Grounds for Denying Leave

Grounds for Denying Leave

Rule 15(a)(2) closes with an instruction that governs every motion to amend: “The court should freely give leave when justice so requires.”¹ That instruction creates a presumption in favor of amendment, and a presumption can be overcome. The rule does not say what overcomes it, so the answer comes from the decisions applying the rule, and they converge on five reasons drawn from Foman v. Davis: undue delay, bad faith or dilatory motive, repeated failure to cure deficiencies by amendments already allowed, undue prejudice to the opposing party, and futility. When none of them appears, the leave sought should be freely given.² A judge who denies leave because the case would look tidier without the change, or because the original pleading seems good enough, has named none of them.

The five grounds protect interests that can outweigh a pleader’s chance to reach the merits. The opponent should not have to rebuild its case around a moving target. The court should not watch its docket stall because a pleader amends at the last moment, or amends again and again without fixing the same flaw. And no one should spend effort on a pleading that could not succeed even if allowed. The grounds divide along those lines. Delay, bad faith, and repeated failure to cure look at the movant’s conduct: when the pleader asked, why, and how many earlier chances it used. Prejudice looks at what the amendment would do to the opponent. Futility looks at the proposed pleading and asks whether it states a claim at all.

Each ground is a separate question with its own evidence, and a denial must rest on one of them. The subsections below take them in this order: the movant’s timing and motive, the effect on the opponent, the legal sufficiency of the amendment, and the movant’s record with chances already given. In every case the presumption favors the pleader, and the party resisting the amendment, or the judge denying it, should be able to point to a concrete reason from the list.

Undue Delay and Bad Faith

Rule 15(a)(2) sets no deadline for seeking leave and tells the court to give leave freely when justice so requires.³ What keeps a party from waiting indefinitely is the first pair of reasons on the Foman list: undue delay, and bad faith or dilatory motive on the movant’s part.⁴ The district court in Rocheux International of N.J., Inc. v. U.S. Merchants Financial Group explains why the two belong together. Each directs the court to the movant’s motives for not amending sooner, whereas prejudice, the subject of the next subsection, directs the court to the effect on the opponent.⁵

Delay alone does not justify denying leave. Picture a motion that arrives late on the calendar but costs neither the court nor the opponent anything. Refusing it would only keep a claim or defense with possible merit out of the case, the very result the instruction to give leave freely exists to prevent. The Third Circuit authorities applied in Rocheux therefore draw a line: delay becomes “undue” when it places an unwarranted burden on the court, and it becomes “prejudicial” when it places an unfair burden on the opposing party.⁶ The court’s burden is a schedule built around pleadings that no longer match the case. The opponent’s burden is the cost of rebuilding preparation that assumed a different case. Until delay reaches one of those burdens, it is only a fact about the calendar, and the common error is to treat a late motion as a refused motion without asking whom the lateness burdens and why the movant waited.

Because the rule weighs the movant’s conduct, the movant’s explanation does most of the work. Prejudice to the opponent is the touchstone for denying an amendment, but the longer the delay goes unexplained, the less the opponent must show to defeat the motion.⁷ That sliding scale follows from what an explanation is worth. A movant who learned something new, such as a fact that surfaced only in discovery, offers a reason that offsets the cost of lateness. A movant who offers nothing leaves the court with that cost and no counterweight, so even a modest showing of harm can tip the balance. Only the movant knows why it waited, which is why the court asks the movant and weighs an unexplained delay against it.

Rocheux shows the scale at its steepest. The defendants answered in February 2007, and discovery closed in November 2008. They first sought to add counterclaims in April 2010, while opposing the plaintiff’s motion for summary judgment, more than three years after the answer and more than sixteen months after discovery closed.⁸ Their own interrogatory answers from August 2007 showed they had known of the potential offsets for roughly thirty months before moving, and they explained themselves only by a belief, which the court found unjustified, that the parties’ installment contract already opened the door to offsets for earlier years.⁹ The counterclaims would also have drastically changed the scope of the claims before the court, so the delay was undue.¹⁰ The court pointed to a Third Circuit decision that found a motion untimely where the party waited until its adversary moved for summary judgment and the proposed claims did not rest on evidence that came to light after discovery.¹¹ A movant warned of a gap early looks much the same. In Wright v. PNC Financial Services Group, Inc., a magistrate judge’s report noted, before the discovery plan was filed, that the complaint made no claim under the Michigan Estates and Protected Individuals Code. The plaintiff waited until he faced a motion for judgment on the pleadings to ask to amend, and the court found the lateness without excuse.¹²

