Ward v. General Motors L.L.C.

Filing 306

OPINION AND ORDER [Regarding the Parties' Motions in Limine and the Admissibility of Plaintiffs Other Similar Incident Evidence] re: (3941 in 1:14-md-02543-JMF, 237 in 1:14-cv-08317-JMF), MOTION in Limine No. 4 to Exclude Evidence of P ast Tobacco Use, filed by Dennis R. Ward, (3924 in 1:14-md-02543-JMF, 220 in 1:14-cv-08317-JMF) MOTION in Limine No. 33 To Exclude Dennis Ward's August 2014 Ignition Switch Videos, filed by General Motors LLC, (229 in 1:14-cv -08317-JMF, 3933 in 1:14-md-02543-JMF) MOTION in Limine to Exclude Evidence Regarding Spoliation of Sensing and Diagnostic Module (SDM) Data, filed by Dennis R. Ward, (233 in 1:14-cv-08317-JMF, 3937 in 1:14-md-02543-JMF) MOTION in Limine No. 3 to Permit Plaintiff to Use Demonstration Videos During His Case in Chief, filed by Dennis R. Ward, (223 in 1:14-cv-08317-JMF, 3927 in 1:14-md-02543-JMF) MOTION in Limine No. 1 re Evidence of March 27, 2014 Traffic Citati on, filed by Dennis R. Ward.Ward's First Motion in Limine and New GM's Thirty-Third Motion in Limine are granted, Ward's Third and Fourth Motions in Limine are denied, and the Court reserves judgment on Ward's Second Motion in Limine pending trial. In addition, Ward's application to admit OSI evidence is granted in part and denied in part. Specifically, Ward may introduce evidence of fifty-six of the sixty OSIs (Pl.'s OSI Discl., Rows 1-56) for purposes of proving notice only. Ward may not introduce evidence of any OSIs (except the one in his own vehicle (id. at Row 57), which may or may not constitute OSI evidence in the first instance) to prove the existence of a defect or causation. Finally, the three unverified consumer complaints (id., Rows 58-60) will not be admitted for any purpose. The Clerk of Court is directed to terminate 14-MD-2543, Docket Nos. 3761, 3924, 3927, 3933, 3937, and 3941 and 14-CV-8317, Docket Nos. 220, 223, 229, 233, and 237, and as further set forth herein. (Signed by Judge Jesse M. Furman on 6/9/2017) (ras)

UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF NEW YORK
-----------------------------------------------------------------------------x
IN RE:

06/09/2017

GENERAL MOTORS LLC IGNITION SWITCH LITIGATION

14-MD-2543 (JMF)

This Document Relates To:
Ward v. General Motors LLC, 14-CV-8317 (JMF)
-----------------------------------------------------------------------------x

OPINION AND ORDER

JESSE M. FURMAN, United States District Judge:
[Regarding the Parties’ Motions in Limine and the
Admissibility of Plaintiff’s Other Similar Incident Evidence]
The next bellwether trial in this multidistrict litigation (“MDL”), brought by Plaintiff
Dennis Ward in connection with his accident while driving his 2009 Chevrolet HHR on March
27, 2014, and familiarity with which is presumed, is scheduled to begin on July 10, 2017. Now
pending are five disputed motions in limine, to wit:
•

•

Ward’s Second Motion, which seeks to exclude evidence and argument that he failed
to preserve, destroyed, or otherwise spoliated sensing and diagnostic module
(“SDM”) data from his car (see Docket No. 3933);

•

1

Ward’s First Motion, which seeks to exclude argument and evidence concerning the
fact that he received a traffic citation for the accident at issue, as well as
characterizations of the citation as a criminal offense (see Docket No. 3927); 1

Ward’s Third Motion and New GM’s Thirty-Third Motion, 2 which concern the
admissibility of two demonstrative videos made by Ward and his son (see Docket
Nos. 3924, 3937); and

Unless otherwise noted, all docket references are to the MDL docket, 14-MD-2543.

2

For reasons that are not entirely clear, plaintiffs in this MDL have adopted the practice of
numbering their motions in limine separately for each trial, restarting each time with the number
one. By contrast, New GM continues in each trial from whatever number it finished at in the
prior trial. Thus, New GM’s Thirty-Third Motion in Limine is its thirty-third motion in limine in
the MDL overall, not its thirty-third motion with respect to Ward.

1

•

Ward’s Fourth Motion, which seeks to preclude evidence and argument regarding his
history of smoking. (See Docket No. 3941).

