Mass Torts
J. G. Fleming'*'

Mass accidents have become a familiar incident of the modern way of life, a
by-product of advancing technology in the production, distribution and use of
toxic agents, dangerous pharmaceuticals, fast modern transport and other
hazardous activities. A single type of product like asbestos or thalidomide, released
on a mass market by one or numerous manufacturers, may inflict injury or disease
on a vast multitude of consumers or their offspring. Or a single accident, like an
aeroplane collision, explosion or escape of poison gas (Bhopal), may bring injury or
death to thousands and dislocation to a whole region. The first is sometimes called
a mass products case, the second a mass accident. Both entail injury to multiple
victims and present adjudicatory problems very different from those faced in
routine accidents. The traditional method of case-by-case adjudication and the
applicable principles of substantive law, still largely based on an individualistic
philosophy of "corrective justice" between man and man, is rather unequal to this
challenge.

TWO ENGLISH CASES
Two mass product disasters have been conspicuous in English experience:

thalidomide in the 1960's, Opren twenty-five years later.

Thalidomide (Contergan)\
Between 1958 and 1961, the Distillers Company manufactured and marketed

under licence in the United Kingdom a drug developed by a small, upstart and
aggressive German pharmaceutical firm, Chemie-Griinenthal. The drug was
recommended as a non-barbiturate, atoxic sedative and, among others, was taken

*Professor of Law, University of California at Berkeley School of Law and Arthur Goodhart
Professor in Legal Science, University of Cambridge. The Child & Co. Oxford Lecture 1988, printed
by kind permission of Professor Fleming and Child & Co..
1. The best account, with special focus on the English scene, is by the Summy Times Insight Team,
Suffer Ihe Children: The Slory of Thalidomide (1979). Sjostrom and Nielsson, Thalidomide and Ihe Power of
Ihe Drug Companies (1972) deals also with litigation in other countries, including a summary of the
German court's justification for terminating the criminal proceedings against officials of
Chemie-Griinenthal. The latter is fully reported in [1971] J.z. 507.

37



THE DENNING LAWJOURNAL

by women to counteract the strains of early pregnancy. By 1961 a number of
newborn infants were born with no, or deformed, limbs, a condition which was
linked to the drug about the same time by two gynaecologists, one in Australia, the
other in Germany. Eventually some 450 victims emerged in Britain, altogether
some 8,000 in thirty different countries. In England between 1962 and 1966
proceedings were commenced by the parents of 70 children, 65 of which were
settled in 1968 on the basis that they receive 40% of what would have been
recoverable if judgment had gone against the defendants.2 The size of the
reduction, far beyond a normal discount or settlement, was due, besides disputed
fault in testing, to doubts whether the common law recognised a cause of action for
pre-natal injury - doubts which, at a later stage of negotiations, were somewhat
lessened by a favourable Australian decision3 and, later yet, prospectively removed
by the Congenital Disabilities (Civil Liability) Act 1976, one of the few lasting
legacies of the thalidomide affair.

By 1969 Distillers had paid out some £1 million to 58 of the claimants. This left
389 other claims which had been started later but eventually qualified under the
Limitation Act. By 1971 Distillers, as an act of grace, offered to set up a trust fund
of £3% million (originally £2.5 million), spread over ten years. These terms were
scathingly castigated by the Sunday Times in an attempt to arouse public
indignation, but Distillers immediately sought an injunction against any discussion
pending acceptance of the settlement terms on the ground that it constituted an
attempt to interfere with the course of justice.4 This legal manoeuvre, supported
by the Attorneys-General of two successive governments, succeeded in stifling to
the very end all factual information regarding the (inadequate) testing procedure
by Distillers and their German licensors. It is an episode, far better remembered
among lawyers, and certainly journalists, than any other aspect of the thalidomide
litigation, because it involved an even more divisiveissue - freedom of the press -
and an eventual censure of the House of Lords by the European Court of Human
Rights.s

At first, even discussion in Parliament was resisted by invoking the rule against
debate on matters sub judice. The matter finally came up on a motion calling on
Distillers to face up to their moral responsibility and for immediate legislation to
establish a trust fund for the children. As a result of continuous agitation in the
Press and pressure on and by the Government the settlement offer was in the end,
i.e. after more than 10 years in the courts, increased to £20 million, estimated to be
well above the full tort measure of damages.

