STC.UNM v. Intel Corporation

Filing 146

MEMORANDUM in Support re 145 MOTION to Bifurcate and Early Trial on Damages by STC. UNM. (Attachments: # 1 Exhibit A - Article, Reverse Bifurcation, # 2 Exhibit B - Litigation Handbook, # 3 Exhibit C - AIPLA Report 2009, # 4 Exhibit D - Intel Amended Resp_ STC First Set of Interr Nos. 1-21)(Pedersen, Steven)

Exhibit A
Article: Reverse Bifurcation

Page 1

2 of 60 DOCUMENTS

Copyright (c) 2006 University of Cincinnati Law Review
University of Cincinnati Law Review
Fall, 2006
75 U. Cin. L. Rev. 213
LENGTH: 22447 words
ARTICLE: REVERSE BIFURCATION
NAME: Drury Stevenson*
BIO: * Associate Professor of Law, South Texas College of Law.
LEXISNEXIS SUMMARY:
... Reverse bifurcation is a trial procedure in which the jury determines damages first, before determining liability. ... In the case of reverse bifurcation, ii depends on the sum of (a) the probability that a trial on damage will yield damages below the minimum amount A requires to maintain his
suit and (b) the probability that damages will be below the threshold that induces the parties to litigate rather than settle (assuming mutual optimism on liability). ... From the standpoint of opportunity costs, therefore, the parties are more likely to have higher opportunity costs by the time they
finish the first phase, giving them a new incentive to settle during the new round of negotiations that
bifurcation provides. ... B. Reverse bifurcation has special effects on settlement because it is inherently harder for either side to predict the outcome of the damages phase than the liability phase. ...
Reverse bifurcation, however, completes the disclosure of the plaintiff's relevant private information
during the first phase, somewhat leveling the playing field between the parties, and ending the prisoner's dilemma described above. ...
HIGHLIGHT: Reverse bifurcation is a trial procedure in which the jury determines damages first,
before determining liability. The liability phase of the trial rarely occurs, because the parties usually
settle once they know the value of the case. This procedure is already being used in thousands of
cases - nearly all the asbestos and Fen-phen cases - but this is the first academic article devoted to
the subject. This article explains the history of the procedure and analyzes why it encourages settlements, simplifies jury instructions, and produces better outcomes for the parties.
TEXT:
[*213]

Page 2
75 U. Cin. L. Rev. 213, *

I. Introduction
The diet drug Fen-phen shrunk a few waistlines, but then bloated judicial dockets with thousands
of product liability cases. Legally, the cases are interesting because almost all of them reverse the
traditional trial order, determining damages first, and liability second. In Fen-phen litigation, most
cases never even proceed to the liability determination because they settle after the first phase, once
the parties know the value of the case.
The same is true for asbestos cases. Originally, asbestos shielded buildings everywhere from
losses caused by fires but left manufacturers exposed to losses from liability over asbestosis and
mesothelioma. Asbestos cases became a staple for many litigation firms; and almost all asbestos
cases, like Fen-phen cases, reverse the order of proceedings and determine damages before liability.
n1 The liability question rarely needs [*214] resolution; the cases usually settle after the determination of damages, which the jury decides before deciding whether the defendant did anything
wrong. n2 Although two of the largest areas of mass tort litigation, as well as a growing number of
cases in other areas of law, n3 [*215] use this nontraditional, bifurcated order for their trials, almost all legal education and academic writing about trial procedure focuses exclusively on the older, liability-then-damages process of unitary trials. "Reverse bifurcation" - the severance of trials
into distinct phases for liability and damages, plus inversion of the usual order - is a significant feature of the modern litigation landscape, but it receives almost no attention in the training of lawyers
or in scholarly research about our legal system. n4
Several rationales exist for separating and inverting the order of issues at trial, many of them
falling under the general rubric of encouraging settlements n5 and yielding better results for the parties. n6 A simple comparison can help clarify the reasons: Our traditional trial system, [*216]
though adversarial, is more like a poker game, where the final prize for the winner is unknown to
the parties when the trial begins, and often increases as the adversarial process goes forward, than it
is like a sports competition for a predetermined prize, title, or trophy. Traditional tort trials are not
like sporting events because the parties usually have different ideas at the outset about the likely
verdict - or range of possible verdicts - and because the stakes of the case can increase as litigation
drags on, new evidence comes in, and juror feeling ratchets upwards. Rather, traditional a tort trial
is more similar to poker, which can have increasing stakes as the parties bet more with each drawn
card, or meet each other's antes; a game that started with bets for a few dollars can reach astronomical amounts by the end. To offer a second analogy, television game shows, similarly, usually have
prize amounts that increase as the contestants move through the rounds; the size of the prize is limited by how quickly the game ends. Traditional unitary trials are therefore more like poker or a
game show. The unpredictable growth of the stakes in a given case can skew the parties' incentives,
and make it harder for anyone to "fold" or settle - just as poker induces players to stay in a game
long beyond what their common sense would dictate.
An insidious aspect is present in many gambling games: as the stakes get higher, the players
must either forfeit (acceding to a loss that may be much larger than the amount they had originally
hoped to win) or raise the stakes. The prospect of forfeiting the entire amount leads many gamblers
to bet much amounts beyond the limits of their disposable funds. The players have too much disincentive to abandon the enterprise even after a game exceeds their risk-tolerance.
The gambler's bind makes casinos controversial from a policy standpoint. Strangely, however,
we run our courts the same way. The same dilemma that can keep players in a poker game after

