PRECEDENTIAL
UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT
_____________
No. 10-2762
_____________
WILLIAM A. GRAHAM COMPANY,
d/b/a THE GRAHAM COMPANY
v.
THOMAS P. HAUGHEY;
USI MIDATLANTIC, INC.,
Appellants
_____________
On Appeal from the United States District Court
for the Eastern District of Pennsylvania
District Court No. 2-05-cv-00612
District Judge: The Honorable Harvey Bartle III
Argued March 22, 2011
Before: FUENTES, SMITH, and VAN ANTWERPEN,
Circuit Judges
(Filed: May 16, 2011)
Aleksander J. Goranin
1
David J. Wolfsohn (Argued)
Woodcock Washburn
2929 Arch Street
12th Floor, Cira Centre
Philadelphia, PA 19104
Henry E. Hockeimer, Jr.
Ballard Spahr
1735 Market Street
51st Floor
Philadelphia, PA 19103
Counsel for Appellee
Floyd Abrams (Argued)
Cahill, Gordon & Reindel
80 Pine Street
New York, NY 10005
Elizabeth S. Campbell
Thomas E. Zemaitis
Pepper Hamilton
18th & Arch Streets
3000 Two Logan Square
Philadelphia, PA 19103
Counsel for Appellants
________________
OPINION
________________
SMITH, Circuit Judge.
2
Defendants Thomas Haughey and USI MidAtlantic,
Inc. appeal a second time from a judgment entered against
them in the Eastern District of Pennsylvania. (Our earlier
decision, to which we will refer as Graham I, is reported at
568 F.3d 425 (3d Cir. 2009).) A jury found Haughey and USI
liable for surreptitiously infringing the William A. Graham
Company‘s copyrights over the course of more than a decade,
and returned a verdict in Graham‘s favor totaling nearly $19
million. To this the District Court added an award of more
than $4.6 million in prejudgment interest. In the defendants‘
view, the jury verdict is so large as to shock the judicial
conscience, and the prejudgment interest award is contrary to
law. We disagree with both contentions and will therefore
affirm.
I
In 1991, Haughey left his job with Graham, an
insurance brokerage, for one with USI, a Graham competitor.
When he changed employers, Haughey took with him two
binders containing hundreds of pages of text describing
various types of insurance coverages, exclusions, conditions,
and similar matter. These binders had been prepared by
Graham employees and were subject to that firm‘s copyrights.
From July 1992 until 2005, Haughey and the rest of USI
made use of these materials in preparing insurance coverage
proposals for presentation to their clients. This use of
Graham‘s creation constituted a long-running copyright
violation, though not in the paradigmatic, ―direct,‖
reproduction-and-sale-of-protected-works form. The
infringement was instead ―indirect,‖ in that the defendants
used the copyrighted materials without permission in order to
sell their own insurance products. This conduct was hidden
3
from view, and Graham apparently did not discover it until
November 2004.
On February 8, 2005, Graham filed suit against
Haughey and USI under the Copyright Act, 17 U.S.C. §§ 101
et seq. The defendants raised the Act‘s three-year statute of
limitations as a defense, but the District Court held that the
―discovery rule‖—which tolls the limitations period until the
plaintiff learns of his cause of action or with reasonable
diligence could have done so—applied to the Copyright Act
and therefore saved at least part of the complaint, subject to
the jury‘s determination of when Graham should have learned
of its cause of action.
The case proceeded to trial. Although the Copyright
Act permits the plaintiff in an infringement action to recover
either statutory damages or ―actual damages and any
additional profits of the infringer,‖ 17 U.S.C § 504(a),
Graham eschewed the statutory damages provision and did
not claim to have suffered any actual damages. It was
therefore left to seek only the infringers‘ profits—that is, ―any
profits of the infringer[s] that are attributable to the
infringement.‖ Id. § 504(b). To succeed on such a claim, a
plaintiff is first required to prove the defendants‘ gross
revenues over the course of the relevant time period, and then
to establish a causal nexus between the infringement and the
profits sought. Graham I, 568 F.3d at 442 (citing Polar Bear
Prods., Inc. v. Timex Corp., 384 F.3d 700, 711 (9th Cir.
2004)). Graham proved gross commissions of about $32
million for USI and $3 million for Haughey personally. The
jury also made the necessary causation finding.
4
Once the plaintiff has done its part, the burden shifts to
the defense to prove that some of its revenues were
―attributable to factors other than the copyrighted work,‖ and
are therefore not recoverable. 17 U.S.C. § 504(b). Graham
conceded that 25 percent of USI‘s revenues were deductible
expenses, reducing its potential recovery against USI to
around $24 million. From there, the defendants argued for
further reductions to account for their own contributions to
their success. The jury credited these arguments in part, and
accordingly awarded Graham $16,561,230 from USI and
$2,297,397 from Haughey—representing about 70 percent of
USI‘s profits, and 75 percent of Haughey‘s, over the course
of the relevant time period. The jury also found that Graham
had not been on notice of the infringement prior to February
9, 2002—which meant that no part of its claim was time-
barred.
