Statistics in Class Action Litigation: Admissibility, Expert - - PowerPoint PPT Presentation

Presenting a live 90‐minute webinar with interactive Q&A

Statistics in Class Action Litigation: Admissibility, Expert Witnesses y, p and Impact of Comcast v. Behrend

Leveraging Statistical Evidence and Expert Testimony to Obtain or Defeat Class Certification

T d ’ f l f

1pm Eastern | 12pm Central | 11am Mountain | 10am Pacific WEDNES DAY, JUNE 19, 2013

Today’s faculty features:

Paul G. Karlsgodt, Partner, Baker Hostetler, Denver Rick Preston & Justin Hopson, Hitachi Consulting, Denver Brian A. Troyer, Partner, Thompson Hine, Cleveland Brian A. Troyer, Partner, Thompson Hine, Cleveland

The audio portion of the conference may be accessed via the telephone or by using your computer's

• speakers. Please refer to the instructions emailed to registrants for additional information. If you

have any questions, please contact Customer Service at 1-800-926-7926 ext. 10.

Tips for Optimal Quality

S d Q lit S

• und Quality

If you are listening via your computer speakers, please note that the quality of your sound will vary depending on the speed and quality of your internet connection. If the sound quality is not satisfactory and you are listening via your computer speakers, you may listen via the phone: dial 1-866-869-6667 and enter your PIN when prompted Otherwise please send us a chat or e mail when prompted. Otherwise, please send us a chat or e-mail sound@ straffordpub.com immediately so we can address the problem. If you dialed in and have any difficulties during the call, press *0 for assistance.

Viewing Qualit y

To maximize your screen, press the F11 key on your keyboard. To exit full screen, press the F11 key again press the F11 key again.

Continuing Education Credits

FOR LIVE EVENT ONLY

For CLE purposes, please let us know how many people are listening at your location by completing each of the following steps:

• In the chat box, type (1) your company name and (2) the number of

attendees at your location attendees at your location

• Click the S

END button beside the box If you have purchased S trafford CLE processing services, you must confirm your participation by completing and submitting an Official Record of Attendance (CLE Form). Y

• u may obtain your CLE form by going to the program page and selecting the

appropriate form in the PROGRAM MATERIALS box at the top right corner. If you'd like to purchase CLE credit processing, it is available for a fee. For additional information about CLE credit processing, go to our website or call us at 1-800-926-7926 ext. 35.

Program Materials

If you have not printed the conference materials for this program, please complete the following steps:

• Click on the + sign next to “ Conference Materials” in the middle of the left-

hand column on your screen hand column on your screen.

• Click on the tab labeled “ Handouts” that appears, and there you will see a

PDF of the slides for today's program.

• Double click on the PDF and a separate page will open.

Double click on the PDF and a separate page will open.

• Print the slides by clicking on the printer icon.

Statistics in Class Action Litigation: Statistics in Class Action Litigation: Admissibility, Expert Witness and Impact of Comcast Corp. v. Behrend - Agenda

 Part I – Introduction (~15 min.)  Part II – A legal framework for evaluating statistical evidence in

Behrend Agenda

class certification after Comcast (~40 min.)

 Part III – Practical tips on presenting and challenging statistics

(~20 min.)

 Question and Answer (~15 min.)

5

Part I – Introduction

6

Introduction to Statistics

“Statistics is the science and art of describing data and drawing inferences from them”* Statistics Statistics Inferential Statistics Descriptive Statistics

Describes relationships, correlations, events Makes inferences, generalizations, estimates, predictions

*(Finkelstein and Levin, p. 1)

7

Terminology of “Statistics”?

 Descriptive statistics

 Used to explain an event or course of events.

 Inferential Statistics

 From the data showing Y, you can infer that X is true.

 Probability

H lik l i thi t b t ?

 How likely is something to be true?

 Regression analysis

 Discussed in Wal-Mart Stores, Inc. v. Dukes and Comcast  Examines the relationship between variables.

Examines the relationship between variables.

 Surveys

 Of X population, Y are likely to respond this way.

 Econometrics

 E.g., “but for the misrepresentation, the price would have been X dollars

lower”

 Compilations of Data

 Not “statistics” per se but may raise some of the same issues  Not statistics per se, but may raise some of the same issues.

Rough Justice & Big Data

Over the past decade, as storage and p g computing power have increased exponentially, it has become increasingly tempting to use statistical sampling as a proxy for the actual adjudication of facts in class or mass actions. j

Sources of Data Growth

• Email, collaboration tools, and mobile devices
• Machine and sensor-generated messages

“Big Data: What It Is and Why You Should Care” IDC (June 2011) Hard Disc Storage Price/GB Solid State Disc Storage Price/GB

• Machine and sensor-generated messages
• Digitization of business records and personal

content

• Instrument devices
• Governance, privacy, and regulatory compliance

requirements

9

How Are Statistics Used to Support Class Certification? Class Certification?