Page 3
75 U. Cin. L. Rev. 213, *

they would prefer to exit - because the stakes are too high to cut their losses - can keep parties in a
lawsuit where the potential losses (and winnings, for the plaintiff at least) often spiral upward. The
parties' incentives get directed away from settlement. Reverse bifurcation corrects for this effect by
settling the amount at stake first, so that the remainder of the proceedings are more like a sports
competition in that the parties know the final amount of the "prize" to be won or lost. Unlike sports
competitions, however, the parties must invest resources to continue. Thus, they estimate their
chances of winning, and the one most likely to lose forfeits the match, perhaps for a token price.
Despite the great success of this innovative procedure - it produces settlements in almost all
cases in which it is used - most litigation continues to use the traditional trial sequence. Likewise,
most legal [*217] education and scholarship perpetuate it. It is time for change. The first step toward change was the widespread adoption of regular bifurcation, where separate trials or hearings
determine liability and damages. n7 Surveys of judges have found widespread endorsement of bifurcation, because judges believe that it produces better results (i.e., results more correlated to the
merits of the case) n8 and increases judicial economy and efficiency.
This Article is the first in the legal academic literature to analyze the general procedure of reverse bifurcation, n9 and it is long overdue because [*218] the procedure is already in use in
thousands of cases and has been for several years. n10 This Article offers some much-needed theoretical discussion of the practice, and calls for an expanded role for this innovative method of structuring trials. Reverse bifurcation fosters settlements; it generally yields verdicts more reflective of
the merits of a case; and it discourages frivolous litigation driven by spite or desire to protect one's
reputation.
Disproportionate damages sometimes reflect the jury's annoyance with a defendant for repeatedly denying all wrongdoing during the liability phase of the trial, which traditional, unified trials
encourage. In the traditional trial format, both parties have an incentive to invest disproportionately
in the either-or liability determination at the beginning of the case; the defendant hopes to escape
without paying compensation and without reputational damage from admitting liability, and the
plaintiff must fend off the possibility of such a complete loss. This strange alignment of the parties'
investment incentives drives the stakes of the liability question upward even as the reliability of the
jury's [*219] decision sags lower and lower. n11 The traditional trial format polarizes the possible outcomes between a higher chance of zero and a larger verdict otherwise.
The value of an individual case in a traditional trial therefore escalates. The defendant has an
incentive to stridently deny all liability at the outset of a trial, uncertain about the potential damages
but certain that damages will be zero if the plaintiff fails to convince the jury about liability first.
n12 If the jurors, however, are even mildly unconvinced by the denial of culpability, they will resent
the repeated, overstated denials as both dishonest and remorseless. Jury resentment is likely to
augment the damages, increasing the value or stakes of the case for both parties. A defendant who
senses this happening has an even greater incentive to deny wrongdoing, even to the point of risking
the appearance of unreasonableness, so the problem becomes an upward spiral. Similarly, as the
value of the case increases, and the defendant's denials become more strenuous, the stakes of the
liability question itself become higher for the plaintiff as well, creating a disproportionate incentive
(urgency) to counter the denials with more inflammatory, prejudicial evidence early on in the case.
The result in a unified trial is a process workflow where the stakes go up as the reliability of the jury's verdict goes down; and this effect can escalate out of control. n13 Reverse bifurcation eliminates this spiraling or cascading effect. Ex ante resolution of the case's value allows the defendant to