After the jury returned its verdict, the District Court
determined, based on the trial evidence, that Graham had in
fact been placed on inquiry notice of the defendants‘ conduct
through the existence of ―storm warnings‖ as early as the fall
of 1991. The court therefore set aside the jury‘s verdict and
held a second trial limited to damages that had arisen within
three years of the commencement of Graham‘s action. The
second jury entered a verdict in the amount of $1.4 million
against USI and $268,000 against Haughey.
The parties cross-appealed. Graham argued that the
District Court‘s ―storm warnings‖ analysis was mistaken, and
that the initial verdict should be reinstated. The defendants
argued that Graham had failed adequately to prove the
requisite causal nexus. We affirmed in part and reversed in
part. Causation, we said, had been adequately proven at the
5
first trial. Graham I, 568 F.3d at 442–43. But while the
District Court had correctly held that the discovery rule
applied to Copyright Act claims, it had erred in finding
―storm warnings‖ in the face of the first jury‘s well-supported
conclusion that Graham could not reasonably have discovered
the infringement at any time before February 9, 2002. Id. at
441. We accordingly remanded the case for consideration of
the defendants‘ argument that the 70 and 75 percent
apportionments of their profits were unsupported by the
evidence and that the verdict was therefore excessive.
The District Court rejected the excessiveness argument
and reinstated the original jury‘s verdict. It also granted
Graham‘s motion for prejudgment interest, which the court
awarded in accordance with the calculations of Dr. Richard J.
Gering. Dr. Gering‘s report, the substance of which the
defendants neither challenged nor rebutted, is premised on
interest beginning to accumulate in 1992, when the first
infringement occurred. The defendants took exception to this
choice of date, arguing that interest should have been
awarded, if at all, only from the date in 2004 on which
Graham discovered its cause of action. They also asserted
that prejudgment interest is not available in infringers‘-profits
copyright cases, as a matter of law. The District Court
rejected these arguments and ordered interest awards totaling
$4,112,859 against USI and $570,542 against Haughey.
The defendants appeal, arguing both that the jury‘s
verdict shocks the judicial conscience and that the
prejudgment interest award is improper. The District Court
had jurisdiction under 28 U.S.C. §§ 1331 and 1338(a); ours is
premised on 28 U.S.C. § 1291.
6
II
We first address the defendants‘ claim that the
damages award is excessive. This is a steep climb. A district
court‘s decision regarding a request for a remittitur is
reversed only for abuse of discretion, and a case is remanded
for a new trial ―only if the verdict is so grossly excessive as to
shock the judicial conscience.‖ Gumbs v. Pueblo Int’l, Inc.,
823 F.2d 768, 771 (3d Cir. 1987) (citations and internal
quotation marks omitted); Tormenia v. First Investors Realty
Co., 251 F.3d 128, 138 (3d Cir. 2000). The climb is made
even steeper by the fact that the defendants bore the burden of
proving the extent to which the verdict should be reduced to
account for ―factors other than the copyrighted work.‖ 17
U.S.C. § 504(b). In the ordinary remittitur case, an aggrieved
defendant need ―only‖ show that the other side‘s evidence did
not justify the award. ―Grossly excessive‖ is a hard thing to
show, but the defendant can proceed by demonstrating that
his opponent‘s case is feeble, or that the damages are out of
proportion to the actual injury. Here, Haughey and USI are in
the position of having to prove that the jury underweighted
their own evidence to the point of shocking the judicial
conscience.
In making their case, the defendants emphasize the
several months and hundreds of man-hours of research and
relationship-building that precede every sale. Before USI can
earn a commission, a salesman has to know what risks to look
for, discern which ones the prospective client faces,
determine which underwriters are willing to insure those risks
and at what prices, and develop a coverage scheme that he
must then sell to the client. This process requires detailed
knowledge of the insurance industry and the sorts of products
7
that are available, and demands that the salesman establish his
credibility and trustworthiness through face-to-face meetings
and sales pitches. All this effort, the defendants insist,
accounts for the vast majority of their earnings. Furthermore,
they argue that only relatively small portions of the text of the
written proposals was lifted from the copyrighted materials,
and that much of this was boilerplate that cannot have had a
large influence on their clients‘ coverage-purchasing
decisions.1
We have some sympathy for the argument that these
efforts accounted for more than 25 or 30 percent of the
defendants‘ earnings (though the defendants coyly decline to
say just how much more). Any such sympathy is not,
however, sufficient to justify overturning the jury‘s verdict.