In addition, Ward seeks an advanced ruling that he may introduce certain other similar incident
(“OSI”) evidence — namely, evidence concerning sixty other crashes allegedly attributable to
the ignition switch defect in certain General Motors cars. (Docket No. 3761 (“Pl.’s OSI Br.”)).
For the reasons stated below, Ward’s First Motion and New GM’s Thirty-Third Motion
are GRANTED, Ward’s Third and Fourth Motions are DENIED, and the Court reserves
judgment on Ward’s Second Motion pending trial. In addition, Ward’s application to admit OSI
evidence is GRANTED in part and DENIED in part. 3
A. Ward’s First Motion in Limine
In his first motion in limine, Ward argues principally that Arizona law precludes New
GM from introducing evidence (or argument) concerning the fact that he was cited for a civil
traffic infraction in connection with the accident at issue and ultimately admitted responsibility
for a civil traffic violation. (Docket No. 3928, at 5-6). New GM does not appear to dispute that
evidence of the citation and Ward’s admission of responsibility would be inadmissible under
Arizona law, which provides that “an admission of the allegation of a civil traffic complaint or a
judgment on the complaint is not evidence of negligence in a civil . . . proceeding.” Ariz. Rev.
Stat. Ann. § 28-1599; see also, e.g., Ingrum v. Tuscon Yellow Cab. Co., 642 P.2d 868, 872 (Ariz.
Ct. App. 1981) (“The fact of citation or non-citation of a driver … is inadmissible in an action
for negligence.”). Instead, New GM contends that, under Erie v. Tompkins, 304 U.S. 64 (1938),

3

Some of the issues decided here may be affected — or even mooted — by the Court’s
decisions on the parties’ Daubert motions and dispositive motions (all of which are now fully
briefed and under advisement). Needless to say, the Court’s rulings are subject to modification
— or even reconsideration — as appropriate in light of subsequent decisions.

2

the admissibility of Ward’s citation and admission of responsibility is a function of federal law
— namely, the Federal Rules of Evidence — rather than state law and that federal law allows
such evidence “as party admissions and/or pursuant to the public records exception.” (Docket
No. 3980 (“Def.’s MIL 1 Opp’n”), at 4-5 (citing cases)).
The Court concludes that Ward has the better of the argument. Although there is general
consensus that the Federal Rules of Evidence ordinarily govern in diversity cases, courts disagree
about whether that is because Erie applies or because the Federal Rules of Evidence (at least in
their original form) were enacted by Congress. See, e.g., Sims v. Great Am. Life Ins. Co., 469
F.3d 870, 877-81 (10th Cir. 2006) (holding that Erie does not apply, but that the original Federal
Rules of Evidence govern in diversity cases because they were enacted by Congress); Feldman v.
Allstate Ins. Co., 322 F.3d 660, 666 (9th Cir. 2003) (applying Erie). In either case, though, the
overwhelming weight of authority holds that “[w]here a state law excludes certain evidence in
order to effect substantive policy considerations,” such evidence should not be admitted by a
federal court sitting in diversity. Sims, 469 F.3d at 881 (emphasis added) (reaching that
conclusion on the ground that Rule 401 of the Federal Rules of Evidence itself “acts to exclude
the evidence since the proposition for which the evidence is submitted is not properly provable
and, therefore, irrelevant to the claim”); see Feldman, 322 F.3d at 666 (reaching the same
conclusion because, “[p]ursuant to Erie and its progeny, federal courts sitting in diversity apply
state substantive law and federal procedural law”); Hottle v. Beech Aircraft Corp., 47 F.3d 106,
110 (4th Cir. 1995) (same); Potts v. Benjamin, 882 F.2d 1320, 1324 (8th Cir. 1989) (same).
More specifically, if the language and policy of a state rule “point to achieving fair, accurate, and
efficient resolutions of disputes, the rule is procedural” and does not apply in diversity cases; by
contrast, “[i]f . . . the primary objective” of a state rule “is directed to influencing conduct

3

through legal incentives, the rule is substantive” and does. Sims, 469 F.3d at 883.
Applying those standards here, the Court concludes that the Arizona statute barring
admission of a civil traffic complaint or a judgment on the complaint is substantive and thus
applies. That is, the purpose of the statute is not “to achieve accuracy, efficiency, and fair play in
litigation, without regard to the substantive interests of the parties.” Id. at 882. Instead, the
language and policy of the Arizona rule make plain that its “primary objective” is to “influenc[e]
conduct through legal incentives,” id. at 883 — namely, to encourage the quick resolution of
traffic citations by allaying motorists’ fears that if they pay a fine, that payment will be used
against them in later proceedings. See, e.g., 22A Fed. Prac. & Proc. Evid. § 5201 (2d ed.)
(“[M]any states have adopted statutes making guilty pleas in motor vehicle prosecutions
inadmissible in subsequent civil litigation. The legislators approved these statutes to encourage
defendants to plead guilty rather than to turn the traffic court into a battleground or discovery site
for ensuing civil litigation.”). It follows that the rule applies in this case and precludes New GM
from using evidence of Ward’s citation and admission of responsibility to prove his negligence.
Cf. Feldman, 322 F.3d at 666–67 (applying California law to exclude recorded conversations);
Gardner v. Chrysler Corp., 89 F.3d 729, 736 (10th Cir. 1996) (applying a Kansas statute
prohibiting the admission of seat belt nonuse evidence); Potts, 882 F.2d at 1324 (applying an
Arkansas statute excluding evidence concerning the failure to secure minors in child seats).
In arguing otherwise, New GM relies almost exclusively on the Third Circuit’s decision
in Rain v. Pavkov, 357 F.2d 506 (3d Cir. 1966), which held that a guilty plea to reckless driving
was admissible in a diversity action despite a Pennsylvania law to the contrary. (New GM cites
several other cases and a treatise as well, but all of those authorities rely, in turn, on Rain.
(Def.’s MIL 1 Opp’n 4-5)). But Rain was decided before the Federal Rules of Evidence were