Rocheux involved a long, unexplained delay over a claim the movant already knew about. Tobinick v. Novella shows that the calendar is not the only measure. The plaintiffs’ two motions for leave came only about a year after the original complaint, but the case had advanced: the district court had dismissed two defendants for lack of personal jurisdiction, dismissed one count, granted summary judgment for another defendant, and denied a preliminary injunction.¹³ The first motion was filed on the deadline for amended pleadings and sought extensive changes, including new factual allegations, a civil conspiracy claim, the return of a previously dismissed defendant, and two new defendants, and the Eleventh Circuit held that denying leave was not an abuse of discretion.¹⁴ Delay is thus measured against the stage the case has reached and the size of the change requested. A year is modest for a party making a small correction and substantial for one who wants to reopen the case after the court has ruled on dispositive questions.

The reason for amending is the other half of the motive inquiry. New facts supply a common and good reason, because the amendment will almost always advance litigation of the merits, but a court may still ask why the facts surfaced only recently.¹⁵ In Beeck, Aquaslide’s president inspected the water slide only while preparing for his deposition and concluded that it was not an Aquaslide product, although the insurers of Aquaslide, of the plaintiff’s employer, and of the slide’s owner had all earlier concluded that it was.¹⁶ The timing, after the statute of limitations had run, could have raised a question about Aquaslide’s motive, but the shared misidentification of the slide removed any whiff of bad faith.¹⁷ A late-arriving legal theory is less readily forgiven than a late-found fact, since counsel might have thought of it sooner. Yet courts routinely allow new claims and defenses, because theories look different on a fuller record, tactics and even the law can change during a case, and a client should not pay for counsel’s slow idea.¹⁸

Bad faith adds purpose to delay. A party that deliberately holds back a claim or defense until the eve of trial, to deprive its opponent of the chance to prepare, gives the court a sound reason to deny leave, and other evidence of bad faith can be equally fatal.¹⁹

Lateness that costs the opponent nothing falls on the other side of the line. In Caribbean Broadcasting Sys., Ltd. v. Cable & Wireless PLC, the D.C. Circuit held that denying leave as untimely was an abuse of discretion, because the limitations period had not run and the defendants suffered no prejudice from the timing.²⁰ In Bausch v. Stryker Corp., the Seventh Circuit found no merit in a finding of undue delay, because the revised complaint advanced no new theory of relief and offered only a clarification and additional factual details; it added that the plaintiff was entitled to wait and see whether the court would find pleading problems that could be corrected.²¹ What separates both decisions from Rocheux is what the amendment asked of the opponent. The Rocheux counterclaims would have reshaped a case in which discovery was over, whereas the opponents in Caribbean Broadcasting and Bausch could point to no added burden.

Delay counts against a movant only because Rule 15(a) leaves the grant of leave to the court’s discretion. Krupski v. Costa Crociere S.p.A. contrasts that discretion with relation back under Rule 15(c), which decides whether an amended pleading is treated as filed on the date of the original. The Supreme Court observed that a court may consider undue delay or dilatory motive under Rule 15(a), but the speed with which a plaintiff moves to amend has no bearing on whether the amended complaint relates back.²² The reason is structural. Rule 15(c) lists its requirements exclusively, and the amending party’s diligence is not among them, so relation back follows once those requirements are met.²³ Relation back is taught in its own right elsewhere; the point here is the contrast. A plaintiff’s slowness is a legitimate argument against granting leave and an illegitimate argument against relation back.

Undue Prejudice

Rule 15(a)(2) tells the court to “freely give leave when justice so requires,” and Foman v. Davis named undue prejudice to the opposing party among the reasons that can justify a refusal.²⁵ Courts treat prejudice as the touchstone for denying leave to amend.²⁶ The previous subsection explained that delay and bad faith direct the court to the movant’s reasons for waiting, while prejudice directs it to what the amendment would do to the opponent. The word undue carries the weight, because some harm to the opponent comes with nearly every amendment, and only harm that makes the opponent’s defense unfair counts toward denial.