Page 4
75 U. Cin. L. Rev. 213, *

make a more balanced answer to the allegations, lowers the incentive for the plaintiff to offset denials with inflammatory evidence, and reduces the likelihood that the jury is punishing the defendant
merely for denying culpability in court.
[*220] The value of reverse bifurcation is its elimination of uncertainty and unpredictability
about the stakes in any given case, n14 thus allowing parties to assess more accurately the relative
costs of continuing with the litigation. Realistic information about the value of the case does more
than foster settlements. n15 It also produces better settlements, that is, agreements more reflective
of the true value of the case. Ex ante valuation of the case also provides a benchmark with which
parties can compare non-monetary stakes, such as the setting of precedent for future cases, reputational damage to the parties, and the professional reputation of the lawyers. n16 It also facilitates the
management of exhibits and testimonial evidence, making it easier for judges to schedule appropriate time periods for hearings, which should reduce wasted time on the docket. Evidentiary rulings
such as objections or motions in limine are narrower and more precise. Ex ante valuation gives parties intermissions between key parts of the trial to resume negotiations toward settlements, with the
parties having better information in each round. n17
Jury instructions would become more intelligible, focusing on simpler, more discreet questions,
instead of muddling up the instruction about fault with suggestions about various remedies or sanctions they could prescribe. n18 The easier question for untrained jurors to [*221] understand - the
amount of compensation required to make the victim whole - comes first. The more abstract questions of "duty" and "intent" may never come before the jury, because of the great likelihood that the
parties will settle after the first phase. Moreover, reverse bifurcation allows parties make independent choices about jury and bench trials, perhaps opting for a bench trial on questions of legal responsibility after the damages are established. n19 Jurors rarely understand all the jury instructions,
n20 and jury instructions are a common source of reversals by [*222] higher courts, most often
because of concerns that the explanation of the [*223] legal rules could have confused the jury.
n21 Relegating the more arcane liability questions to a second phase that is unlikely to occur, or to a
bench hearing, could therefore eliminate many reversals and retrials.
A bench hearing for the second phase also resolves one concern about reverse bifurcation - that
it might cause prejudice in the jury, because they see all the evidence about the victim's pain and
suffering before they reach the liability decision. n22 This concern is not overly worrisome, however, because it is moot where the jury's role ends after the first phase, which almost always occurs in
reverse bifurcation cases. In the exceptional case that fails to settle after the first phase, a bench
hearing [*224] (or even a second jury) for the second phase resolves the problem, n23 and even
the same jury could be used if the parties agreed. n24
Reverse bifurcation has an additional advantage: it allows better-informed public policy decisions about liability rules. This is, in some ways, the flip side of concerns about manipulating the
jurors' emotions with graphic presentations of the victim's horrific injuries or losses as in a traditional trial. Rather, this means that the court knows the scope of the harm or costs, which is especially important in cases where a policy decision is unavoidable. To use the classic Hand Formula
as an illustration:
B < Lp [*225] B refers to the burden or cost for the defendant (potential or present) of averting the harm; L refers to the victim's total losses; and p refers to the probability of such harm or
losses in similar circumstances. n25 This adjustment for probability is appropriate because more
resources should be devoted to avoiding tragedies that are very likely to occur without such safe-

Page 5
75 U. Cin. L. Rev. 213, *

guards, while very remote harms justify less expenditure. This formula became famous, however,
not for its novelty, but instead for how well it captured what common law courts had been doing all
along, and have done since. n26 "B <" captures the idea of the liability question; "B<" means "liable," and "B>" means "not liable."
Notice, however, that one can only determine the "<" or ">" if we know the amounts, and that
"B" is somewhat inseparable from the ultimately question of liability. The traditional trial sequence
places the decision about "duty" at the beginning, in both unitary trials and with regular bifurcation.
Duty is the "<", usually determined mostly with reference to B, and without a conclusive L. The
fact finder's perception of B is going to be muddled with the ultimate question of whether the defendant was "supposed" to take the precaution or bear the cost. This muddling produces disturbing
results if the L derived from the pre- [*226] determined "B<" is visibly incongruent with the real-world L sitting at the plaintiff's table, the actual victim's losses. The Hand Formula is elegant, but
it forces the conclusion that we should determine L first, as that is often the most concrete number
known. The Hand Formula favors reverse bifurcation. n27
Liability designates who should bear the costs or burden of insuring against certain harms. n28
When we say that no duty or liability exists, we merely put the burden of insuring against that harm
onto the plaintiff (and other potential plaintiffs); when we affix liability on the defendant, we force
parties in the defendant's position to insure against the particular harm. n29 Where to affix liability
becomes a question of where to assign the burden or costs of prevention or insurance against loss.
The amount of loss at stake is useful information for this determination, and therefore determining
the stakes first becomes important. The focus should be on the harms first, and then the legal rules
that address the harm; reverse bifurcation approaches cases in this order. n30
The Coase Theorem suggests that parties would normally reach mutually agreeable settlements
if there were no transaction costs. n31 The role that transaction costs play in settlements implies another benefit of reverse bifurcation. A Coasian view of trials suggests that the specific procedures or
assigned burdens of proof are irrelevant to the outcome of a case, if the parties' information is complete and their transaction costs are low. n32 In other words, the splitting of issues, much less the
order of issues, should have no bearing on decisions to settle a case, or on the jury's verdict, in an
ideal world. In reality, however, parties do not [*227] always reach agreement, because they do
not have adequate information, i.e., too many uncertainties involved n33 or because they cannot
transact easily, as where a corporate defendant adopts a never-settle policy. The latter occurs relatively infrequently but may in fact explain the handful of cases that proceed to the second phase in
bifurcated trials. n34 Eliminating uncertainty or obtaining adequate information may therefore best
explain why so many parties settle after the first phase. n35
The parties enter a trial with different private information: The plaintiff-victim has better information about the value of the injuries, and the defendant has superior information about the degree
of caution or callousness that resulted in the tortious act. Reverse bifurcation forces disclosure of
the plaintiff's best information first, without a reciprocal contemporaneous disclosure of the defendant's inside information; but the procedure also diminishes the value of the defendant's private
information, because the unlikelihood of the second phase makes it less relevant. Modern pre-trial
discovery rules, on the other hand, may result in more disclosure of the defendant's private information because so much of the negligence or breach of duty will turn on objective considerations,
such as the costs of preventing the harm. The value of the victim's injury, on the other hand, in-