4

enacted and turned, instead, on application of Rule 43(a) of the Federal Rules of Civil Procedure.
Under that Rule, evidence that met “any one of [three] prescribed tests [was] admissible in the
federal courts. In the event of a conflict between state and federal law, the rule which favor[ed]
admissibility . . . [was] followed.” Rain, 357 F.2d at 510. Thus, Rain’s analysis and conclusion
have only limited applicability and utility today, and the Court declines to follow the decision.
Instead, applying the more modern federal law discussed above, the Court concludes that the
Arizona statute barring evidence of a civil traffic complaint and a judgment on such a complaint
applies to this case. It follows that Ward’s first motion in limine must be and is granted.
B. Ward’s Second Motion in Limine
Next, Ward moves to exclude evidence and argument that he failed to preserve,
destroyed, or otherwise spoliated SDM data from his car. (Docket No. 3934 (“Pl.’s MIL 2
Br.”)). The SDM is a device that records certain data in connection with an event such as a
crash. (Docket No. 3935 (“Def.’s MIL 2 Decl.”), Ex. 2 (“Pitman Rpt.”), at 7-8). Where, as in
this case, a crash is a “non-deployment event” (that is, it is below the threshold where the airbags
are designed to deploy), the data are “soft written” to the SDM — meaning that the data are not
permanently stored, but can be overwritten by other events or the absence of subsequent events.
(Id.). Soft-written data in a 2009 Chevrolet HHR’s SDM are apparently overwritten after 250
ignition cycles. (Id.). In this case, Ward’s car was taken after his March 2014 accident to a
repair facility, where it was repaired. (Def.’s MIL 2 Decl., at Ex. 1 (“Ward Depo. Tr.”), at 15,
104). Ward resumed driving his car in August or September 2014, and then drove it for over two
years (during which time, it is safe to assume, the car exceeded 250 ignition cycles) until, in
October 2016, the SDM was finally downloaded. (Id. at 113; Pitman Rpt. 6) At that point, the
SDM did not contain any relevant event data. (Pitman Rpt. 6). Notably, assuming that data from

5

Ward’s accident was captured by the SDM in the first place, there is no evidence establishing
when the data was overwritten. (Id.; Pl.’s MIL 2 Br. 4 n.21).
Ward argues that, on these facts, New GM cannot proffer evidence to support a “finding
of spoliation that would warrant the imposition of ‘severe’ sanctions, such as an adverse
inference instruction or dismissal.” (Pl.’s MIL 2 Br. 1). Significantly, New GM more or less
concedes the point, clarifying that it “is not seeking spoliation sanctions” (at least “at this
stage”). (Docket No. 3986 (“Def.’s MIL 2 Opp’n”), at 2). That is for good reason, because —
among other things — there is no evidence that the SDM data would have been favorable to New
GM. See In re: Gen. Motors LLC Ignition Switch Litig., No. 14-MD-2543 (JMF), 2015 WL
9480315, at *3 (S.D.N.Y. Dec. 29, 2015) (“Scheuer Spoiliation Op.”) (denying New GM’s
motion for spoliation sanctions on similar grounds). Nevertheless, New GM “seeks to introduce
the facts related to plaintiff’s destruction of the SDM data” on the ground that it is “otherwise
relevant.” (Def.’s MIL 2 Opp’n 2). Specifically, it contends that the evidence is relevant in
three ways: (1) to the cross-examination of Ward’s expert “as it undermines the reliability of
both his reconstruction assumptions and his opinions on the severity of the crash”; (2) to prevent
confusion or prejudice arising from the admission of OSI evidence involving accidents where
SDM data was recovered; and (3) to impeach Ward, because “it directly contradicts his long-held
claim that his airbags should have deployed” and because his “decision to continue driving the
car enough times to override the SDM data demonstrates that plaintiff did not believe his car was
defective or caused his crash.” (Id.). Additionally, New GM argues that, even if an adverse
inference sanction is not justified, a permissive adverse inference instruction — “that the jury
may (but need not) draw adverse inferences based on the destruction of the SDM data” — is
“appropriate.” (Id. at 2-3).

