaintiff's complaint states a valid facial claim for relief.").9

Finally, we note that the rule we adopt today serves the policies animating our default-judgment jurisprudence and avoids prudential obstacles inherent in the contrary rule Wooten's position assumes. As our cases make plain, default judgments raise fundamental concerns of fairness — both to the plaintiff, who is entitled to judgment when the defendant fails to heed the commands of the judicial system, and to the defendant, who is entitled not to be held liable on claims that are procedurally or substantively infirm. Indeed, it is this delicate balance that informs our general policy disfavoring default judgments. See, e.g., Harper Macleod Solicitors v. Keaty & Keaty, 260 F.3d 389, 393 (5th Cir.2001) ("Federal courts generally disfavor default judgments, preferring to resolve disputes according to their merits.") (citing, inter alia, Lindsey v. Prive Corp., 161 F.3d 886, 892 (5th Cir.1998)); Sun Bank of Ocala v. Pelican Homestead & Sav. Ass'n, 874 F.2d 274, 276 (5th Cir.1989) (per curiam) ("The Federal Rules of Civil Procedure are designed for the just, speedy, and inexpensive disposition of cases on their merits, not for the termination of litigation by procedural maneuver. Default judgments are a drastic remedy, not favored by the Federal Rules and resorted to by courts only in extreme situations." (footnote omitted)). At the same time, this policy "is counterbalanced by considerations of social goals, justice and expediency." Rogers, 167 F.3d at 936 (internal quotation marks omitted). A rule permitting a party to cure facially deficient pleadings through evidence introduced at a damages hearing would disturb the careful compromise our cases have struck between fairness, finality, and justice.

As explained above, it is well settled that default judgments must be based on competent pleadings. Recognizing an exception to this established rule in cases where necessary facts omitted from the complaint emerge for the first time at a damages hearing would inject uncertainty into this body of law and unfairly disadvantage defaulting defendants.

To illustrate why this is the case, it is helpful to conceptualize the acceptance of supplemental evidence as a de facto amendment to the complaint. This Court has long recognized that an amended complaint supersedes the original complaint and deprives it of all legal effect unless the former expressly refers to or adopts the latter. King v. Dogan, 31 F.3d 344, 346 (5th Cir.1994) (per curiam); Boelens v. Redman Homes, Inc., 759 F.2d 504, 508 (5th Cir.1985). Correspondingly, "district courts routinely set aside entries of defaults when plaintiffs file amended complaints." Freilich v. Green Energy Res., Inc., 297 F.R.D. 277, 283 (W.D.Tex.2014) (citing cases). This approach is consistent both with the basic idea that default judgment may be entered only on an existing, valid complaint and with the general principles of notice and fairness enshrined in the Federal Rules governing pleadings and final judgments.

For example, it is widely accepted that "when [a] complaint is amended [the] defendant should be entitled to amend the answer to meet the contents of the new complaint." 6 Wright et al., supra, § 1476. Similarly, Federal Rule of Civil Procedure 5(a)(2) provides that "[n]o service is required on a party who is in default for failing to appear[, b]ut a pleading that asserts a new claim for relief against such a party must be served on that party under Rule 4." And Rule 54(c) declares that "[a] default judgment must not differ in kind from, or exceed in amount, what is demanded in the pleadings." The purpose of these rules is to ensure that defendants have notice of the contours of all claims upon which they may be held liable, and can therefore decide on the basis of the pleadings whether to defend the action. See Varnes v. Local 91, Glass Bottle Blowers Ass'n of U.S. & Can., 674 F.2d 1365, 1368 (11th Cir.1982) ("Rule 4, and Rule 5(a) as it applies to parties in default for failure to appear, reflect a policy that a defendant should receive notice of all claims for relief upon which a court may enter judgment against him."); 10 Wright et al., supra, § 2663 ("The theory of [Rule 54(c)] is that the defending party should be able to decide on the basis of the relief requested in the original pleading whether to expend the time, effort, and money necessary to defend the action.").

Allowing de facto amendment of an inadequate complaint through testimony taken at a hearing on damages, without notice to the defendant and a concomitant opportunity to respond, would needlessly conflict with these rules and the policies they serve. Cf. Silge v. Merz, 510 F.3d 157, 161-62 (2d Cir.2007) (explaining the purposes of Rule 54(c) and observing that the notice of the plaintiff's request for damages beyond those sought in the original complaint was delivered after the entry of default and "may therefore have seemed to come too late in the day for defendants to undo the consequences of their prior decision to default"); Varnes, 674 F.2d at 1368-69, 1370-71 (holding that improper service of an amended complaint required that the default judgment rendered thereon be vacated under Rule 5(a)(2) and denying the plaintiff's request for default on the original complaint because "she should not be able to engraft necessary portions of the amended complaint onto the original complaint to perfect the original complaint... for the sole purpose of legitimating a clerk's default").10