2. S. v. Dis/if/ers Co [1970J 1 W.L.R. 114 (two representative actions to set a standard for assessing
damages under the settlement).
3. Wall v. Rama [19721 V.R. 353.
4. AII.-Gm. v. Times Newspapers [1973] QB. 710; [1974] A.C. 273. The injunction was eventually

discharged in 1976, shortly before the last four cases were settled.
5. The Sundll;Y Times v. UK [1979] 2 E.H.R.R. 245.

38



MASS TORTS

Lessons
At least two major lessons emerge from this protracted saga. First, that the
ultimate outcome was achieved not through the legal process but through
extra-legal means. Second, that the outcome was not bottomed on principles of
legal liability but on an overriding sense of justice.

The dismal failure of the legal process revealed several fundamental flaws in the
ability of the English system to cope with effective personal injury litigation,
especially of mass claims. It is a frequently voiced boast that the division of the
English legal profession into solicitors and barristers tends to assure the most
skilful representation of the client. Alas, this viewpoint focuses primarily on the
choice of barrister and trivialises the role of the solicitor. Few solicitor firms,
outside trade union solicitors, specialise in personal injury claims and acquire
anything like the expertise and stamina necessary for energetic litigation. Also,
litigation being less remunerative than conveyancing and other non-litigious work
creates a temptation for less than a hundred per cent effort and, as Hazel Genn
has shown, for cajoling clients into under-value settlements so as to get paid more
quickly than in case of protracted proceedings. The original thalidomide claimants
did not give careful consideration to the choice of a solicitor with the special
expertise and experience their cases required, and there has been criticism of the
way in which he dealt with the matter. This points to two critical failures of the
afore-mentioned image of superior client representation. First, the choice of
solicitor is a blind man's bluff in the absence of advertising or other facilities for
identifying specialists in personal injury litigation. Fortunately, the intervening
years have seen a great improvement in this regard, so much so that it has brought
upon the Law Society angry complaint from the BMA.6 Secondly, the critical task
of assembling the evidence falls to the solicitor; the barrister may not even
establish contact with witnesses prior to the trial and is therefore entirely
dependent on, and may be handicapped by, the solicitor's preparation.7

The solicitor's task is aggravated by his inability under English procedure to
obtain judicial assistance for discovery of evidence (other than of documents).
Under American practice wide-scale discovery, including the deposition of
adversaries and witnesses, is a routine procedure with the result that all relevent
evidence available from all sources is likely to be known to both parties prior to
trial. By contrast, the thalidomide defendants successfully employed every legal
device to withhold information and forbade the use of information from other
sources such as the contemporaneous German proceedings.

The illuminating forensic history of the thalidomide affair by the Insight team of
the Sunday Times, Suffer the Children, contains a startling indictment of ineptitude

6. See The Times, 18 April 1988, protesting the Law Society's public call on would-be claimants to
contact lead solicitors. A recent publication, The Legal SOO (1988), addressed to potential solicitor
recruits, contains information on firms specialising in personal injury work.
7. Defending that rule as ensuring the barrister's primary loyalty to the Court, is a letter by Mr Gray

QC in The Times, 18 April 1988.

39



THE DENNING LAWJOURNAL

in the cause of the claimants. Nor was it confined to the matters already
mentioned. The most egregious was the alleged pressure put on clients to accept
the defendant's pusillanimous settlement offer which was conditioned on
unanimous acceptance. This included the threat of loss of Legal Aid - an
unpropitious model for current proposals to channel Legal Aid to a "lead team" in
mass litigation.8 The most troublesome aspect of this story is that these flaws are
systemic and unlikely to be effectively remedied without radical reforms affecting
the legal profession and legal procedure. It compares most unfavourably with
American legal representation on behalf of tort claimants, conducted by
specialised and publicity seeking plaintiffs attorneys driven by the incomparable
incentive of the contingent fee. The current proposal, discussed later, for
modifYing Legal Aid can only marginally affect this comparison.