6

Significantly, however, there is no evidence in the current record suggesting that Ward
knew about the SDM, let alone that the SDM could have provided data relating to his accident or
that such data could be destroyed through continued use of the car. In the absence of such
evidence (not to mention, the absence of evidence suggesting that relevant data existed in the
first place and would have been favorable to New GM), the mere fact that Ward failed to
preserve the SDM data has little or no relevance and probative value. See Fed. R. Evid. 401-402.
And in the absence of such evidence, permitting New GM to suggest through questions or to
explicitly argue that the jury should draw an adverse inference based on the destruction of the
SDM data would plainly risk unfair prejudice to Ward. See Fed. R. Evid. 403. New GM’s
arguments to the contrary are unpersuasive because they confuse the mere absence of the data
with the possible reasons for that absence. That is, to cross-examine Ward’s expert and
distinguish Ward’s accident from any other incidents for which SDM was available, it does not
matter why the SDM data is now unavailable; all that matters is that the data is unavailable.
Similarly, New GM is plainly entitled to cross-examine Ward about the fact that he continued to
drive his car despite the accident and to argue from that fact that he did not believe the car was
defective or dangerous. To make that argument, however, New GM need not introduce evidence
or suggest that, by continuing to use the car, Ward spoliated the SDM data.
Thus, on the current record, the Court agrees with Ward that New GM should be
precluded from introducing evidence or arguing that he failed to preserve, destroyed, or
otherwise spoliated SDM data from his car and that the jury should be told — “in a neutral
manner (ideally by stipulation) that does not suggest that either party was responsible,” Scheuer
Spoiliation Op., 2015 WL 9480315, at *5 n.6 — only that the SDM data is unavailable. That
does not resolve the matter, however. Ward contends that New GM should be held to the current

7

record, as it had an opportunity to ask him questions about the SDM during his deposition and
did not do so. (Docket No. 4022, at 1, 3). But he cites no authority for the proposition that a
party is limited to asking questions at trial on subjects that it inquired about at a pretrial
deposition; nor does he affirmatively proffer, let alone establish, that he did not know about the
SDM data and the potential for its loss through continued use of his car. Accordingly, the Court
reserves judgment until trial on whether to grant Ward’s second motion in limine in all respects
(and, relatedly, on whether to grant New GM’s request for a permissive adverse inference
instruction). At trial, the Court will give New GM a limited opportunity to establish, in a manner
that avoids any risk of prejudice, that Ward understood the significance of the SDM data and the
potential for its loss. The parties should confer with respect to how New GM should be
permitted to do so — by questioning Ward outside the presence of the jury; asking neutral, nonleading questions in the presence of the jury; or some other means — and be prepared to discuss
the matter at the final pretrial conference.
C. Ward’s Third Motion in Limine and New GM’s Thirty-Third Motion in Limine
Ward’s Third Motion and New GM’s Thirty-Third Motion concern the same issue: the
admissibility of two videos made by Ward and his son around the time Ward retained counsel. 4
The first video depicts Ward using his key, on his keychain, to start his 2009 HHR and then,
several times, intentionally applying an unspecified amount of force (in an action the Court
would describe as “flicking”) to rotate the key out of the “run” position and cause the car to lose

4

Needless to say, the result of two motions addressing the same issue is six memoranda of
law covering largely the same ground rather than the usual three. In future bellwether cases, the
parties should meet and confer before filing motions in limine to minimize, if not eliminate, such
duplicative briefing. The parties should confer and submit a proposed order to that effect if they
believe that one is necessary.

8

power. (Docket No. 3939 (“Pl.’s MIL 3 Decl.”), at Ex. 2). The second video involves a similar
“demonstration” by Ward’s son in his Chevrolet Silverado, only the flicking never results in the
key rotating out of the “run” position. (Docket No. 3938 (“Pl.’s MIL 3 Br.”), at 3; Pl.’s MIL 3
Decl., Ex. 3). Ward expressly notes that he does not seek to offer the videos as substantive
evidence (Docket No. 3989 (“Pl.’s MIL 33 Opp’n”), at 8), and for good reason: As out-of-court
statements offered to prove the truth of a matter asserted, the videos would plainly constitute
hearsay. See, e.g., United States v. Martinez, 588 F.3d 301, 311 (6th Cir. 2009). Instead, Ward
seeks to use the videos as demonstrative exhibits “for the purpose of illuminating” his testimony.
(Pl’s MIL 33 Opp’n 8). Specifically, he contends that the videos “demonstrate the propensity of
Mr. Ward’s ignition switch to rotate in non-severe conditions, as well as the physical
characteristics of Mr. Ward’s key chain and Mr. Ward’s seating position in his car. Mr. Ward
will testify to each of these facts as a live witness at trial; the demonstrative videos will simply
help to clarify this testimony for the jury.” (Id. at 8-9).
These arguments are disingenuous at best. Ward does not need the videos to depict the
physical characteristics of his keychain and seating position in the car; as New GM notes, he
may, without objection, use still frames from the first video to do both. (Docket No. 3990
(“Def.’s MIL 3 Opp’n”), at 8). And using the videos to “demonstrate” the propensity of Ward’s
switch to rotate (Pl.’s MIL 33 Opp’n 8), to “refute” New GM’s contentions about the probability
of inertial rotation (Docket No. 4024, at 2 n.1), or to “prov[e] that the ignition switch will, in
fact, rotate . . . in less-than-extreme conditions” (id. at 4) is to use them to prove the truth of a
matter asserted in violation of the rules against hearsay; labeling the videos “demonstrative”
rather than “substantive” exhibits does not render that use any more permissible. And in any
event, the videos suffer from a host of other problems. First, the video of the Silverado is totally