 The existence of a common practice

p

 A relationship between the defendant’s conduct

and some injury to class members

 The total damages or other impact caused by a  The total damages or other impact caused by a

practice

 The percentage of people impacted by a

g y practice.

 Given a set of characteristics, the probability that

a person was impacted by a practice a person was impacted by a practice.

 Common reliance

 Truly common reliance, e.g. “fraud on the market”

R li b “ t” f th l

 Reliance by “most” of the class

Why Do Statistics Matter In Class Actions? Class Actions?

 Wal-Mart Stores, Inc. v. Dukes created a more demanding

standard for class certification

 Hannaford showed courts won’t speculate on the ability to

provide necessary data post-certification

 Comcast sets higher level of scrutiny (logical fallacies, causal

li k b t i j & d ) link between injury & damage)

 The lower courts are starting to fill in the gaps left by the Dukes

Court’s analysis—see, for example, Duran v. U.S. Bank National Association Association

 Both sides are likely to attempt to create a more well-developed

factual record of the people, process, and technology

 Statistics often provide an appealing way to illustrate how  Statistics often provide an appealing way to illustrate how

aggregate or common method of injury is possible, and can create diverse individual outcomes

 Data is more available and accessible than ever before

11

Part II – Case Law on the Use of Statistics in Class Certification

12

Dukes in Review Dukes in Review

 Did not change the landscape regarding statistics and class

certification but confirmed the necessity of rigorous scrutiny. y g y

 The Court examined the statistical analyses and found inferential

gaps between the policy that statistics were claimed to show and what they actually showed.

 Illustrated and confirmed inherent limitations of statistical and  Illustrated and confirmed inherent limitations of statistical and

aggregate proof.

 Confirmed that, validity of statistics aside, conceptual gaps are

critical.

 Even if statistics showed the claimed pattern, that pattern would

not establish commonality.

 Whether any individual decision was discriminatory would still

require individual proof require individual proof.

 Condemned use of “trial by formula.”  Gave a strong hint in favor of Daubert being required at class

certification, but did not answer the question directly. , q y

13

Comcast Corp. v. Behrend, 133 S Ct 24 (2012) 133 S. Ct. 24 (2012).

 Daubert question left unanswered again.  Issue decided was not the issue initially certified for review.

ssue dec ded as

• t t e ssue

t a y ce t ed o e e

 Question presented was whether court could grant class certification

without deciding whether expert testimony was admissible.

 Question decided was whether “certification was improper because

plaintiffs failed to establish that damages could be measured on a classwide basis” through expert testimony they presented.

 Basic Facts

 Plaintiffs alleged that Comcast was engaged in “Clustering,” or the

concentration of operations in a geographic region to increase market h share.

 Four theories of harm – clustering:



Made it profitable to withhold local sports programing from competitors



Reduced competition from “overbuilders”  The only common impact accepted by the district court district court.



Reduced the level of “benchmark” competition



Increased Comcast’s bargaining power  Expert performed a regression analysis that purported to show the overall

price impact of clustering. p p g

14

Framework for Evaluating Statistical E id Off d f Cl C tifi ti Evidence Offered for Class Certification

1.

What is the theory of liability (wrongdoing/conduct)?

2

What is the theory of common impact?

2.

What is the theory of common impact?

3.

What elements/questions are claimed to be matters of common statistical proof?

a.

Liability

b.

Reliance, causation, injury (impact questions)

c.

Damages

4.

By what methodology?

5

I th th d l h i l ti i d bl f idi

5.

Is the methodology, when rigorously scrutinized, capable of providing such common proof? E.g.,

a.

Faulty assumptions

b.

Analytical errors

c.

Ignored evidence

d.

Does it properly tie the theories of liability, impact, and damages together?

1)

Alleged Conduct → Common Impact (reliance/causation/injury)

2)

Common Impact → Damages

)

p g 15

Unresolved Questions? Unresolved Questions?

 Comcast makes clear that there must a common impact and the

evidence (expert or otherwise) must provide a logical connection evidence (expert or otherwise) must provide a logical connection between the alleged wrongdoing, a common impact, and damages.

 Must expert opinions offered in support of class certification

satisfy Daubert standards?

 How will the requirement that damages be provable with

common evidence on a classwide basis be interpreted by courts and applied in the antitrust and other contexts?

16

Framework Applied to Comcast Framework Applied to Comcast

• What is the theory of liability (wrongdoing/conduct)?
• Clustering
• Clustering
• What is the theory of common impact?
• Four mechanisms proposed, but only deterrence of “overbuilding” accepted

as subject to common proof. j p

• What elements/questions are claimed to be matters of common

statistical proof?

• Damages
• By what methodology?
• Econometric regression analysis
• Is the methodology, when rigorously scrutinized, capable of providing

such common proof?

• No, it assumed the validity of all four proposed theories of common antitrust

impact rather than being limited to deterrence of overbuilders.