Graham has pointed to substantial evidence of its own that
supports the conclusion that the misappropriated documents
were an important element of the defendants‘ overall sales
strategy. Use of standardized, well-thought-out language
allows salespeople to demonstrate credibility, knowledge of
their insurance products, and understanding of the businesses
and risks being insured. As the District Court noted, ―the
[w]orks were virtually the only source of written insurance
policy explanations within USI.‖ Graham‘s eponymous
principal testified that his firm ―would not have been
successful without these documents.‖ Parts of the
1
The defendants also assert that the verdict includes
damages based on revenue arising from non-infringing proposals
or sales made without proposals. Properly understood, this
concerns not apportionment but the causal link between some part
of the damages award and the defendants‘ infringement. This
court already rejected the causation argument on the defendants‘
initial appeal. See Graham I, 568 F.3d at 442–43.
8
copyrighted materials described in some depth the various
forms of coverage that a client might want to purchase, and
would have been quite valuable to a salesman needing to
explain and summarize his offerings. All of USI‘s account
managers had access to copies of the books Haughey had
pilfered from Graham, and support staff were explicitly
reminded to make use of them. USI hired a temporary
employee to type the contents of the two huge binders into a
word processing program so that the text could be
electronically referenced and easily copied into new
documents. The defendants made use of Graham‘s language
in some 857 sales proposals prepared for 315 different clients
over the course of thirteen years. If that were not enough, the
defendants willfully destroyed a number of pre-1995
documents relevant to the case after being ordered by the
court to preserve them. This entitled Graham to a spoliation
instruction (which the defendants have not appealed) allowing
the jury to infer that in its early years the infringement was
actually more widespread than the evidence at trial showed.
The ―shocks the conscience‖ or ―miscarriage of
justice‖ standard for a grant of a new trial exists ―to ensure
that a district court does not substitute its judgment of the
facts and the credibility of the witnesses for that of the jury.‖
Sheridan v. E.I. Dupont de Nemours & Co., 100 F.3d 1061,
1076 (3d Cir. 1996) (citations and internal quotation marks
omitted). This concern is even more pressing at the appellate
level, where the judges have not had the opportunity to
observe the trial. In light of the conflicting evidence that we
have just outlined, a ruling for the defendants here would
constitute an impermissible substitution of the court‘s
assessment of the facts for the jury‘s. The verdict does not
shock the judicial conscience and will be affirmed.
9
III
Defendants raise several arguments against the District
Court‘s award of prejudgment interest. Specifically, they say
that interest is not available under the Copyright Act; that
even if it is, it should not be granted in infringers‘-profits
cases; and that in any event it should not have begun to
accumulate in this case until Graham had actually discovered
the infringement.
A
Although Congress has not enacted express statutory
authorization for prejudgment interest, see S. Rep. 97-275, at
11–12 (1981); Monessen Sw. Ry. Co. v. Morgan, 486 U.S.
330, 339 n.8 (1988), its silence is of no moment. ―Far from
indicating a legislative determination that prejudgment
interest should not be awarded, . . . the absence of a statute
merely indicates that the question is governed by traditional
judge-made principles.‖ City of Milwaukee v. Cement Div.,
Nat’l Gypsum Co., 515 U.S. 189, 194 (1995). Similarly, the
fact that Congress amended the Patent Act in 1946 to provide
for prejudgment interest, see 35 U.S.C. § 284, says nothing
about how we should interpret the Copyright Act. Although
patents are analogous to copyrights, the fact is that the Patent
Act amendment was a direct reaction to a series of cases
requiring exceptional circumstances for an award of
prejudgment interest covering the period of time preceding
liquidation of damages. See GM Corp. v. Devex Corp., 461
U.S. 648, 651–53 (1983). There is no analogous line of
decisions in the copyright context, and therefore no reason for
Congress to have amended the Copyright Act in response to a
perceived problem in its judicial interpretation.
10
In point of fact, this court‘s precedents indicate that
congressional silence favors permitting prejudgment interest
awards. It is a ―long-standing‖ rule of our federal common
law that, ―in the absence of an explicit statutory command
otherwise, district courts have broad discretion to award
prejudgment interest on a judgment obtained pursuant to a
federal statute.‖ Skretvedt v. E.I. Dupont de Nemours, 372
F.3d 193, 205–06 (3d Cir. 2004); see also Pignataro v. Port
Auth., 593 F.3d 265, 273–74 (3d Cir. 2010) (affirming this
rule and indicating that interest ―should be awarded based on
considerations of fairness‖). Of course, when the federal
courts engage in the development of gap-filling common law,
we must do so with the statute‘s policy goals in mind. See,
e.g., Rodgers v. United States, 332 U.S. 371, 373 (1947)
(―[I]n the absence of an unequivocal prohibition of interest on
such obligations, this Court has fashioned rules which granted
or denied interest on particular statutory obligations by an
appraisal of the congressional purpose in imposing them and
in the light of general principles deemed relevant by the
Court.‖) (citations omitted); Skretvedt, 372 F.3d at 206. So if
it were the case that the purposes of the Copyright Act ran
counter to those of prejudgment interest awards, a rule against
interest might be appropriate.