9

irrelevant; how the ignition switch in Ward’s car compared to the ignition switch in his son’s
vehicle is not relevant to any issue in dispute. Second, the video of Ward’s car is only
marginally probative, as it depicts him intentionally applying force to his keychain while Ward
alleges that his accident was caused by inadvertent rotation of the ignition switch in his car.
Third, allowing the jury to see the video risks confusion and unfair prejudice, as the amount of
force Ward applied to his keychain is unknown. And finally, the question of whether Ward’s
switch had sufficient torque resistance (the subject of the videos and any testimony about them)
is the province of expert, not lay, testimony. Ward is concededly not an expert, and the “tests”
depicted in the videos were plainly not done in accordance with scientific standards.
Notably, the videos do not come close to meeting the standard for admission cited in
Ward’s own briefs. Quoting from Bellinger v. Deere & Co., 881 F. Supp. 813, 816 (N.D.N.Y.
1995), Ward insists that “[d]emonstration evidence is admissible at the discretion of the court
where the conditions of the demonstration are similar to those at the time in question and where
it is proffered merely to show the mechanical principles involved in the operation of the
equipment in question.” (Pl.’s MIL 3 Br. 5). As Ward himself concedes, however, the
conditions in the video are plainly dissimilar to those at the time of the accident as the car was
not even moving (see Pl.’s MIL 33 Opp’n 13), and Ward does not proffer the videos to show the
mechanical principles involved in the operation of the ignition switch. In fact, he admitted in his
deposition that he made the videos to demonstrate that his switch was “messed up” and thus
defective. (Pl.’s MIL 3 Decl., Ex. 1, at 262). Nor are the other cases upon which Ward relies
any more helpful, as they did not involve anything of the sort proposed here — namely,
presentation through a lay witness of self-serving videos depicting unscientific demonstrations
made for the purpose of litigation in order to prove precisely what the plaintiff needs to prove in

10

that litigation. See, e.g., Clevenger v. CNH Am., LLC, 340 F. App'x 821, 825 (3d Cir. 2009)
(affirming a decision to admit a demonstrative video depicting the defendant’s expert operating
the machinery at issue); Edwards v. ATRO SpA, 891 F. Supp. 1085, 1086-88 (E.D.N.C. 1995)
(allowing a videotaped demonstration by the plaintiff’s expert in part to familiarize the jury with
“a piece of commercial equipment many have never seen,” noting that the case did “not involve
a product such as an automobile, with which a jury would be uniformly familiar”); Veliz v.
Crown Lift Trucks, 714 F. Supp. 49, 51-54 (E.D.N.Y. 1989) (approving the use of a
demonstrative video by the defendant’s expert to demonstrate “the physical and mechanical
principles involved in the braking of lift trucks” and a “brief, live demonstration . . . to
familiarize the jury with the operations of lift trucks in general”).
In short, the Court concludes that the videos (and, relatedly, any testimony about the
“tests” that they depict) should not be admitted at trial. Accordingly, New GM’s Thirty-Third
Motion in Limine is granted, and Ward’s Third Motion in Limine is denied.
D. Ward’s Fourth Motion in Limine
Ward’s fourth and final motion in limine seeks to preclude evidence and argument
regarding his history of smoking. (See Docket No. 3941). In his opening memorandum of law,
Ward contends that such evidence is irrelevant and prejudicial and should be precluded pursuant
to Rules 401, 402, and 403. (See Docket No. 3942, at 3-6). Those arguments, however, are
patently meritless, substantially for the reasons provided by New GM in its opposition
memorandum of law. (See Docket No. 3983, at 2-6). Specifically, evidence of Ward’s smoking
is plainly relevant to his life expectancy and his claim for future damages, not to mention
relevant to and probative of the cause of his alleged current and future symptoms (and, relatedly,
of the reliability and credibility of Ward’s expert’s opinion that “all” of his disabilities are

11

“causally related to the accident”). (See id. at 2-6). Additionally, like the plaintiff in Cockram,
Ward “provides no reasons why evidence of [his] tobacco use generally would be . . . so
prejudicial as to substantially outweigh any probative value,” particularly since the alleged
prejudice is a product of what makes the evidence so probative: that smoking is unhealthy. In re:
Gen. Motors LLC Ignition Switch Litig., No. 14-MC-2543 (JMF), 2016 WL 4410008, at *5
(S.D.N.Y. Aug. 18, 2016) (citing Fed. R. Evid. 403). Perhaps revealingly, in his reply
memorandum of law, Ward switches gears somewhat, asserting that New GM lacks admissible
evidence with respect to the effects of smoking on health and life expectancy. (See Docket No.
4021, at 1 n.1, 4). But that argument is both procedurally improper — a party may not raise new
arguments in a reply, see, e.g., Docket No. 1770, at 2 — and “premature. If and when New GM
seeks to introduce evidence of [Ward’s] tobacco use at trial, [Ward] may make appropriate
objections.” Gen. Motors LLC Ignition Switch Litig., 2016 WL 4410008, at *5. Thus, Ward’s
Fourth Motion in Limine is denied without prejudice to objection to specific evidence at trial.
E. The Admissibility of OSI Evidence
Finally, Ward seeks to admit evidence of fifty-eight other incidents to prove New GM’s
knowledge and notice of the alleged defect and eight other incidents to prove the existence or
magnitude of the defect as well as causation. (Pl.’s OSI Br. 1). The applicable legal standards,
discussed by the Court in connection with three earlier bellwether trials in this MDL, are
undisputed. See In re: Gen. Motors LLC Ignition Switch Litig., No. 14-MD-2543 (JMF), 2016
WL 4410030 (S.D.N.Y. Aug. 18, 2016) (“Cockram”); In re: Gen. Motors LLC Ignition Switch
Litig., No. 14-MD-2543 (JMF), 2016 WL 796846, at *2 (S.D.N.Y. Feb. 25, 2016)
(“Barthelemy/Spain”); In re: Gen. Motors LLC Ignition Switch Litig., No. 14-MD-2543 (JMF),
2015 WL 9463183, at *1 (S.D.N.Y. Dec. 28, 2015) (“Scheuer”). Those standards provide that,