17

Framework Applied to Dukes Framework Applied to Dukes

• What is the theory of liability (wrongdoing/conduct)?
• Managers exercise policy of discretion to discriminate against women
• Managers exercise policy of discretion to discriminate against women.
• What is the theory of common impact?
• Pattern or practice of discrimination (general policy of discrimination)
• What elements/questions are claimed to be matters of common
• What elements/questions are claimed to be matters of common

statistical proof?

• Pattern or practice of discrimination (conduct, injury) and damages
• By what methodology?

By what methodology?

• Pattern or practice of discrimination
• Sociologist’s “social framework” opinion
• Statistician’s comparison by region of numbers of women promoted versus in hourly pool
• Labor economist’s comparison of rates of promotion with those of competitors
• Labor economist s comparison of rates of promotion with those of competitors
• Damages
• Trial by formula (sampling) scheme based on depositions of selected class members and

extrapolation and allocation to all class members

18

Framework Applied to Dukes Framework Applied to Dukes

 Is the methodology, when rigorously scrutinized, capable of

providing such common proof? providing such common proof?

 Pattern or practice of discrimination: No, methodologies flawed with

conceptual gaps and failures of inference

 “Social framework” opinion lacked any quantitative substance.

Social framework opinion lacked any quantitative substance.

 Regional disparities and comparisons to competitors could not establish a

uniform practice affecting each store.

 Even if they could, they could not show that each female employee was

bj t t di i i ti hi h f th subject to discrimination, or which of them were.

 Damages: No, trial by formula rejected as matter of law

 Violates Rules Enabling Act and deprives the defendant of due process.  Winners and losers cannot be averaged across the class  Winners and losers cannot be averaged across the class.  Fundamentally is not even an attempt to prove damages with common

evidence but a shortcut around individual issues.

19

Issue Continuity and Primacy of Injury in Dukes Injury in Dukes

 Commonality requires same injury.

 To prove commonality plaintiffs must show that they can  To prove commonality, plaintiffs must show that they can

prove by common evidence that each class member suffered the same injury (i.e., based on common facts).

 Why no common question of pattern or practice of

Why no common question of pattern or practice of discrimination?

 Alleged discrimination that does not affect 100% is not

“common mode” of exercising discretion, thus not a li f di i i ti ( l t l i t t ti )? policy of discrimination (employment law interpretation)?

 For women not subject to discrimination, there is no

question of a pattern or practice affecting them (Rule 23 interpretation)? interpretation)?

 Certification would erroneously assume all were injured?  No question of a policy of discrimination that necessarily must

be answered for every class member?

20

McLaughlin

 Plaintiffs alleged implicit representation that light cigarettes are

healthier; sought $800 billion.

 Plaintiffs relied upon sixteen experts, including economists who

d t ti ti l d t i l proposed statistical and econometric analyses.

 Judge Weinstein certified nationwide class of light cigarette

consumers under RICO, applying “price impact” theory of reliance similar to the theory of fraud on the securities market. y

 Reversed by McLaughlin v. American Co., 522 F.3d 215 (2d

• Cir. 2008).

 Individual proof was required: reliance, loss causation, injury, damages

(and limitations) (and limitations).

 Market for light cigarettes is not efficient.  Individual facts presented to show non-reliance by customers.  Expert’s survey evidence “pure speculation.”  Statistical analysis did not prove the relevant facts.  Rejected “fluid recovery” approach of awarding aggregate “class”

damages followed by “simplified proof of claim procedure” and cy pres.

21

Framework Applied to McLaughlin Framework Applied to McLaughlin

• What is the theory of liability (wrongdoing/conduct)?
• Advertising falsely promoting health benefits of light cigarettes
• What is the theory of common impact?
• Advertising “distorted the body of public information,” raising demand.
• What elements/questions are claimed to be matters of common statistical

proof?

• Reliance, causation, injury, and possibly damages
• By what methodology?
• Reliance: consumer surveys about preferences
• Causation and injury: econometric regressions and surveys purporting to show

j y g y p p g “price impact” and "loss of value" in the aggregate (fraud on the market)

• Is the methodology, when rigorously scrutinized, capable of providing such

common proof? No:

• Reliance: fraud on the market inapplicable; ignored contrary evidence regarding

pricing sales consumer and plaintiff behavior pricing, sales, consumer, and plaintiff behavior

• Loss causation: fraud on the market does not apply; circular; contrary to evidence

regarding market behavior

• Injury: loss of value and price inflation theories legally invalid, speculative,

unsupported by, and contrary to, evidence

22

In re Hannaford In re Hannaford

 In re Hannaford Brothers Company Data Security Breach

Litigation, No. 2:08-md-01954-DBH (D. Me., March 20, 2013). g , ( , , )

 Payment card system compromised by hackers.  Court initially denied motion to dismiss for lack of standing on the

ground that expenditures made to protect against future identity theft constituted injury in fact (But see the Supreme Court’s theft constituted injury in fact. (But see the Supreme Court s more recent decision in Clapper).