In summary, then, the eventual successful outcome for the thalidomide victims
was achieved, not through but notwithstanding the legal process. It was due to the
dedicated pursuit of a handful of doctors, scientists and reporters who ultimately
succeeded in arousing public indignation through the media and Parliament, and
forcing Distillers to capitulate to terms of social justice.

The second lesson was closely related to the first, viz. that the settlement did not
reflect principles of tort law so much as superseding notions of social justice,
consonant to the imprecation of the Sunday Times that "there are times when to
insist on the letter of the law is as exposed to criticism as infringement of another's
legal rights.,,9 It will be recalled that the original settlement offer was discounted
to 40% in part because of evidential uncertainties brought about in large measure
by the defendant's own conduct. This blanket of information in particular
obscured proof of the defendant's testing procedures and knowledge or suspicion
of the drug's teratogenic potential, elements essential for proof of negligence.
Beyond that lay the problem of a duty of care to the unborn, already alluded to.
The only official response to the public indignation over thc thalidomide tragedy
(apart from a small grant to the trust fund in order to offset income tax) was the
appointment of a Royal Commission, the Pearson Commission, to report on the
general problem of Civil Liability and CompensationJor Personal Injury. But though
its Report in 1979 recommended, inter alia, the introduction of strict liability for
harmful products to cover situations like thalidomide, it remained stillborn. When
eventually strict liability was introduced, it was under compulsion of an EC
Directive. Moreover, it was the British Government which insisted on the optional
incorporation of a "development risk" defencelO and further diluted that defence

8. Another, 'IOrribile die/II, was an attempt, foiled only by the Court of Appeal, to make the children of
dissenters wards of the court and thus obtain judicial co-operation. See SuJJer Ihe Children, SIIPra n.l,
ch.1 I.
9. 24 September 1972. The writer had in mind that the defendant's last year profits were £64.8

millions and that its assets are worth £421 millions.
10. Allowing the defendant to plead that "the state of scientific and technical knowledge at the time
when he put the product into circulation was not such as to enable the existence of the defect to be
discovered": Art. 7(e). The defence has not been adopted by the Francophone countries and the
Netherlands.

40



MASS TORTS

in the course of enacting it into the domestic legislation of the Consumer
Protection Act 1987.II Thus in the end, twenty-five years of soul searching have
brought us little, if any, nearer to matching the law to popular expectations, and
bode ill for any replay, heaven forbid, of the thalidomide tragedy. A recurring
refrain calls for no-fault compensation for dangerous drugs, such as was put into
place in Sweden and Germany in the aftermath of thalidomide. 12 But why, it has
been asked, stop at drugs and not extend the programme to all products; indeed,
having regard to need, why differentiate between any causes of injury for accident,
or, better still, between incapacity from whatever source? 13

Opren
Once again a licensed drug, developed abroad, for dealing with arthritic disease

was withdrawn from the American and British markets amidst accusations that it
caused serious side effects, including even death. The American manufacturers
denied responsibility, contending that all but 30 among 1,300 British claimants
suffered symptoms not attributable to the drug and that these and others were
caused by excessive doses being prescribed. Attempts by 450 British claimants to
have their cases tried in Indiana, headquarters of the manufacturer Eli Lilly, were
rejected by an Indiana court in 1983 on grounds ofjOrum non conveniens. Last year
Eli Lilly offered a settlement of £21/4million for the 1,200 odd surviving claimants,
which was harshly attacked by their lawyers and the media on grounds both of the
defendant's profitability and the "grotesque disparity" in the treatment of
American and British claimants, the former being paid "gigantic sums" including
a $6 million verdict, while the latter were receiving an everage of only $1,800.14

Conspicuous in the Opren controversy was the impecuniosity of many
claimants, which at one time threatened abandoning their c1aimsl5 until a good
fairy in the shape of Mr Godfrey Bradman came to their rescue.16 Their plight
emphasized the failure of the Legal Aid system to ensure access to justice for that
vast portion of the population who were just above the poverty line but lacked the
resources to pursue claims in lengthy litigation with a risk, in case of failure, of
having to pay the fees not only of their own lawyers but also the defendant's.
Indeed, even plaintiffs under legal aid have to face the risk of having to bear a, for
them, frightening percentage of legal costS.17