12

in a product liability case such as this, OSI evidence may be admitted to prove “negligence, a
design defect, notice of a defect, or causation.” Hershberger v. Ethicon Endo-Surgery, Inc., No.
10-CV-0837, 2012 WL 1113955, at *2 (S.D. W. Va. Mar. 30, 2012); see also McClure v.
Walgreen Co., 613 N.W.2d 225, 234 (Iowa 2000) (holding that OSI evidence “was obviously
relevant to the punitive-damage issue of willful and wanton conduct”). But before such evidence
“may be admitted for any purpose, the proponent must establish [the prior accidents’] relevance
by showing that they occurred under the same or substantially similar circumstances as the
accident at issue.” Schmelzer v. Hilton Hotels Corp., No. 05-CV-10307 (JFK), 2007 WL
2826628, at *2 (S.D.N.Y. Sept. 24, 2007); see also Barker v. Deere & Co., 60 F.3d 158, 162 (3d
Cir. 1995) (noting that every Court of Appeals that has considered the admissibility of prior
accidents in products liability cases has applied the substantial similarity standard). “Whether a
prior accident occurred under ‘substantially similar’ conditions necessarily ‘depends upon the
underlying theory of the case, and is defined by the particular defect at issue.’” Lidle v. Cirrus
Design Corp., 505 F. App’x 72, 74 (2d Cir. 2012) (summary order) (quoting Guild v. Gen.
Motors Corp., 53 F. Supp. 2d 363, 367 (W.D.N.Y. 1999)).
Significantly, the requisite degree of similarity varies according to the purpose for which
OSI evidence is offered. For example, “[e]vidence proffered to illustrate the existence of a
dangerous condition necessitates a high degree of similarity because it weighs directly on the
ultimate issue to be decided by the jury.” Four Corners Helicopters, Inc. v. Turbomeca, S.A.,
979 F.2d 1434, 1440 (10th Cir. 1992). Similarly, where OSI is offered to prove causation, courts
tend to consider multiple factors — namely, whether “(1) the products are similar; (2) the alleged
defect is similar; (3) causation related to the defect in the other incidents; and (4) exclusion of all
reasonable secondary explanations for the cause of the other incidents.” Watson v. Ford Motor

13

Co., 389 S.C. 434, 453 (2010); see also 1 McCormick on Evid. § 200 (7th ed. 2013) (noting that
courts are more likely to allow OSI evidence to show causation “when the defendant contends
that the alleged conduct could not possibly have caused the plaintiff’s injury”). By contrast, the
substantial similarity standard is “relaxed” where OSI evidence is offered to show notice; that is,
“the similarity in the circumstances of the accidents can be considerably less than that which is
demanded when the same evidence is used for one of the other valid purposes.” Schmelzer, 2007
WL 2826628, at *2. As the substantial similarity inquiry is “fact-specific,” a “district court is
owed considerable deference in its determination of substantial similarity.” Bitler v. A.O. Smith
Corp., 391 F.3d 1114, 1126 (10th Cir.), as clarified on reh’g, 400 F.3d 1227 (10th Cir. 2004). If
substantial similarity is established, “[a]ny differences in the accidents . . . go to the weight of the
evidence” rather than its admissibility. Four Corners, 979 F.2d at 1440.
In the present case, the parties dispute how these principles apply to several categories of
alleged OSI evidence. 5 The Court will address each category in turn.
National Highway Traffic Safety Administration (“NHTSA”) Customer Complaints.
First, Ward seeks to admit — to prove notice, the existence of defect, and causation — three
unverified complaints to NHTSA. (Pl.’s OSI Br. 22-26; see Pl.’s OSI Discl., Rows 58-60). Such
reports, however, are hearsay, and thus inadmissible as a matter of course. See, e.g., Landis v.
Jarden Corp., 5 F. Supp. 3d 808, 815 (N.D.W. Va. 2014); Hendricks v. Ford Motor Co., No. 12CV-71, 2012 WL 4478308, at *2 (E.D. Tex. Sept. 27, 2012); Guild v. Gen. Motors Corp., 53 F.