The law tolerates ordinary harm because of why parties amend. A party amends to improve its position, so an opponent facing a new claim or a changed theory is necessarily worse off than before. If that disadvantage were enough to refuse leave, almost no amendment would ever be allowed, and the instruction to give leave freely would mean nothing.²⁷ Call this ordinary disadvantage merits prejudice: it consists only of the changed litigating posture, as when a defendant must answer two claims instead of one. The harm that counts is preparation prejudice, meaning the need for new discovery, new witnesses, or more time to defend, and it is the key factor in the analysis.²⁸ A pleading tells the opponent what to prepare for, so a defendant chooses depositions, experts, and strategy by reading it. When an amendment changes what must be proved, the opponent must redo some of that work or can no longer do it at all, and no later ruling restores the lost time.

Preparation prejudice grows as the trial date approaches and is often near its maximum at trial.²⁹ Time is the resource an amendment consumes. Soon after the answer, the opponent has a full discovery period in which to investigate the new matter, and the court has a distant trial date it can keep. Near trial, discovery has closed and witnesses and experts are committed, so the same amendment lands on a calendar with no room to absorb it. GEOMC Co. v. Calmare Therapeutics Inc. applies that logic to a new counterclaim. Early in the case, a counterclaim may be as broad as one filed in answer to an original complaint, but late in the case a counterclaim that goes beyond the new claims in the latest amended complaint usually causes escalating prejudice and undue expansion of the litigation, so it should normally be refused.³⁰ Some decisions state the standard even more generously, saying that prejudice “ordinarily is not considered to have occurred unless the Motion is made during or after the actual trial.”³¹ That formulation shows how seldom courts find prejudice early in a case. Prejudice before trial remains possible, as the next decisions show.

The party resisting the amendment carries the burden of showing prejudice. Beeck states the allocation directly: prejudice must be shown, the burden rests on the party opposing the amendment, and the trial court must inquire into prejudice in light of the particular facts of the case.³² The burden follows the information, since only the opponent knows how much of its preparation depended on the pleading as written. An opponent that can name the discovery it would have to take, the witnesses it would have to find, or the strategy it would have to abandon gives the court something concrete to weigh against the presumption in favor of amendment. An opponent that only objects gives the court nothing. The required showing also moves with the movant’s explanation, because the longer the delay goes unexplained, the less the opponent must show.³³

Two decisions with similar calendars show where the line falls. In Material Supply International v. Sunmatch Industries, a defendant offered a statute-of-limitations defense after discovery closed and just before trial, and the D.C. Circuit held that refusing the amendment was an abuse of discretion, because the plaintiff already knew of the defense from an earlier memorandum and needed no discovery to prepare to try it.³⁴ In Wright v. PNC Financial Services Group, Inc., the plaintiff sought to amend only after facing a motion for judgment on the pleadings, and the court found the timing extremely prejudicial to the bank because discovery had since closed, a conclusion it reached alongside lateness without excuse and the plaintiff’s failure to say what new facts he would add.³⁵ Both motions arrived after discovery closed, so the calendar does not separate them. What separates them is what each amendment asked of the opponent: the plaintiff in Material Supply needed nothing it did not already have, while the bank in Wright could no longer use discovery to meet whatever the amendment would add.

Harrison v. Rubin rests on the same reasoning. The D.C. Circuit found an abuse of discretion in denying an amendment whose claims and defenses required essentially the same proof as the original pleading.³⁶ When the proof is the same, the opponent’s preparation already covers the amended pleading, so the delay costs it nothing. A sound answer to an opponent’s objection therefore asks whether the amendment would force the opponent to do something it can no longer do, such as take discovery, find witnesses, or change its strategy. An objection that the opponent must defend an extra claim, or simply that the motion is late, does not answer that question.