Reviewing those two sets of purposes, we find that
they are well aligned with one another. According to the
relevant House Report, the aims of the Copyright Act‘s
damages-plus-profits provision are two-fold: ―Damages are
awarded to compensate the copyright owner for losses from
the infringement, and profits are awarded to prevent the
infringer from unfairly benefiting from a wrongful act.‖ H.R.
Rep. No. 94-1476, at 161 (1976). See also, e.g., Polar Bear
Prods., 384 F.3d at 718 (―[T]he purpose of § 504(b) is to
11
compensate fully a copyright owner for the misappropriated
value of its property and ‗to avoid unjust enrichment by
defendants, who would otherwise benefit from this
component of profit through their unlawful use of another‘s
work.‘‖) (citation omitted). Far from being contrary to these
goals, the purposes of prejudgment interest—―making the
claimant whole and preventing unjust enrichment‖—parallel
them exactly. Fotta v. Trs. of the UMW Health & Ret. Fund
of 1974, 165 F.3d 209, 212 (3d Cir. 1998). Although interest
may allow an injured party to recoup the time-value of his
loss, its usefulness is not, as the defendants would have it,
confined to the provision of just compensation. Requiring
only that a losing defendant pay back the principal amount of
a wrongfully obtained sum permits him to retain the money‘s
time-value as a windfall in the form of an interest-free loan.
See Consol. Rail Corp. v. Certainteed Corp., 835 F.2d 474,
478 (3d Cir. 1986). The defendants‘ insistence that a plaintiff
seeking to recover an infringer‘s profits, as Graham does,
should not receive interest because he ―has suffered no loss at
all‖ is therefore unpersuasive, for interest is just as
appropriate to achieve full disgorgement as to ensure just
compensation.
Given our general rule permitting interest awards and
the consistency of that rule with the Copyright Act‘s
purposes, the defendants are left with few straws at which to
grasp. Seizing on language in the House Report to the effect
that § 504‘s purpose was to give courts ―unambiguous
directions concerning monetary awards,‖ H.R. Rep. No. 94-
1476, at 161, they assert that Congress intended to limit the
forms of available relief to those it specifically enumerated—
that is, statutory damages, actual damages, and/or the
infringers‘ profits, full stop. But they have not bothered to
12
quote the very next clause of the sentence on which this
argument is based, which advises that the Act was intended
―to provide the courts with reasonable latitude to adjust
recovery to the circumstances of the case, thus avoiding some
of the artificial or overly technical awards resulting from the
language of the existing statute.‖ Id. (emphasis added). The
statute was not meant to circumscribe the courts‘ authority on
the subject of interest awards; it certainly did not do so
explicitly.
The defendants finally argue that the procedures
applicable in infringers‘-profits cases make an interest award
inappropriate. Specifically, they cite the Act‘s burden-
shifting scheme and the fact that the measure of damages is
determined through the ―inherently arbitrary and artificial‖
process of submission to a jury. But regardless of who bears
what burden, the jury system is the process by which courts
typically determine facts, including the measure of copyright-
related profits and attribution percentages. A jury‘s verdict in
a case like this one is an assessment of the degree to which
the defendants have illegally enriched themselves. Nothing
about such a finding or the process through which it comes
about contradicts the aims to be achieved through an award of
prejudgment interest.
The defendants‘ arguments have not persuaded us that
our usual rule should not apply to copyright cases generally
or to infringers‘-profits cases in particular. Furthermore, we
think it self-evident that, as between a copyright owner and an
infringer, the former has the stronger equitable claim to the
time-value of income derived from his creation. We therefore
hold that prejudgment interest is available in copyright cases
13
at the District Court‘s discretion, exercised in light of
―considerations of fairness.‖ Pignataro, 593 F.3d at 274.