5

Ward includes in his application a prior incident in his own vehicle. (Pl.’s OSI Br. 16;
see Pl.’s OSI Br., Ex. 1 (“Pl.’s OSI Discl.”), Row 57). The Court need not decide whether that
incident is subject to analysis under the OSI principles set forth above, as New GM does “not
dispute that [P]laintiff can offer factual testimony about this incident at trial.” (Docket No. 3831
(“Def.’s Opp’n Br.”), at 15).

14

Supp. 2d 363, 368 (W.D.N.Y. 1999). Additionally (and relatedly), as the Court observed during
the Scheuer trial, such complaints are not the kind of materials upon which an expert would
reasonably rely. (Scheuer Trial Tr. at 718). Ward argues otherwise by alleging that these reports
were “relied upon by New GM’s outside consultant,” the Virginia Tech Transportation Institute
(“VTTI”), which “characterized the incidents as involving inadvertent ignition switch rotations
due to knee-key contact.” (Pl.’s OSI Br. 16). But that argument does not overcome the hearsay
problem. Moreover, as New GM points out, VTTI identified the three complaints when it
performed a “database search” for “knee-key related incidents,” but the Institute never actually
verified or analyzed the incidents following the search. (Def.’s Opp’n Br. 20). Additionally, and
as VTTI itself observed, there are reasons to doubt the reliability of the complaints: All three
were filed after the initial recall, and two of the three were made more than a year after the
incidents to which they related. (Id. at 21). Finally, and in any event, the three complaints lack
sufficient information about the relevant incidents for the Court to conduct the requisite
substantial similarity analysis, even under the more relaxed standard applicable to notice.
Accordingly, Ward may not offer evidence pertaining to these three incidents for any purpose.
Existence of a Defect and Causation. Next, Ward seeks to admit evidence of four
incidents (Melton, Caban, Gamage, and Paluszek) to prove the existence of a defect and
causation. (Pl.’s OSI Br. 22-26; see Pl.’s OSI Discl., Rows 20, 36, 44, 48). Applying the
heightened standard of substantial similarity that applies to evidence offered to prove the
existence of a defect and causation, the Court agrees with New GM that the incidents — none of
which involved the ignition switch actually at issue in this case — are inadmissible for those
purposes. First, the Melton and Caban incidents are easy calls. The former involved not only a
different ignition switch, but a different kind of vehicle (a 2005 Chevrolet Cobalt) altogether.

15

(See Def.’s Opp’n Br. 16-17). Additionally, the circumstances of the accidents (for example, in
Ward’s case, the road was dry, while Melton involved wet conditions) and the stature of the
drivers (Ward is a 6 foot, 218 pound male while Melton was a 29-year-old female) differed
significantly. (See id.). Against these points of dissimilarity, Ward asserts that both incidents
involved “rough road conditions” and “a key that had a slotted keyhead design and several
keychain attachments” (Def.’s Opp’n Br. 24-25), but those similarities are not enough to satisfy
the heightened standard applicable to proving causation and the existence of a defect.
Caban did involve the same vehicle as the present case, but an earlier — and different —
version of the ignition switch, the very feature of the vehicle alleged to be defective here. (Id. at
17). Moreover, Ward provides no factual support for his assertion that the two incidents were
“substantially similar” (Pl.’s OSI Br. 24-25) — a shortfall that is all the more notable given that
the Court excluded evidence relating to the Caban incident in the Barthelemy/Spain trial on
similar grounds. See Barthelemy/Spain, 2016 WL 796846, at *4; see also, e.g., Ruminer v.
General Motors Corp., 03-CV-0349 (GTE), 2006 WL 287945, at *14 (E.D. Ark. Feb. 6, 2006)
(finding that Plaintiff had failed to meet his burden when he made “no effort to demonstrate or
explain how” the OSIs were substantially similar to the alleged product defects in his case);
Great N. Ins. Co. v BMW of N. Am., LLC, 02-CV-1153, 2015 WL 3936229, at *10 (S.D. Ohio
June 26, 2015) (excluding OSI evidence for purposes of proving causation because the plaintiffs
had failed to establish that “they [would] lay the proper foundation to prove the prior incidents
were sufficiently similar”).
By contrast, the Gamage and Pluszek incidents both involved the same vehicle as the
incident in this case, and Ward does provide some information concerning the two. But the
differences between the incidents and Ward’s preclude use of the earlier incidents to prove the