 At class certification, plaintiffs argued that they would be able to

prove common impact through statistical evidence about the likelihood that the breach would have caused individual damages.

 Court denied certification based in part on the fact that no expert

report supporting the common impact theory had been report supporting the common impact theory had been presented.

 Side note – on motion for reconsideration, plaintiffs have argued

that common impact can be inferred from “reason and common sense” and that no expert testimony is necessary sense and that no expert testimony is necessary.

23

Framework as Applied to In re Hannaford Hannaford

• What is the theory of liability (wrongdoing/conduct)?
• Negligence resulted in system vulnerability and loss of card information.
• What is the theory of common impact?
• Allegation is that negligence led to compromise of card information for entire
• class. Specific financial impacts of compromise, if any, would vary.

Wh l / i l i d b f i i l

• What elements/questions are claimed to be matters of common statistical

proof?

• Damages.

By what methodology?

Questions – Is “fluid recovery” = “trial by formula” (Dukes )? Is it a valid way to calculate damages on a class-wide basis? (Comcast)

• By what methodology?
• Plaintiffs argued that they could compute overall damages on a statistical basis

(fluid recovery). Problem is in determining which customers.

• Is the methodology, when rigorously scrutinized, capable of providing such

a ages o a c ass w e bas s? (Co cast)

Is the methodology, when rigorously scrutinized, capable of providing such common proof?

• Methodology was only hypothesized, so court did not consider this question.

24

Duran Duran

 Duran v. U.S. Bank National Association, No. A125557 &

A126827 (Cal. App., Feb. 6, 2012), review granted, No. A126827 (Cal. App., Feb. 6, 2012), review granted, No. S200923 (Cal., March 19, 2012).

 Same expert (Dr. Drogin) as in Dukes.  Trial court used Drogin’s analysis as a model but came up with

g y p its own simplified analysis.

 Trial court applied “statistical” analysis to estimate the number of

employees within the class that had been misclassified for ti

• vertime pay purposes.

 Court of Appeal held:

 Methodology violated due process because it denied defendant

• pportunity to provide relevant evidence and individualized defenses
• pportunity to provide relevant evidence and individualized defenses

relating to classification of each employee.

 Methodology was flawed because sample was arbitrary.  Sampling would have been improper even if used to calculate

d d t th hi h i f damages due to the high margin of error.

25

Framework as Applied to Duran Framework as Applied to Duran

• What is the theory of liability (wrongdoing/conduct)?
• Misclassification of employees.

p y

• What is the theory of common impact?
• Employees have a statistical likelihood of having been misclassified. Impact is

better characterized as “aggregate” rather than common.

• What elements/questions are claimed to be matters of common statistical

proof?

• Fact of misclassification (impact).
• By what methodology?
• Regression analysis proposed, but court applied its own formula.
• Is the methodology, when rigorously scrutinized, capable of providing such

f? common proof?

• No. Problems with qualification, methodology and reliability of underlying data.

Judge not qualified, sample arbitrary, and error rate too high.

• Furthermore, allowing a general, statistical approach for determining a likelihood

26

Furthermore, allowing a general, statistical approach for determining a likelihood that a given employee might have been misclassified violated due process.

F b k Facebook

 In re Facebook, Inc. PPC Advertising Litigation, No. C

09-3043 PJH slip op (N D Cal Apr 13 2012) FRCP 09 3043 PJH, slip op. (N.D. Cal. Apr. 13, 2012), FRCP 23(f) review granted (9th Cir. July 18, 2012).

 Allegation that Facebook breached “cost-per-click”

agreements with advertisers by charging for “invalid” agreements with advertisers by charging for invalid clicks.

 Plaintiffs proposed that their experts could create a

methodology that would distinguish between valid and methodology that would distinguish between valid and invalid clicks.

 Court rejected this argument, finding that “there is no

way to conduct this type of highly specialized and y yp g y p individualized analysis for each of the thousands of advertisers in the proposed class.”

27

Framework as Applied to Facebook Framework as Applied to Facebook

• What is the theory of liability (wrongdoing/conduct)?
• Illegally charging for invalid clicks.

g y g g

• What is the theory of common impact?
• Invalid clicks cause charges that are not legally justifiable. This may be better

characterized as a theory that would permit formulaic resolution of individual impacts.

• What elements/questions are claimed to be matters of common statistical

proof?

Indi id al impacts and damages

• Individual impacts and damages.
• By what methodology?
• Algorithm claimed to be able to distinguish valid from invalid clicks.

Is the methodology when rigorously scrutinized capable of providing such

• Is the methodology, when rigorously scrutinized, capable of providing such

common proof?

• No. Too many variables involved in distinguishing valid from invalid clicks.

28

In re Neurontin Sales and Mktg. P ti Litig Practices Litig.

 Causation problem: which off-label prescriptions were caused by

allegedly fraudulent promotion? g y p

 Plaintiffs relied upon econometric analysis to try to show causation of

“all” off-label prescriptions.