Moreover, this approach would present practical problems that will not burden the regime we endorse today. First, treating testimony at a prove-up hearing as curative of inadequate pleadings would give trial judges impermissible latitude to assist individual parties through the conduct of such hearings. On one hand, a judge sympathetic to a plaintiff's claim — or frustrated with a defendant's nonfeasance — could enter default judgment on the plaintiff's facially deficient pleadings, intending to accept testimony at the damages hearing that would fill the gaps in the plaintiff's prima facie case. The default judgment would then be immune from attack on the sufficiency of the pleadings, depriving the defendant of a critical — and longstanding — avenue of relief. On the other hand, a judge skeptical of a plaintiff's claim — or sympathetic to a defendant's plight — could dismiss the deficient complaint sua sponte or decline to enter default judgment rather than hold a hearing that would enable the plaintiff to shore up the pleadings. Either way, such discretion would both disserve the policies underlying default judgments and undermine fundamental fairness.

Second, and relatedly, permitting prove-up testimony to effectively amend a complaint would unfairly give plaintiffs a second bite at the default-judgment apple, without providing countervailing safeguards for defendants. In the Rule 55(b) hearing context, there is a critical difference between evidence that "establish[es] the truth of an[] allegation" and evidence that makes out an allegation in the first instance. To take a simple example: In a negligence suit arising out of a car accident, the plaintiff's pleadings assert that the defendant drove through a stale red light and collided with the plaintiff's car. After the defendant defaults, the court holds a damages hearing. During the hearing, the court takes testimony to "establish the truth" of the well-pleaded factual allegation that the traffic light was red, and the plaintiff testifies that the light was in fact green. The court may now decline to enter default judgment because there is no proof of one of the facts necessary to assess liability. See, e.g., Au Bon Pain Corp. v. Artect, Inc., 653 F.2d 61, 65 (2d Cir.1981) ("[A] district court has discretion under Rule 55(b)(2) once a default is determined to require proof of necessary facts and need not agree that the alleged facts constitute a valid cause of action...." (citing 10A Wright et al., supra, § 2688)). If, however, the plaintiff's pleadings allege only that the defendant drove "negligently," testimony at the damages hearing that the defendant drove through a stale red light would not "establish the truth of an[] allegation" — "negligence" is a bare conclusion of law that the defendant does not admit by his default, Nishimatsu, 515 F.2d at 1206 — but rather would make a novel factual allegation. Accordingly, reading this testimony into the original pleading would let the plaintiff amend the complaint without affording the defendant an opportunity to respond.11

This approach would also undermine our policy of enabling defendants intelligently to weigh the costs of default against the costs of defending an action. For instance, a defendant named in an ostensibly frivolous complaint may opt to default rather than incur the costs of defending the suit, only for the trial court to elicit testimony in a prove-up hearing that would make the claim plausible. Armed with this information, the defendant may well have elected to defend the claim; but if the testimony is impliedly incorporated into the original complaint, then the defendant cannot revisit his earlier decision. Nor, for that matter, can the defendant avail himself of the procedural protections associated with amendments to pleadings and variances between pleadings and default judgments.12

In sum, the Federal Rules, our precedents and those of our sister circuits, and policy and practical considerations lead us to the conclusion that a fatally deficient complaint cannot be cured by testimony at a prove-up hearing. Rather, a district court in these circumstances has three options. It may (1) dismiss the complaint sua sponte under Rule 12(b)(6) without prejudice, allowing the plaintiff to amend and refile;13 (2) grant leave to amend the complaint to include the facts presented at the hearing; or (3) treat the hearing evidence as constituting a de facto amendment to the complaint and then allow the defendant to answer the complaint as amended. By requiring or inferring amendment, all three approaches "open up" the default and entitle the defendant to another opportunity to answer.14

Turning to the facts before us, we hold that the district court abused its discretion in entering default judgment against McDonald Transit. As explained above, Wooten's barebones pleadings were inadequate to support a default judgment. Although Wooten's testimony at the prove-up hearing, if included in the original complaint, would likely have satisfied our pleading standards, it cannot properly be considered part of the pleadings. Accordingly, the district court should have dismissed the complaint sua sponte before entering a judgment of default or, upon receiving Wooten's testimony at the hearing, either granted leave to amend the complaint or treated the complaint as functionally amended. In any event, McDonald Transit should have been served with an amended complaint containing Wooten's "plausible" factual allegations, and should have received a corresponding opportunity to answer (or properly default), before judgment. We do not condone McDonald Transit's conduct in the district court,15 but we cannot uphold the entry of default judgment on pleadings that are infirm or are subject to de facto amendment without the notice to the defendant that is ordinarily required.16

IV. CONCLUSION

For the foregoing reasons, we VACATE the district court's entry of default judgment and REMAND with instructions to dismiss the complaint with leave to amend.

JACQUES L. WIENER, Circuit Judge, dissenting.