16

existence of a defect or causation. Most significantly, the vehicles in the Gamage and Pluszek
incidents featured an earlier — and different — ignition switch from the vehicle at issue here.
(Pl.’s OSI Br. 2-5; Def,’s Opp’n Br. 2-4). Whether Ward’s switch can be said to suffer from the
same defect as the earlier switch and, if so, whether his accident was caused by that defect are
perhaps the central issues in this case. (Pl.’s OSI Br. 3-7, 10-12; Def.’s Opp’n Br. 10-15). Ward,
however, provides no theory by which two crashes involving the earlier — admittedly defective
— switch under somewhat different circumstances could possibly shed light on those questions.
As it was in the Barthelemy/Spain case, “the contrast to Scheuer is telling. There, the plaintiff’s
theory was that the airbags in his car should have deployed when he crashed and that the only
explanation for the fact that they did not deploy was inadvertent switch rotation. The fact that
there were other crashes (under relatively similar circumstances, no less) in which the only
explanation found for airbag non-deployment was the ignition switch defect plainly tended to
support that theory (and tended to support the plaintiff’s experts’ opinions to that effect).” 2016
WL 796846, at *6. Here, by contrast, there is nothing about the Gamage and Pluszek evidence
that tends to make Ward’s theories of defect and causation “more or less probable than it would
be without the evidence.” Fed. R. Evid. 401. Accordingly, the evidence is not admissible to
prove the existence of a defect or causation.
Notice. Finally, Ward seeks to admit evidence (largely in the form of documents
generated by New GM) relating to fifty-six other incidents — fifty-two of which were admitted
in Cockram — solely for purposes of proving notice. (Pl.’s OSI Br. 19-20; see Pl.’s OSI Discl.,
Rows 1-56). Again, New GM opposes the request on the ground that the vehicles in most, if not
all, of those incidents were different from the ignition switch in Ward’s vehicle. (Def.’s Opp’n
Br. 10-15). But that argument lacks the same force under the “relaxed” standard that applies

17

when OSI evidence is offered to prove notice. Scheuer, 2015 WL 9463183, at *2. That is, Ward
alleges in his Amended Complaint that his ignition switch suffered from the same type of defect
as the earlier admittedly defective ignition switch: insufficient torque that enabled the switch to
inadvertently move from the “run” position to the “off” or “accessory” position, thereby
disabling critical safety systems in the vehicle. (14-CV-8317, Docket No. 157, ¶¶ 19, 28).
Moreover, there is evidentiary support for that allegation: Among other things, New GM itself
concluded that the torque in the new switch was below specification, warned drivers of vehicles
with the new switches that large keychains and jarring road conditions could result in inadvertent
rotation, and ultimately extended the initial recall to include vehicles with the new switch (albeit
allegedly because such vehicles might have been serviced or repaired with one of the earlier,
admittedly defective switches). (Pl.’s OSI Br. 1-5). In the final analysis, New GM’s arguments
about the differences between the two switches go to weight rather than admissibility. See, e.g.,
Scheuer, 2015 WL 9463183, at *2; Jodoin v. Toyota Motor Corp., 284 F.3d 272, 280 (1st Cir.
2002) (“When the relevant elements are sufficiently similar, we further emphasize that other
differences are for defendants to highlight and the jury to weigh in its deliberations.”). In fact, to
exclude the other incidents would arguably “prevent the jury from understanding the scale,
significance, and duration of Old GM’s and New GM’s conduct with respect to the ignition
switch defect” allegedly at issue in this case. Cockram, 2016 WL 4410030, at *4.
Remaining Issues. Two collateral issues remain. First, New GM contends that the Court
should limit, if not preclude, Ward’s introduction of OSI evidence involving airbag nondeployment because his expert “concedes that airbag deployment was not warranted in this
accident.” (Def.’s Opp’n Br. 22). As in Barthelemy/Spain, the Court “will not permit Plaintiff to
introduce excessive evidence of airbag non-deployment given [his] concession that [his] crash

18

did not exceed to deployment threshold.” 2016 WL 796846, at *5 n.2. But the Court will not go
further than that and preclude such evidence altogether. Second, New GM objects to testimony
by Ward’s expert, Steve Loudon, about the OSIs on the ground that he conceded that he could
not say whether they were “actually similar” to Ward’s accident and that he did not know
whether inadvertent rotation caused or contributed to causing fifty-six of the OSIs. (Def.’s
Opp’n Br. 23). Any such objection, however, should have been raised by way of a Daubert
motion, not by way of New GM’s opposition to Ward’s offer of OSI proof.
CONCLUSION
For the reasons stated above, Ward’s First Motion in Limine and New GM’s Thirty-Third
Motion in Limine are granted, Ward’s Third and Fourth Motions in Limine are denied, and the
Court reserves judgment on Ward’s Second Motion in Limine pending trial. In addition, Ward’s
application to admit OSI evidence is granted in part and denied in part. Specifically, Ward may
introduce evidence of fifty-six of the sixty OSIs (Pl.’s OSI Discl., Rows 1-56) for purposes of
proving notice only. Ward may not introduce evidence of any OSIs (except the one in his own
vehicle (id. at Row 57), which may or may not constitute OSI evidence in the first instance) to
prove the existence of a defect or causation. Finally, the three unverified consumer complaints
(id., Rows 58-60) will not be admitted for any purpose.
The Clerk of Court is directed to terminate 14-MD-2543, Docket Nos. 3761, 3924, 3927,
3933, 3937, and 3941 and 14-CV-8317, Docket Nos. 220, 223, 229, 233, and 237.

SO ORDERED.
Dated: June 9, 2017
New York, New York

19