B
The defendants‘ sole remaining opportunity to reduce
their liability is their argument that the District Court
calculated its interest assessment improperly by counting
from the date in 1992 on which the first infringement
occurred rather than from the date in 2004 when Graham
discovered it. The basis for this argument is the interplay
between our prior holding in this case to the effect that, under
the ―discovery rule‖ Graham‘s cause of action did not
―accrue‖ for statute of limitations purposes until it discovered
its injury, see Graham I, 568 F.3d at 433–41, and the
Supreme Court‘s statement in West Virginia v. United States
regarding the period during which interest is awarded:
―Prejudgment interest serves to compensate for the loss of use
of money due as damages from the time the claim accrues
until judgment is entered, thereby achieving full
compensation for the injury those damages are intended to
redress.‖ 479 U.S. 305, 310 n.2 (1987) (emphasis added).
The defendants argue that, if the claim did not accrue until
2004 and the interest clock does not start ticking until accrual,
interest should not have begun to accumulate until 2004. This
syllogism rests on a misapprehension of the nature of the
discovery rule—albeit one that is quite common and that we
ourselves have, unfortunately, helped to propagate.
In keeping with West Virginia, we determine the date
on which to begin counting interest by asking when the claim
in question accrued. As a general matter, a cause of action
―accrues‖ when it has ―come into existence as an enforceable
14
claim or right.‖ Black‘s Law Dictionary (9th ed. 2009).
Stated another way, accrual is ―[t]he event whereby a cause of
action becomes complete so that the aggrieved party can
begin and maintain his cause of action.‖ Ballentine‘s Law
Dictionary (3d ed. 1969). ―‗Accrue‘ derives from the Latin
words ‗ad‘ and ‗creso,‘ to grow to; thus it means to arise, to
happen, to come into force or existence.‖ Strassburg v.
Citizens State Bank, 581 N.W.2d 510, 514 (S.D. 1998) (citing
Berry v. Branner, 421 P.2d 996, 998 (Ore. 1966); Black‘s
Law Dictionary (6th ed. 1968)). This is an objective feature
of any extant claim: the question is whether all of its elements
have come into existence such that an omniscient plaintiff
could prove them in court. At that point the cause of action is
―complete,‖ and has therefore accrued. Thus the
Pennsylvania Supreme Court has recently stated that ―a cause
of action accrues when the plaintiff could have first
maintained the action to a successful conclusion.‖ Fine v.
Checcio, 870 A.2d 850, 857 (Pa. 2005). And in Delaware, ―a
cause of action ‗accrues‘ . . . at the time of the wrongful act,
even if the plaintiff is ignorant of the cause of action.‖ Wal-
Mart Stores, Inc. v. AIG Life Ins. Co., 860 A.2d 312, 319
(Del. 2004) (citations omitted). See also, e.g., Stokes v. Van
Wagoner, 987 P.2d 602, 603 (Utah 1999) (a claim ―accrues at
the time it becomes remediable in the courts, that is when the
claim is in such a condition that the courts can proceed and
give judgment if the claim is established‖) (citation and
internal quotation marks omitted); Wittmer v. Ruegemer, 419
N.W.2d 493, 496 (Minn. 1988) (accrual is the point at which
―the action can be brought without being subject to dismissal
for failure to state a claim‖) (citation omitted); Aetna Life &
Cas. Co. v. Nelson, 492 N.E.2d 386, 389 (N.Y. 1986) (―The
Statute of Limitations begins to run once a cause of action
accrues, that is, when all of the facts necessary to the cause of
15
action have occurred so that the party would be entitled to
obtain relief in court.‖) (citations omitted); Robinson v.
Weaver, 550 S.W.2d 18, 19 (Tex. 1977) (―For the purposes of
application of statutes of limitations, a cause of action
generally can be said to accrue at the time when facts come
into existence which authorize a claimant to seek a judicial
remedy.‖) (citation omitted). Accrual has to do with the
existence of a legally cognizable right to obtain a judicially
sanctioned remedy, not the practical capacity to file a lawsuit.
This definition fits nicely with the purposes of
prejudgment interest. As West Virginia indicated, such
interest is often aimed at ―achieving full compensation for the
[plaintiff‘s] injury.‖ 479 U.S. at 310 n.2. Complete redress
will require that interest cover the entire period of the injury‘s
existence, which implies that injury and accrual generally
happen at the same time. The Court has elsewhere suggested
as much. See Kansas v. Colorado, 533 U.S. 1, 14 (2001)
(indicating that prejudgment interest should be awarded, in
appropriate cases, ―from the time of injury‖). Similarly, in
cases where the goal is the defendant‘s disgorgement rather
than rectification of the plaintiff‘s injury, fully attaining the
goal of prejudgment interest will require that it be awarded
from the date of the first illicit profit. To generalize, for
prejudgment interest to fully serve its purpose, it needs to be
awarded from the date on which the plaintiff first had a right
to collect the principal sum—that is, the date of accrual, as
that term has traditionally been defined. If this is the right
way to determine the accrual date—and thus the date on
which interest began to accumulate—then the District Court
was correct in awarding interest from the first infringement.