11. Section 5(e): " ... was not such that a producer ... might be expected to have discovered". This
version is currently being contested by the E.C. Commission.
12. See Fleming, "Drug Injury Compensation Plants", 30 Am. J. Camp. L.297 (1982).
13. See Stapleton, Disease and the Compmsation Debate (1986).
14. Times, II December 1987.
IS. This was precipitated by a ruling (C.A.) requested by the Government, that the legal costs would
have to be borne per capita by all claimants, including those not qualifYing for Legal Aid: see Law
Magazille, 12 June 1987.
16. See Law Magazine, 26 June 1987.
17. Say, 10 per cent of an estimated £1.5 million in the whooping cough vaccine litigation. Total costs
were estimated at £5 million: Law Magazine, 2 October 1987. Successful plaintiffs do not derive any
benefit from Legal Aid because the residue of legal costs (after the costs taxed against the defendant)
are retained by Legal Aid out of the award: see Davies v. Eli Lilly rs Co. [1987] 1 W.L.R. 1136, at
p.1140 ("Legal Aid helps those who lose cases, not those who win them": per Lord Donaldson M.R.).

41



THE DENNING LAW JOURNAL

Their condition compared, in the view of media commentators, very
unfavourably with that of American claimants. The latter had - so they said - the
advantage, first of all, of strict products liability. In reality, the more difficult hurdle
was that of causation, from proof of which strict liability would no more exempt
claimants than a regime of fault liability. We shall see later how American law has
sought to ease that burden. In any event, a more significant advantage enjoyed by
American litigants was the availability of the contingent fee, hailed as "the key of
the door to the courthouse". Under what Goodhart has called the "American
rule", each litigant bears his own lawyer's fees and there is no fee shifting from
victor to loser. Moreover, in tort cases, the plaintiff does not even have to pay his
own lawyer if he loses, the risk being thus borne by the attorney rather than his
client. Since individual plaintiffs are generally risk averse, if not altogether unable
to pay legal fees, this arrangement removes the fearful consequences of losing
their case, and indeed provides a partial explanation for American litigiousness
and a greater willingness to pursue marginal causes. There is, of course, a price to
be paid: if successful, the plaintifPs lawyer becomes entitled to a portion of the
award - one third to one half - in order to remunerate him for his efforts in this
and other, unsuccessful cases. Thus, in a sense, the successful litigant subsidised
the unsuccessful; the lawyer, for his part, is an entrepreneur who can evaluate the
odds and spread the risk among a larger number of clients. 18

For several reasons, the most cogent being tradition, the contingent fee has been
condemned in England (and many other countries) as champerty, in company with
such other un-Christian vices as usury and gambling. Quite frankly, access to
justice has not traditionally ranked as an important social goal in this country. Not
only is actual litigation deplored and discouraged, but the very pursuit of legal
claims is viewed with ambivalence. This has borne down hardest on the
low-income classes and accounts for the systemic (even systematic) denial of
justice for tortious injuries during the nineteenth century and beyond. In order to
correct this historic injustice the Times leader writer, among others, therefore
advocated the "careful" introduction of contingent fees, especially in cases such as
Opren: "it was reasonable to ask lawyers to share some of the risks; their advice
would be none the worse for it", and "it would be a mistake to regard ['the
ambulance chasing' lawyer] as the complete refutation of the principle on which
the system is based." 19

Some years ago Justice, the association of reform-minded lawyers, unsuccessful-
ly argued for a modified version of contingent fees in its submission to the Royal
Commission on Legal Services.20 A more propitious time was now when the
clamour over Opren had aroused public concern.21 This was a mass tort with

18. See Fleming, The Americall Tort Process (1988), ch.6.
19. Times, 16 December 1987.
20. Justice. Lawyers and the Legal System: A Critique oj Legal Services ill Ellglalld and Wales (1977). See
White, "Contingent Fees: A Supplement to Legal Aid?'" 41 M.L.R. 286 (1978).
21. See my article "How Enterprising Lawyers Could Help the Less Well Off', Fillancial Times, 11
March 1988.