 In first opinion, 244 F.R.D. 89 (D. Mass. 2007), Judge Saris gave

plaintiffs opportunity to show through “statistical proof” that essentially all plaintiffs opportunity to show through statistical proof that essentially all prescriptions in each category were caused by fraud.

 Second class certification motion also denied, 257 F.R.D. 315 (D. Mass.

2009): N t ffi i t k t

 Not an efficient market.  Defendant’s right to present evidence defeats predominance.  Closer scrutiny of expert opinions for class certification was mandated

than presumed in earlier opinion. Wh t’ i i th t l th b t ti ll ll (>99%) f

 Where expert’s opinion was that less than substantially all (>99%) of

prescriptions were caused by fraud, individual inquiry required.

 Where expert’s opinion was that substantially all prescriptions were

caused by fraud, the expert analysis was flawed.

29

Framework Applied to In re Neurontin Neurontin

• What is the theory of liability (wrongdoing/conduct)?
• Off-label marketing
• Off-label marketing
• What is the theory of common impact?
• Off-label marketing inflated sales by misleading prescribers, payors, consumers

about benefits (claimed not to be a theory of fraud on the market) ( y )

• What elements/questions are claimed to be matters of common statistical

proof?

• Reliance, causation, injury
• By what methodology?
• Statistical analysis purporting to show percentages of prescriptions for each off-

label indication allegedly caused by off-label marketing

30

Framework Applied to In re Neurontin Neurontin

• Is the methodology, when rigorously scrutinized, capable of

providing such common proof? No. providing such common proof? No.

• Consumers
• Indications < 99% caused by off-label promotion → which
• nes?
• Indications > 99% caused by off-label promotion →

methodology flawed by false assumption that all detailing was

• ff-label and fraudulent, and failure to account for other factors

affecting off-label use affecting off-label use.

• Third-Party Payors
• Methodology could only provide aggregate percentages across

all TPPS. all TPPS.

• Could not account for differences in knowledge, preferences,

etc.

• Individual TPP inquiries still required.

31

Whirlpool/Kenmore Cases Whirlpool/Kenmore Cases

 Alleged design defect causing numerous models of

front-loading washers to develop mold odor front loading washers to develop mold odor.

 Injury/damage issues: most users had experienced

no odor.

 Plaintiffs argued that 35% complained of odor  Plaintiffs argued that 35% complained of odor.  Defendants argued that much smaller percentages

experienced odor. The Sixth Circuit affirmed certification Glazer v

 The Sixth Circuit affirmed certification, Glazer v.

Whirlpool Corp., No. 10-4188 (May 3, 2012), and the Seventh Circuit (Posner, J.) reversed denial of certification Butler v Sear Roebuck and Co Nos certification, Butler v. Sear, Roebuck and Co., Nos. 11-8029, 12-8030 (Sept. 28, 2012).

 Both decisions vacated and remanded by the

Supreme Court in light of Comcast. Supreme Court in light of Comcast.

32

Framework Applied to Whirlpool/Kenmore Cases Whirlpool/Kenmore Cases

• What is the theory of liability (wrongdoing/conduct)?
• Design defect causing mold odor

Design defect causing mold odor

• What is the theory of common impact?
• All machines share design defect making them prone to mold odor.
• What elements/questions are claimed to be matters of common

statistical proof?

• Proximate causation, injury
• By what methodology?
• Percentage of consumers who allegedly complain of mold odor
• Percentage of consumers who allegedly complain of mold odor
• Is the methodology, when rigorously scrutinized, capable of providing

such common proof?

• Both Sixth and Seventh Circuits held that certification was proper even if

some class members had not suffered injury.

• Recognition it is not a common question?
• Both vacated by Supreme Court.

33

In re Zyprexa

 Judge Weinstein’s certification of off-label TPP economic

loss class under RICO reversed by the Second Circuit.

 UFCW Local 1776 & Participating Health & Welfare Fund v. Eli

UFCW Local 1776 & Participating Health & Welfare Fund v. Eli Lilly & Co., 620 F.3d 121 (2d Cir. 2010).

 “Excess price” analysis could not provide common proof of

 but-for (transactional) causation, because drug pricing is inelastic.

but o (t a sact o a ) causat o , because d ug p c g s e ast c

 proximate (direct) causation, because alleged chain of causation

was incomplete.

 “Excess sales” theory could not provide common proof of

y p p causation because, e.g.,

 it assumed away all other factors affecting prescriptions.  there was individualized evidence of non-reliance.  it ignored alternative prescriptions and costs, some of which

could even have cost more.

34

Framework Applied to Zyprexa Framework Applied to Zyprexa

• What is the theory of liability (wrongdoing/conduct)?
• Off-label marketing
• Off-label marketing
• What is the theory of common impact?
• Off-label marketing inflated price and sales of Zyprexa (fraud on the market).
• What elements/questions are claimed to be matters of common
• What elements/questions are claimed to be matters of common

statistical proof?