Page 6
75 U. Cin. L. Rev. 213, *

cludes a subjective element that is unknowable until the victim or the victim's family appears in
court; to this extent, reverse bifurcation levels a subtle inequality in the pre-trial discovery rules.
Part II of this Article discusses the advantages of bifurcation generally. All of the academic literature to date has focused on regular [*228] bifurcation, so this brief survey will reflect that
emphasis. Part II will also describe cases where courts have used reverse bifurcation. Part III uses
game theory models to illustrate the incentive effects of reversing the order of trial proceedings, and
attempts to demonstrate reverse bifurcation's superiority to both the traditional trial format and regular bifurcation. Part IV focuses on the kinds of cases that make reverse bifurcation more or less
suitable from a public policy standpoint. Part V anticipates and answers some objections regarding
due process concerns, and Part VI concludes.
II. Background: Academic Literature and Cases
A. Academic Commentators on Bifurcation
Scholarly discussion of trial bifurcation began in earnest n36 in the late 1950s and early 1960s,
with advocates presenting it as a technique for expediting trials and reducing court congestion.
Much earlier in our legal history, prior to Blackstone's era, there were actions of "account-render" in
the law courts (predating the formation of equity courts) that used severance or bifurcation, trying
damages and liability separately. n37 Certain affirmative defenses in criminal cases (insanity, duress, consent, etc.) had been tried separately from the prosecutors' charges for generations. The
mid-century scholarly interest in bifurcation coincided with its adoption or frequent use by courts in
certain regions.
Professors Zeisel and Callahan published a seminal empirical study in 1973, focusing on federal
district courts in Illinois. They concluded that regular bifurcation reduced trial time by almost 25%,
reduced jury verdicts for the plaintiffs by 30% (from 42% to 12%), and lowered plaintiff-favoring
settlements by 8% (from 32% to 24%). These results cemented the conventional wisdom in the legal profession that bifurcation favored defendants (because of the lower number of plaintiff [*229]
verdicts) but also fostered judicial economy significantly. n38 Use of the procedure grew slowly but
steadily through the 1960s and 1970s.
A more dramatic change occurred during the 1980s and 1990s. The advent of class actions,
mass torts (especially asbestos), and large punitive damages created an environment where each issue of the case (e.g., liability, causation, damages) could constitute a protracted, complex trial on its
own; n39 some states, such as Pennsylvania, created specialized complex litigation courts to handle
such cases. n40 Recent surveys of judges indicate overwhelming support for bifurcation as a means
of preserving judicial economy, n41 but relatively few judges admit to actually using it in their
courts more than five or ten times in a three-year period. n42 Most of the cases using bifurcation
appear in the complex litigation courts.
Recently, commentators have taken a more nuanced look at bifurcation. n43 The earlier finding
that bifurcation resulted in more defense verdicts has been tempered with the observation that verdicts for plaintiffs, while less frequent, are significantly larger. In other words, defendants have better odds of obtaining a verdict in their favor, but if they lose at the liability phase, the risk of high
damage awards (both compensatory and punitive) is much greater. n44
[*230] Juror compromise appears to be one cause of this pattern. In a traditional, unified trial,
a majority of jurors favoring the plaintiff can persuade the dissenters to agree to a lower damage