42



MASS TORTS

more than a thousand plaintiffs. If, instead of handling each claim as a separate
unit, they could be aggregated as in an American class action, and a team of lead
counsel could represent the whole class before the court, considerable economies
could be achieved and legal aid, funnelled to that team, be afforded for the benefit
of all. This was the new proposal of the Law Society to which in substance the
Lord Chancellor gave his blessing in an amendment to the Legal Aid Bill last
February.22 Procedural changes on the lines of a class action, he added, could be
accomplished without legislation by Rules of Court. We shall try presently to see
what can be learned from American experience in such aggregated proceedings. In
any event, it is doubtful if su.::h a reform would have produced a different and more
satisfactory resolution of the Opren litigation.

On 9 December 1987, Hirst] took the unusual step of canvassing in open court
the terms of a settlement reached by the six leading firms of solicitors representing
the main body of plaintiffs and Eli Lilly's solicitors, with the agreement also of the
Government defendants, the Committee on Safety of Medicine and the licensing
authority.23 After stating that the Court was neutral, Hirst] urged all plaintiffs in
the strongest terms to accede lest the settlement break down, and warned that their
legal aid might otherwise be discontinued. You will recall a similar episode when
the lead counsel for the thalidomide plaintiffs made the same point in order to
corral all into a unanimous settlement. This tactic raised not only serious concern
on the early occasion about counsels' conflict of interest but also the larger
question whether legal aid does not unduly subordinate the interest of individual
litigants to administrative efficiency and what is perceived to be the best interest of
the group as a whole. This problem would become endemic under the
contemplated modification of legal aid in multi-plaintiff litigation, where all
claimants are represented by only one legal team. Does such a team owe loyalty to
each of its many clients? In the United States critics of class actions, among them
many plaintiffs lawyers, raise the same objections, though they are often suspected
of concern more for their own fees than their clients' welfare.

THE AMERICAN EXPERIENCE
Class actions

In the light of American experience,24 English advocates of class actions may be
promising themselves too much. Those responsible for the introduction of class
actions by the Federal Rules of Civil Procedure in the mid-1960's did not
contemplate their application to tort litigation at all. Their principal reason was
that tort claims even in mass accidents were unlikely to satisfY one of the essential
conditions for class certification, viz. commonality, because damages vary with
each plaintiff and are central to their claim. Even more, in mass product disasters
causation looms large and offers few, even subsidiary, common issues. To revert to

22. Davies v. Eli Lilly ($ Co., N.L.]. Law Rep., 18 December 1987, p.1l83 .
. 24. See in more detail Fleming, supra n.18, ch.7.

43



THE DENNING LAW JOURNAL

the Opren case by way of illustration, even if the drug was capable of causing any of
the harmful effects complained of, the question whether it actually caused them in
each individual case would depend on numerous individualistic factors varying
with each patient, such as that his or her deteriorating vision might have been the
result of the synergistic effect of other drugs or of consuming excessive doses or
just of ageing, not unnatural with a group of arthritic patients. Thus, only a
negative answer to the only common issue, viz. was the drug capable of causing any
of the injuries, could have disposed of all claims uno iau (in one blow) as indeed it
did in the recent whooping cough test action.25

Another reason for regarding class action treatment as unsuitable for tort claims
is that its primary purpose is to make claims for minor losses, which individually
would not be worth pursuing, viable when aggravated by numerous plaintiffs, such
for example as systematic overcharging of interest by credit card issuers or
discriminatory practices in employment or leasing. Class actions for such
"non-viable, irrecoverable" claims provide an incentive for private law
enforcement of social policies and thus occupy a place along with punitive
damages and fee shifting (allowing recovery of legal costs, including attorney fees,
from the loser as an exception to the general rule). By contrast, tort claims are
usually sufficiently substantial in amount to attract lawyers on a contingent fee and
can therefore be left to their own devices.

American appellate courts have therefore hitherto given no encouragement to
certification of class tort actions, to the chagrin of many trial judges overwhelmed
with a flood of claimants and the discordant voices of their lawyers, many of whom
have reason to dislike class actions because it will adversely affect their fees unless
they participate as lead attorneys. More official support has been forthcoming for
class actions under a different qualification, viz. where a common fund is
inadequate to satisfy all claims, as where the defendant's assets are liable to be
exhausted by successive awards of punitive damages and late-comers are at risk of
recovering nothing. Notably, it is only in this category that all claimants, willy nilly,
can be forced into a class action; in cases of commonality so-called "opting-out"
must be tolerated for constitutional and other technical reasons.