16
To obtain relief from the District Court‘s judgment, the
defendants must therefore argue for some other method of
determining the date on which Graham‘s cause of action
accrued. To that end, they cite Graham I‘s articulation of the
―discovery rule‖ as a doctrine that delays the date on which a
cause of action accrued. For reasons that we shall explain, we
conclude that this was a mischaracterization.
Accrual, as we have said, occurs once events satisfying
all the elements of a cause of action have taken place. At that
point, the period prescribed by the applicable statute of
limitations ordinarily begins to run—time begins to count
against the plaintiff, such that if enough of it goes past he can
no longer obtain relief. See, e.g., Fine, 870 A.2d at 857
(―[T]he statute of limitations begins to run as soon as the right
to institute and maintain a suit arises.‖). There exist,
however, various statutory and judge-made rules that operate
to toll the running of the limitations period—that is, ―to stop
[its] running‖; ―to abate‖ it, Black‘s Law Dictionary (9th ed.),
supra, or ―[t]o suspend or interrupt‖ it, Ballentine‘s Law
Dictionary, supra. These tolling doctrines include those for
infancy, see, e.g., Kach v. Hose, 589 F.3d 626, 640–41 (3d
Cir. 2009) (discussing 42 Pa. Cons. Stat. § 5533(b)), the
pendency of a class action which includes absent class
members‘ claims, see In re Cmty. Bank of N. Va., 622 F.3d
275, 299 (3d Cir. 2010) (citing Am. Pipe & Constr. Co. v.
Utah, 414 U.S. 538, 551, 553 (1974)), and the dictates of
equity, see John R. Sand & Gravel Co. v. United States, 552
U.S. 130, 133–34 (2008) (statutes of limitations ―typically
permit courts to toll the limitations period in light of special
equitable considerations‖ unless the particular statute is
deemed ―jurisdictional‖). Time that passes while a statute is
tolled does not count against the limitations period. This can
17
operate to exclude a chunk of time in the middle of the
limitations period—the clock could start, then stop when a
class action is filed, and then start again once certification is
denied. Perhaps more frequently, a tolling rule directs the
court to ignore time at the beginning of the limitations
period—an infant in Pennsylvania is not affected by any
statutory time limit until he achieves the age of majority
(though he could theoretically file suit before that date). This
latter form of tolling has the same practical effect, for
limitations purposes, as a delay in the accrual of the cause of
action: either way, time does not count against the limitations
period until some event external to the cause of action has
occurred.
The discovery rule has been characterized both as
delaying the accrual of a cause of action and as tolling the
running of the limitations period. See 4 Wright & Miller,
Federal Practice and Procedure § 1056 & nn.43.1–43.2 (3d
ed. 2002 & Supp. 2010). The distinction between the two
concepts is ―often confusing,‖ id., but because it makes no
difference for purposes of deciding whether a claim survives
a statute-of-limitations defense, the question has rarely been
analyzed with semantic precision. For instance, none of the
myriad decisions forming the genealogy of Graham I‘s
statement that the discovery rule pertains to accrual defines
precisely what it means for a cause of action to accrue; nor do
they explain why accrual rather than tolling is the relevant
concept. See Disabled in Action of Pa. v. Se. Pa. Transp.
Auth., 539 F.3d 199, 209 (3d Cir. 2008); Romero v. Allstate
Corp., 404 F.3d 212, 222 (3d Cir. 2005); Union Pac. R.R. v.
Beckham, 138 F.3d 325, 330 (8th Cir. 1998); Connors v.
Hallmark & Son Coal Co., 935 F.2d 336, 341 (D.C. Cir.
1991); Dixon v. Anderson, 928 F.2d 212, 215 (6th Cir. 1991);
18
Cada v. Baxter Healthcare Corp., 920 F.2d 446, 450 (7th Cir.
1990); Corn v. Lauderdale Lakes, 904 F.2d 585, 588 (11th
Cir. 1990); Alcorn v. Burlington N. R. Co., 878 F.2d 1105,
1108 (8th Cir. 1989); Keystone Ins. Co. v. Houghton, 863
F.2d 1125, 1127 (3d Cir. 1988); Jensen v. Snellings, 841 F.2d
600, 606 (5th Cir. 1988); Davis v. A.G. Edwards & Sons, Inc.,
823 F.2d 105, 107 (5th Cir. 1987); Alexopulos v. S.F. Unified
Sch. Dist., 817 F.2d 551, 555 (9th Cir. 1987); Cullen v.