Futility

Rule 15(a)(2) tells the court to freely give leave to amend when justice so requires, and that instruction assumes the proposed change has something to offer.³⁸ An amendment that would accomplish nothing falls outside that generosity, and courts call it futile. In Shane v. Fauver the Third Circuit listed futility alongside undue delay, bad faith, dilatory motive, and prejudice as grounds that can justify denying leave, and it defined the term: an amendment is futile when the complaint, as amended, would fail to state a claim upon which relief could be granted.³⁹ That definition looks only at the pleading the movant proposes to file. Delay and bad faith ask why the movant waited, and prejudice asks what the change does to the opponent, but futility asks whether the new pleading would survive if the opponent attacked it.

The attack in question is the familiar motion to dismiss for failure to state a claim. Applying Third Circuit authority, a federal district court in Pennsylvania explained that courts assessing futility use the same standard of legal sufficiency that governs a Rule 12(b)(6) motion, so futility turns on the likelihood that the content of the amendment will survive a renewed motion to dismiss.⁴⁰ The District of Nevada described the test in the same terms and applied it to proposed counterclaims, working through them one at a time and starting from the elements of each cause of action, such as tortious interference with contract under Nevada law.⁴¹ Because the standard is borrowed whole, the analysis you already use on a motion to dismiss carries over unchanged, with the proposed pleading in place of the complaint on file.

The reason for the ground appears when leave is granted anyway. A pleading that cannot survive a Rule 12(b)(6) motion is filed, the opponent moves against it, and the court dismisses it, leaving the case where it stood after one more round of briefing. Refusing leave at the start saves that round for everyone. The same reasoning limits the ground. If the court asked at the leave stage whether the movant would actually win, the movant would have to prove a claim before being allowed to plead it, and the policy of liberal amendment would collapse into an early trial on the merits. A Maryland federal court therefore said that a review for futility is not an evaluation of the underlying merits of the case, and it quoted the Fourth Circuit’s instruction that “[u]nless a proposed amendment may clearly be seen to be futile because of substantive or procedural considerations, . . . conjecture about the merits of the litigation should not enter into the decision whether to allow amendment.”⁴² The word clearly matters there, since a defect that is arguable rather than evident is a matter for a later motion. The same court added that leave is warranted, at minimum, when at least some of the movant’s claims are not futile.⁴³

Henderson v. Wells Fargo Bank, N.A. shows both sides of the line inside one opinion. Henderson sued Wells Fargo over, among other things, insurance placed on his escrow account, and the court dismissed most of his claims because they were either abandoned or failed as a matter of law.⁴⁴ He had amended once before, but the court observed that he had not had the benefit of the motion to dismiss or of its opinion, and it still refused him a further chance to replead the dismissed claims, because further attempts to amend them would be futile.⁴⁵ No rewording can rescue a claim that the law does not recognize. His remaining contract claim stood differently, because it failed only for want of an allegation that he had performed by obtaining insurance meeting the requirements of paragraph 5 of the Deed, and the court permitted an amended complaint addressing that deficiency.⁴⁶ A missing allegation of that kind is something the pleader can supply, so the amended claim might well state a claim. The permission was also limited to that claim, because the amended complaint could not include any claim the court had dismissed.⁴⁷

Appellate courts reverse futility rulings that treat a curable or sufficient pleading as hopeless. In Bausch v. Stryker Corp. the Seventh Circuit held that denying leave was an abuse of discretion in part because the district judge erred in concluding that the amended complaint was futile on the merits because its claims would still be preempted, and the revised complaint had clarified that the claims were focused only on violations of federal law and added factual details, especially about an FDA warning letter.⁴⁸ In Caribbean Broadcasting the D.C. Circuit held the proposed second amended complaint was not futile, because its allegations supported the subject matter jurisdiction whose absence had led to dismissal.⁴⁹ What separates those reversals from the refused claims in Henderson is whether the proposed pleading answers the defect: the new allegations in Bausch and Caribbean Broadcasting were directed at the ground on which the earlier pleading failed, while no allegation could revive Henderson’s abandoned or legally barred claims. A defect in the original complaint does not by itself make the amendment futile, because the question is whether the amended pleading cures it.

Step four does the real work, because it separates a claim the law forecloses from a claim the pleader has simply drafted badly. A court that denies leave on a ground a better-drafted allegation would cure has used futility to penalize drafting, and that is the kind of ruling the reversals above corrected. The point is developed further in the later section on amendment after dismissal, where the plaintiff usually asks for leave to repair exactly the defect the court has just identified.