• But-for and proximate causation, injury
• By what methodology?

By what methodology?

• Econometric analyses purporting to show aggregate “excess price” and

“excess sales”

35

Framework Applied to Zyprexa Framework Applied to Zyprexa

• Is the methodology, when rigorously scrutinized, capable
• f providing such common proof? No:
• f providing such common proof? No:
• “Excess price” analysis could not provide common proof of
• but-for (transactional) causation, because drug pricing is inelastic

(physicians do not consider price).

 proximate (direct) causation, because

proximate (direct) causation, because

 alleged reliance by physicians is independent of price negotiation and

payment by TPPs (advised by PBMs)

 variations in price negotiation by TPPs showed that the alleged chain of

causation was incomplete.

“E l ” th ld t id f f

 “Excess sales” theory could not provide common proof of

causation because, e.g., it assumed away all other information

and factors affecting prescriptions.

 TPPs continued to pay for Zyprexa.

TPP b bl id f diff t t f ff l b l i ti

 TPPs probably paid for different percentages of off-label prescriptions.  some prescribing doctors not misled.  it ignored alternative prescriptions and costs, some of which could even have

cost more.

36

Rhodes

 Rhodes v. E.I. Du Pont de Nemours and Co., 253 F.R.D. 365

(S.D.W.V. 2008)

 Medical monitoring claim based on contamination of drinking  Medical monitoring claim based on contamination of drinking

water with C-8.

 Problems with toxicologist’s and epidemiologist’s quantitative

• pinions offered to establish common proof:
• pinions offered to establish common proof:

 Did not address the question of the relationship between exposure

and a significantly increased risk of health problems; and f

 Did not provide any common proof that any given individual

suffered a significantly increased risk of the exposure.

 Preliminary and insufficient data was used.  Failed to rule out other variables.  Proposed remedy was a precautionary public health measure, not

something that can be awarded as a tort remedy.

37

Framework Applied to Rhodes Framework Applied to Rhodes

• What is the theory of liability (wrongdoing/conduct)?
• Contamination of drinking water

g

• What is the theory of common impact?
• All area residents were exposed to C-8.
• What elements/questions are claimed to be matters of common statistical

proof? proof?

• Significant exposure, significantly increased risk of disease, need for medical

monitoring

• By what methodology?
• Toxicologist’s risk assessment and physician/epidemiologist’s epidemiological

survey

• Is the methodology, when rigorously scrutinized, capable of providing such

common proof? No

• Background risks unknown and overlooked
• Improper focus on “safe level” and not significantly increased risk
• Preliminary and insufficient data
• Aggregate estimates of exposure and risk could not be applied to individuals

gg g p pp

38

Part III – Practical tips on p presenting and challenging statistics statistics

39

Tips for Dealing With Experts

Draw Inferences (optional) Analyze Collect Data Considerations



How collected? Trusted source?



Is the method / measurement process reliable (consistent performance with repetition)? V lid?



Has the methodology been peer reviewed? Discredited?



Can the results be generalized?



How are charts/graphs d?



What variables were omitted?



Did the expert answer the right question?



How do I estimate whatever is i i ?



Valid?



Recorded properly?



Categories appropriate?



What is the non-response rate ( )? Wh ? presented?



What method is used to select the units (or scale)?



Do analyses reach different

• pinions?

missing?



Is the sample size big enough to be predictive?



How accurate are the predictions? (survey)? Why?

• pinions?

predictions?



Does the analysis prove a fact to be true, or does it assume the fact is true?

40

Ask, “What is missing? Who would know?”

40

Common Statistical Flaws

Ill lit



Illusory commonality



When (even reliable) statistics only purport to answer a question for X or X% of a class, or show that X or X% of a proposed class is affected, commonality does not exist (indeed, is disproved).



Discrimination (Dukes)



Consumer fraud (Zyprexa, Neurontin)



Breach of contract (e.g., timeliness of payment)



Overlooked factors and intervening causes.

Alt ti d i ht b i f



Alternative drugs might be more expensive for some.



Some people smoke lights for flavor or because they are “cool.” 

Circularity/Assumed Reliance

Wh t i l i t dl h th t ti b d l



When an econometric analysis purportedly shows that causation can be proved on a class- wide basis through a “price effect,” the analysis may assume reliance or causation rather than prove them. 

Erroneous assumptions Erroneous assumptions



All off-label marketing is fraudulent (legal/factual error).



Third-party payors have similar rates of reimbursement for off-label prescriptions (factual error).