Margiotta, 811 F.2d 698, 725 (2d Cir. 1987); Norco Constr.,
Inc. v. King Cnty., 801 F.2d 1143, 1145 (9th Cir. 1986);
Dowty v. Pioneer Rural Elec. Coop., 770 F.2d 52, 56 (6th Cir.
1985); Shapiro v. Cook United, Inc., 762 F.2d 49, 51 (6th Cir.
1985); Howard v. Lockheed-Georgia Co., 742 F.2d 612, 614
(11th Cir. 1984); Metz v. Tootsie Roll Indus., 715 F.2d 299,
304 (7th Cir. 1983); Trotter v. Int’l Longshoremen’s &
Warehousemen’s Union, 704 F.2d 1141, 1143 (9th Cir.
1983); Cline v. Brusett, 661 F.2d 108, 110 (9th Cir. 1981);
Pauk v. Bd. of Trustees, 654 F.2d 856, 159 (2d Cir. 1981);
Singleton v. New York, 632 F.2d 185, 191 (2d Cir. 1980);
Bireline v. Seagondollar, 567 F.2d 260, 263 (4th Cir. 1977);
Young v. Clinchfield R. Co., 288 F.2d 499, 503 (4th Cir.
1961). In the main, these courts simply had no need to
discuss the issue in depth, because it held no practical import
for the results in the cases before them. The Eighth Circuit in
Alcorn, for example, asserted that ―[a] limitations period
accrues when a claimant knows, or should know through an
exercise of reasonable diligence, of the acts constituting the
alleged violation.‖ 878 F.2d at 1108. But this statement
cannot be correct. A limitations period does not accrue: it
neither ―come[s] into existence as an enforceable claim or
right,‖ Black‘s Law Dictionary (9th ed.), supra, nor
―become[s] complete,‖ Ballentine‘s Law Dictionary, supra.
A limitations period runs, and after some period of time it
19
expires. The Seventh Circuit‘s opinion in Cada suggests
similar confusion: the court declared that ―[a]ccrual is the
date on which the statute of limitations begins to run. It is not
the date on which the wrong that injures the plaintiff occurs,
but the date—often the same, but sometimes later—on which
the plaintiff discovers that he has been injured.‖ 920 F.2d at
450. This has it backwards. As we explained above, accrual
is defined in terms of the objective existence of a viable cause
of action, not in terms of whether the limitations clock has
started. It happens to be the case that the limitations period
generally commences once a claim exists, but a running clock
is not the sine qua non of accrual.
Even the Supreme Court has on occasion confused the
two concepts. In Merck & Co. v. Reynolds, --- U.S. ---, 130
S. Ct. 1784, 1793 (2010), the Court described the discovery
rule as ―a doctrine that delays accrual of a cause of action
until the plaintiff has ‗discovered‘ it.‖ As with the court of
appeals cases just discussed, Merck was not concerned with
the precise mechanics of the discovery rule. It dealt only with
the meaning of the word ―discovery‖ in the general federal
statute of limitations, 28 U.S.C. § 1658, and did not consider
the question we now address. Indeed, a close reading of §
1658 reveals that the Court‘s statement regarding the
discovery rule was not grounded in rigorous analysis.
Subsection (a) states, in accordance with ordinary limitations
rules, that ―[e]xcept as otherwise provided by law, a civil
action arising under an Act of Congress . . . may not be
commenced later than 4 years after the cause of action
accrues.‖ Subsection (b) then goes on to enact a partial
discovery rule:
20
Notwithstanding subsection (a), a private right
of action that involves a claim of fraud, deceit,
manipulation, or contrivance in contravention of
a regulatory requirement concerning the
securities laws . . . may be brought not later
than the earlier of—
(1) 2 years after the discovery of the facts
constituting the violation; or
(2) 5 years after such violation.
The statute provides an extended limitations period in a
certain class of cases. It does not alter the date of accrual, but
operates ―[n]otwithstanding‖ the fact that the cause of action
has already accrued. The Court‘s suggestion to the contrary
is neither technically accurate nor necessary to its holding.
Given the unimportance of the difference between
tolling and delayed accrual to the outcomes of the various
decisions cited above and the consequent failure of so many
courts to recognize the distinction, we do not regard their
articulations of the meaning of the discovery rule to be
anything more than nonbinding obiter dicta. Moreover, we
are unaware of any decision considering the precise argument
that the defendants have raised here. We therefore address
what we conclude is the correct meaning of the discovery
rule, and its application to this case, starting from a more or
less blank slate.
This being another question of federal common law,
we turn again to the usual tools of judicial decisionmaking,
beginning with the ordinary legal definitions of the terms
involved. Accrual happens at the moment when events
fulfilling all the elements of a cause of action have transpired.