Despite my deep respect for the panel majority, I am convinced that its opinion sends the wrong message to the district courts of this circuit and, more troubling, that it will eviscerate the role of default judgments in the efficient administration of civil litigation. I have long accepted that the noble experiment with notice pleading has been relegated to the trash bin of history, as recently illustrated by the Supreme Court's opinions in Twombly1 and Iqbal.2 I also accept that Wooten's threadbare complaint is factually deficient in the extreme. Nevertheless, when hereafter asked to grant default judgments based on petitions that are anything short of absolute perfection, our district courts will almost certainly refuse to do so without first affording the recalcitrant defendants yet another bite at the apple (here, a third bite!). Despite our disfavoring default judgments, they do serve an important purpose in the prompt and efficient administration of civil litigation. Yet today's majority ruling will unnecessarily prolong that process and likely eliminate the positive role played in it by default judgments. This is why I am compelled to dissent.

I note first that, despite stating that "a fatally deficient complaint cannot be cured by [such] testimony," the majority opinion does allow the plaintiff to supplement his complaint with evidence introduced at a Rule 55(b)(2) prove-up hearing. Indeed, the text of Rule 55(b)(2) and our own precedent in Nishimatsu leaves open this avenue for curing a deficient complaint.3 Where the majority and I differ is in what should follow the prove-up hearing. The majority would have the district court grant the defaulting defendant yet another opportunity to answer and seek Rule 12(b)(6) or summary judgment relief; I would affirm the court's grant of the default judgment.

The majority bases its holding primarily on policy considerations, analogizing to Federal Rule of Civil Procedure 5(a)(2), which requires the defendant to receive notice if the plaintiff's pleadings add a "new claim for relief."4 The majority warns that, unless we extend this rule to follow Rule 55(b)(2) hearings, we will (1) give the plaintiff the right to "amend the complaint without affording the defendant an opportunity to respond," and (2) "undermine our policy of enabling defendants intelligently to weigh the costs of default against the costs of defending an action."

But defendants who have defaulted simply are not situated similarly to their nondefaulting counterparts. A defaulting defendant has, in fact, already had an opportunity to respond (in this case, two opportunities). Moreover, a complaint supplemented by facts revealed at a Rule 55(b)(2) hearing does not add "a new claim for relief"5 — it merely makes whole an existing claim, albeit a deficient one.

At first glance, holding a defendant to its initial default may appear to be unfair. But, doing so is a consequence of the defendant's own inaction. Furthermore, we serve "our policy of enabling defendants intelligently to weigh the costs of default" so long as we give them a clear rule to follow. If defendants know that they will be bound by the facts proven at a Rule 55(b)(2) hearing, they can weigh that factor when making their initial decision whether or not to default.

Giving me even greater concern is the practical consequence of the majority's decision. It requires our district courts to rule, sua sponte and without briefing, on the sufficiency of plaintiffs' complaints, all the while giving those courts every incentive to err on the side of insufficiency. Consider a hypothetical in which the viability of the plaintiff's complaint is a close call, and a Rule 12(b)(6) motion would likely be contested vigorously but for the defendant's default. After the district court has held a Rule 55(b)(2) hearing, the majority would then have it make a ruling sua sponte as to whether the complaint was deficient in the first place. I cannot believe that a district court would ever hold the complaint sufficient and enter default judgment on it: If the court should do so at that point, the defendant would lose nothing by entering the case, essentially to litigate its Rule 12(b)(6) case via a Rule 60(b) motion. The majority thus provides yet another safety net for the previously defaulting defendant, who would also enjoy a generous standard of review in its effort to convince us on appeal that the district court erred in holding the original complaint sufficient.6

On one hand, a district court that enters a default judgment on a complaint that's sufficiency is truly a close call would face a significant probability of reversal. If, on the other hand, the district court should require amendment and notice to the defaulting defendant, the plaintiff might not even have the right to appeal.7 Thus, when faced with a marginal complaint, a district court will likely opt to err on the side of requiring amendment (either actual or de facto) and providing the defendant in default a new opportunity to respond.

This result is inordinately lopsided and, even worse, favors the wearer of the black hat over the wearer of the white hat. Our default judgment jurisprudence carefully balances our preference for judgments on the merits with "considerations of social goals, justice and expediency."8 We do not honor this balance, much less retain it, by allowing the defaulting defendant to lie behind the log until after a Rule 55(b)(2) hearing, then have the option to jump into the fray and litigate the merits as though his default had never occurred. Under today's decision, the defendant may comfortably sit back while the plaintiff goes to the trouble, time, and expense of a Rule 55(b)(2) hearing, leaving the district court to grapple with legal issues that are truly the defendant's own duty to raise and support. I see this as grossly unfair to the innocent plaintiff and a waste of judicial resources. Far better and fairer, in my view, to hold the defaulting party to his default. These are the reasons why I must respectfully dissent.