All class members were unaware the drug was unapproved (factual error). 41

Common Impact & Wrongdoing Common Impact & Wrongdoing

 Defendant Goal: Identify a disconnect between the evidence required

and the statistician’s expert opinion and the statistician s expert opinion

 May not require a strong statistical background to evaluate

 Related Cases  Related Cases

 Dukes: Conclusions don’t go further than showing that disparities exist  Rhodes: Did not address the question of the relationship between exposure

and a significantly increased risk of health problems

 Fact Patterns & Tools for Challenging Experts

 Relevance – The analysis addressed the wrong issue  Underlying data – The source(s) aren’t known, authoritative, complete, or

aren’t being interpreted correctly.

 Assumption vs. conclusion – The analysis assumes the fact it’s proving  Other logical fallacies

Circular logic key variables ignored

 Other logical fallacies – Circular logic, key variables ignored 42

Challenging Common Impact Challenging Common Impact

 Defendant Goal: Identify any potential for individualized outcomes  Related Cases

 Dukes: fewer promotions doesn’t mean that all women suffered discrimination  Comcast: analysis took into account the combined value of four different  Comcast: analysis took into account the combined value of four different

potential impacts, only one of which the trial court had found could be a common impact.

 McLaughlin: individual facts (descriptive statistics) presented to show non-

li b t reliance by customers

 Rhodes: did not provide any common proof that any given individual suffered

a significantly increased risk of the exposure

 Recurring Theme – Inability to specify root causes and predictors that

are common to all class members

43

Challenging Common Impact - Visualizing Root Causes Visualizing Root Causes



Is there a common “answer” for all class members—i.e. did the same set of circumstances apply to each class member; “Yes” in Halliburton; “No” in Dukes I h h h l i ( h h d )?



Is there perhaps some other explanation (other than gender)?



Root Cause (Ishikawa) Diagram

Personal What Else? Class Definition “’[A]ll women [w]ho have been or may be subjected to Wal Mart’s challenged Traits

Gender Personal decisions Age Family Situation

a se Region

Dept Store

to Wal-Mart s challenged pay and management track promotions policies and practices.”

Full-/part-time Tenure Role Previous job Discretion Mobility Workload Expectations

Employment Status Management Behaviors Policies & Procedures

j Performance

Paraphrasing: While disparity may exist, the underlying root causes are likely to be different among class members 44 members 44

Challenging Common Impact – H N t & V i bl C l it Human Nature & Variable Complexity

Outset of Analysis

Tension & Tension & Vested Interest Few Variables Many Variables

Consideration of many variables can lead to:

• Class re-definition
• Sub-classing
• Removal of damages

categories Consideration of just a few variables can lead to:

• Agreement on priorities, focus
• Expedited timeframes

“Return to Sanity”

categories

• Class de-certification

Objections: “Let’s keep it simple” “It’s too complicated” “It’s not manageable” Objections: “Yes, but we’re not considering . . .” “We seem to be in denial of how many moving pieces there are . . . “ “This is too simplistic”

“Point of Litigation/Judicial Discomfort”

45

g p

Challenging Common Impact – F t P tt Fact Patterns

 Fact Patterns and Tools for Challenging Experts

Statistics used to estimate percentage of class members to whom the

 Statistics used to estimate percentage of class members to whom the

defendant may be liable (Trial by formula)

 This violates due process according to Duran.

 Statistics used to aggregate and apportion damages (Trial by

formula)

 No, according to dictum in Dukes, but some courts may be more

welcoming of this argument.

 Winners and losers

 Some class members are actually better off as a result of the alleged

practice

 Subclasses may cure this problem, but problem might be in identifying who

goes in which category g g y

46

Challenging Common Impact – F t P tt ( t ) Fact Patterns (cont.)

 Single policy or practice (Dukes)

 Does a single policy exist? (McReynolds)  Does a single policy exist? (McReynolds)

 Is there a way to prove a causal link between the policy and some

alleged harm?

 Can the causal link be resolved by reference to common, classwide

evidence.

 Mass reliance/common impact—ask whether

 Legal theory is such that individual reliance is not required (if so, still have

to consider the separate question of causation) to consider the separate question of causation)

 Reliance question can be both proved and resolved by reference to

common evidence.

 If a theory of damages is provided, it should align with the theory of

common impact

47

Challenging Common Impact – E l ti g M th d l gi Evaluating Methodologies



Defendant Goal: Evaluate the applicability of the statistician’s methodology



Key message: Take a hard look at the statisticians methodology. It can have a big impact on y g gy g p the case outcome



Example Cases



Dukes: Trial by formula not allowed.



McLaughlin: The expert’s survey methodology deemed “pure speculation” g p y gy p p



Neurontin: Inability to ID root cause 1) Where expert’s opinion was that less than substantially all (>99%)

• f prescriptions were caused by fraud, individual inquiry required; 2) Where expert’s opinion was that

substantially all prescriptions were caused by fraud, the expert analysis was flawed



Zyprexa: Root cause not pinpointed by expert Rh d 1) P li i d i ffi i t d t d 2) F il d t l t th i bl



Rhodes: 1) Preliminary and insufficient data was used. 2) Failed to rule out other variables.