21
But, with rare and irrelevant exceptions (for instance, a false
imprisonment case where the plaintiff is not actually harmed,
see Restatement (Second) of Torts § 35), knowledge of an
invasion of one‘s rights is not something that a plaintiff must
prove in order to prevail. To be sure, by the time a case goes
to trial the plaintiff will know what has happened to him, or
(if the named plaintiff is incompetent) he will be imputed
with knowledge; but we do not ordinarily dismiss a complaint
pursuant to Rule 12(b)(6) for failing to allege knowledge. In
order to defer accrual, the discovery rule would have to add
an additional component to the substantive definitions of the
claims to which it applies. That simply cannot be right. Rules
regarding limitations periods do not alter substantive causes
of action. Accordingly we do not think the discovery rule
should be read to alter the date on which a cause of action
accrues.
Since it cannot be an accrual doctrine, the discovery
rule must instead be one of those legal precepts that operate to
toll the running of the limitations period after a cause of
action has accrued, as sundry cases have stated. E.g.,
Richison v. Ernest Grp., Inc., 634 F.3d 1123, 1127 n.1 (10th
Cir. 2011) (―The discovery rule, of course, tolls the statute of
limitations until a plaintiff acquires sufficient information,
which, if pursued, would lead to the true condition of
things.‖) (citation and internal quotation marks omitted);
Epstein v. C.R. Bard, Inc., 460 F.3d 183, 187 (1st Cir. 2006)
(―The so-called ‗discovery rule‘ provides that the limitations
period is tolled until ‗events occur or facts surface which
would cause a reasonably prudent person to become aware
that she or he had been harmed.‘‖) (citation omitted); Fine,
870 A.2d at 858 (The discovery rule ―act[s] to toll the running
of a statute of limitations,‖ that is, ―to exclude from the
22
running of the statute of limitations that period of time during
which a party who has not suffered an immediately
ascertainable injury is reasonably unaware he has been
injured, so that he has essentially the same rights as those
who have suffered such an injury.‖); Wal-Mart, 860 A.2d at
319 (―Even after a cause of action accrues, the ‗running‘ of
the limitations period can be ‗tolled‘ in certain limited
circumstances. Under the ‗discovery rule‘ the statute is tolled
where the injury is ‗inherently unknowable and the claimant
is blamelessly ignorant of the wrongful act and the injury
complained of.‘‖) (citations omitted); Burkholz v. Joyce, 972
P.2d 1235, 1236 (Utah 1998) (―[I]n certain circumstances the
discovery rule may operate to toll the period of limitations
until the discovery of the facts forming the basis for the cause
of action.‖) (citations and internal quotation marks omitted).
This conclusion fits with the usual definitions of ―toll‖ and
―accrue,‖ as we have explained, but its appeal does not end
there.
To cast the discovery rule as changing the date of
accrual, so as to delay the onset of interest charges, would
warp its fundamentally plaintiff-friendly purpose. The rule is
an exception to the usual principle that the statute of
limitations begins to run immediately upon accrual regardless
of whether or not the injured party has any idea what has
happened to him. It is grounded in the notion that it is unfair
to deny relief to someone who has suffered an injury but who
has not learned of it and cannot reasonably be expected to
have done so. Treating the discovery rule as altering the date
of accrual would turn it into a means for defendants to protect
themselves from having to fully compensate plaintiffs‘ losses
and disgorge their own wrongful gains. It would, moreover,
give defendants additional incentive to conceal their tortious
23
or otherwise illegal acts: a fraudster would owe no interest on
his purloined cash until discovery of the theft, and would thus
be allowed to benefit from an interest-free loan. This is
emphatically not what the discovery rule is designed to do.
We hold that the ―accrual‖ of a cause of action occurs
at the moment at which each of its component elements has
come into being as a matter of objective reality, such that an
attorney with knowledge of all the facts could get it past a
motion to dismiss for failure to state a claim. The federal
discovery rule then operates in applicable cases to toll the
running of the limitations period. Prejudgment interest,
however, may be awarded in appropriate cases from the
initial accrual date.
Applying this holding to the present case, we conclude
that Graham‘s copyright infringement claim accrued in 1992,
when the first infringement took place. The limitations
period was then tolled until Graham discovered the
infringement in 2004. The District Court acted within its
discretion when it assessed prejudgment interest beginning in
1992.
24
IV
Having thus decided that the verdict is not conscience-
shocking and that the District Court did not abuse its
discretion or commit any other reversible error, we will
affirm its judgment in all respects.2
2
The defendants have renewed their argument that the
discovery rule ought not apply to this case at all. Graham I
decided that issue against them, 568 F.3d at 437, and the
defendants acknowledge that they have raised it here only for
purposes of preserving it for a future certiorari petition. Even were
we to revisit the question, we would reaffirm the prior panel‘s
decision.
25