Repeated Failure to Cure

Rule 15(a)(2) tells the court only that it should freely give leave to amend when justice so requires, so the reasons that can override that instruction come from the decisions applying it.⁵¹ The list that courts draw from Foman v. Davis includes the movant’s repeated failure to cure deficiencies by amendments previously allowed.⁵² The earlier grounds looked at why the movant waited, what the amendment would do to the opponent, and whether the new pleading could state a claim at all. This ground looks backward at what the movant did with chances already given. An earlier amendment is only a fact about the docket, while repeated failure to fix a problem that someone has pointed out is a pattern, and the pattern is what can justify refusing leave.

The factor is tied to repetition because of what an amendment is for. A pleader who has been told what is wrong, whether by the opponent’s motion or by the court’s own ruling, holds the information needed to repair the pleading, and the first amendment after that notice is the occasion to use it. If the repair works, the case moves on. If the new pleading leaves the same gap in place, the pleader has shown that another round is unlikely to improve anything, and each further round costs the opponent another motion and the court another ruling on a pleading that is not getting better. A refusal at that point treats the pleader fairly, because the pleader had both notice and opportunity. The logic also runs the other way: a pleader who has not been told of a defect cannot have failed to cure it, which is why a plaintiff whose complaint might be saved by better drafting generally receives at least one chance to amend before the court dismisses with prejudice.⁵³

The facts that moved the answer in that scene are the court’s notice and the buyer’s two failed responses to it. A case with a defect but without that history comes out the other way. In Garcia v. Chiquita Brands Int’l, Inc. the plaintiffs’ claims were barred on the face of their complaint by Colombia’s ten-year limitations period, and the complaint had not said which plaintiffs were minors who might benefit from minority tolling.⁵⁴ The district court denied their Rule 59(e) motion to amend. The Eleventh Circuit found that none of the recognized reasons for refusal applied: there was no evidence of bad faith or delay, and the plaintiffs had not repeatedly failed to cure the minority-tolling deficiency, because they had not yet had a chance to plead facts on it.⁵⁵ Their proposed amended complaint identified each plaintiff who was a minor before March 25, 2010, which was enough to plausibly state a tolling claim for those plaintiffs, and the defendant could point to no prejudice from one more opportunity.⁵⁶ The litigation had lasted more than a decade, but length alone, without other evidence of prejudice or bad faith, does not justify denying the chance to amend, so dismissal with prejudice was an abuse of discretion.⁵⁷ Both the buyer and the Garcia plaintiffs had defective pleadings. What separates them is whether a chance to cure had already come and gone.

Courts phrase the factor differently. The wording used above asks about a pattern of failed cures, while the five-factor test applied in JPMorgan Chase Bank, N.A. v. KB Home asks whether the plaintiff has previously amended.⁵⁸ A count of earlier amendments is a rough measure of the chances a movant has already used, and a summary of the governing principles lists the number of prior amendments among the considerations that guide the court’s discretion.⁵⁹ The count means little, though, without knowing what each earlier amendment responded to. An amendment made as of course under Rule 15(a)(1), as taught earlier in this chapter, usually comes before the opponent has moved or the court has ruled, so it cannot have been a failed attempt to cure a defect that nobody had yet named.

Henderson, the futility case above, shows a court counting this way. The plaintiff there had amended once before, and the court acknowledged it, but it also observed that he had not had the benefit of the bank’s motion to dismiss or of the court’s opinion.⁶⁰ The court refused further amendment of the dismissed claims, though the reason it gave was futility, since those claims were abandoned or failed as a matter of law.⁶¹ For the one claim that failed only because the complaint did not allege that he had performed by obtaining insurance meeting the deed’s requirements, the court identified that gap and allowed a new complaint to address it.⁶² The first amendment had come before the defect was named, and the court offered a chance to cure once it was named. That sequence is the reverse of repeated failure.

An opponent who answers a motion for leave by saying that the plaintiff has already amended has therefore pointed to very little, unless the earlier amendment followed a named defect and left it standing. The factor also remains separate from futility, since futility asks whether the proposed pleading could state a claim, while this factor asks what the movant did with earlier chances. A court may rely on both when a pleader has been shown the same legal gap more than once, but each rests on its own showing.