Duran: 1) Methodology was flawed because sample was arbitrary. 2) Sampling would have been improper even if used to calculate damages due to the high margin of error



Facebook: there was no way to conduct this type of highly specialized and individualized analysis for each of the thousands of advertisers in the proposed class 

Fact Patterns & Tools for Challenging Experts



A cohesive story often requires collaboration between the attorney to identify the legal risks, the business owner to identify the process, system, & resource risks, and finally the statistician to help quantify the extent to which the data supports the theory

48

Confidence in Confidence

Question for the C t At h t Courts: At what point do we get to an acceptable level

• f common proof?

p

49

Challenging Common Impact – V lidit f D t Validity of Data

 Defendant Goal: Evaluate the usefulness and validity of data  Fact Patterns & Tools for Challenging Experts



Is there an actively-engaged, senior executive-sponsored data governance body in place? p



What evidence is available to demonstrate that a robust data quality management program is used?



Are data stewardship roles & responsibilities well-defined?



Is a standards-based data management process and procedure framework in place? g p p p



Does the company have day-to-day reliance on purpose-built data governance tools and performance metrics?



Does a robust and active master data management program exist?



Is a financial information management program in place? Is a financial information management program in place?



How well and how often are errors identified, analyzed for root cause, and corrected?



How well-managed is the quality of data coming into the system, both manually and in an automated fashion?



What data in the system can’t be considered the source of truth?



What data in the system can t be considered the source of truth? 50

Challenging Common Impact – E t C d ti l Expert Credentials

 Defendant Goal: Evaluate the appropriateness of expert credentials  Fact Patterns & Tools for Challenging Experts



Does the statistics expert have extensive knowledge of the subject area s/he is analyzing?



Or is the expert simply a statistician with no particular understanding of the subject at issue? If so, is there a relevant supporting expert?



In what ways should the statistician’s testimony be supplemented by other experts with subject area or data quality knowledge?



What are the strengths and weaknesses of the foundational philosophies and historical tendencies of this expert’s approach? this expert s approach?



Consider an expert’s subfield emphasis—e.g. econometrics, biostats, product testing



Consider the expert’s attention to the entire lifecycle of a statistic—e.g. initial data profiling, study design, collection method



What evidence of error exists in the expert’s published materials? Commentary by other statisticians?



What evidence of error consideration is given in the expert’s own published materials? Are none, some,

• r all potential deficiencies noted by the expert?



How extensive and meaningful were peer reviews for published materials, assuming they exist at all?

51

For Further Study



David H. Kaye & David A. Freedman, Reference Guide on Statistics, Reference Manual on Scientific Evidence 2d Ed. (Federal Judicial Center 1981) (http://www.fjc.gov/public/pdf.nsf/lookup/sciman02.pdf/$file/sciman02.pdf)



Robert Ambrogi, Statistics Surge as Evidence in Trials, IMS Newsletter, BullsEye: August 2009 (http://www ims-expertservices com/newsletters/aug/statistics- August 2009, (http://www.ims expertservices.com/newsletters/aug/statistics surge-as-evidence-in-trials-081409.asp)



Edward K. Cheng, A Practical Solution to the Reference Class Problem, 109

• Colum. L. Rev. 2081 (2009)

(http://www.columbialawreview.org/assets/pdfs/109/8/Cheng.pdf)



Denise Martin, Stephanie Plancich, and Mary Elizabeth Stern, Class Certification in Wage and Hour Litigation: What Can We Learn from Statistics? (Nera Economic Consulting 2009) (http://www.nera.com/extImage/PUB_Wage_Hour_Litigation_1109_final.pdf)



Dukes plaintiff’s Expert Dr Richard Drogin’s Statistical Report



Dukes, plaintiff s Expert Dr. Richard Drogin s Statistical Report (http://www.walmartclass.com/all_reports.html)



Michael O. Finkelstein and Bruce Levin, Statistics for Lawyers: Second Edition (Springer, 2001)



Finkelstein, Michael O., Basic Concepts of Probability and Statistics in the Law Finkelstein, Michael O., Basic Concepts of Probability and Statistics in the Law (Springer, 2009)



Olive Jean Dunn and Virginia A. Clark, Applied Statistics: Analysis of Variance and Regression, Second Edition (John Wiley & Sons, 1987)

52

Thank You Thank You

 Topics covered

I i i t f t ti ti & th f d t

 Increasing importance of statistics & growth of data  Basic statistical concepts and use in litigation  Case studies

Case studies

 Practical tips

 Questions?

53

Today’s Panel Today’s Panel

Paul G. Karlsgodt Brian A. Troyer Baker Hostetler Denver, CO pkarlsgodt@bakerlaw.com Thompson Hine Cleveland, OH Brian.Troyer@thompsonhine.com p g @ y @ p J i H Ri k P Justin Hopson Hitachi Consulting Denver, CO Rick Preston Hitachi Consulting Denver, CO JHopson@hitachiconsulting.com Rick.Preston@hitachiconsulting.com

54