As a result, almost all the spectacular mass tort trials in the United States have
relied on other than class action certification. Only in the Agent Orange case, among
the more conspicuous mass tort cases, was the trial judge's certification of a class
action upheld on appeal; I will later return to other, far more dramatic innovations
engineered by the redoubtable Judge Weinstein in the course of ultimately putting
a painful and controversial settlement to bed.

The most common aggregative proceeding is the consolidation of claims,
especially by multi-district panels which can assign elaims from federal registries
all over the country to a single judge for pre-trial disposition. This includes not
only the broad-range discovery possible under American procedure (thereby

25. See /11 re Paris Aircraft Crash of March 3, /974,399 F. Supp. 732 (D.C. Cal. 1975).

44



MASS TORTS

saving a great deal of duplication), but extends also to motions on the pleadings
and even of summary judgment on particular issues, such as choice of law. Indeed
in practice, it almost invariably culminates in a settlement into which the judge
cajoles the more or less reluctant attorneys and their clients. A paradigm is the
Turkish airline litigation before the late Judge Pierson M. Hall, a highly
experienced aviation judge.26 Thus, just as in England,27 consolidation
accompanied by one or more test trials on an issue oflaw or damages and followed
by settlement remains the usual mode of disposition of mass tort claims.

Another procedural device, first attempted by the Johns-Manville Corporation
in coping with asbestos c1aims,28 and later followed by the A. H. Robbins Co.,
manufacturer of the Dalkon Shield,29 is to seek shelter under Chapter 11 of the
Bankruptcy Act. This permits corporate reorganisation where creditors' claims
exceed the resources of the enterprise but it is in the creditors' interest to
encourage its continued functioning in order to obtain additional compensation
out of future profits.

Burden of proof
One of the most frequent difficulties encountered in mass accident cases relates

to proof of causality. It has two aspects. First, there is the problem of the
indeterminate defendant. It is often unclear which of several manufacturers of,
say, a drug produced the particular unit of the product that harmed the plaintiff.
The generic character of the product, the inconspicuousness of the exposure
event, and the long latency period frequently prevent precise identification of the
responsible manufacturer. Secondly, there is the problem of the indeterminate
plaintiff. Especially in pollution cases, the plaintiff can often rely only on general
statistical information to suggest that the defendant's emission merely increased
the number of sufferers beyond those who could have contracted the disease in
any event from other human agents or perhaps legally non-responsible
background risks. Does this sufficiently identifY the plaintiff as one injured, rather
than merely threatened, by the defendant?

The traditional requirement that the plaintiff prove causality against each
defendant on a balance of probabilities reflects our notions of procedural fairness
in the individualized confrontation typical of random accidents. It is argued,
however, that this "rule is neither rational nor a just means of resolving the
systematic causal indeterminacy presented by mass exposure cases.,,30

27. One of the earliest sensational test actions arose from the Thetis submarine disaster in which 99
lives were lost (see DUllcall v. Cammell Laird [1942] A.C.264). Recent examples are Thompsoll v. Smiths
Ship Repairers [1984] QB.450, involving more than 20,000 claimants for loss of hearing in shipbuilding
work, and Loveday v. We/kome Foulldatioll, supra n.25, the whooping cough case.
28. See Hensler, Felstiner, Selvin and Ebener, Asbestos ill the Courts: the Challenge of Mass Toxic Torts
(Rand Corp. 1985).
29. See Mintz, At AllY Cost: Corporate Greed, Women alld the Da/koll Shield (1985).
30. Rosenberg, "The Causal Connection in Mass Exposure Cases: a 'Public Law' Vision of the Tort
System", 97 Han'. L. Rev. 849, at p.858 (1984).

45



THE DENNING LAW JOURNAL

This postulate calls essentially for modification of conventional substantive law
in order to exploit the procedural advantages of class actions in mass tort cases. To
what extent has substantive law in America already bent to this challenge?

1. The /ndetenninate Deftndant
Modifications of the conventional rule, which places the burden of proof on the
plaintiff to identify which one among a group of potential culprits was responsible
for his injury, have actually preceded the advent of class actions. Most of these
emanated from California, and especially the more radical of them have not, at
least not yet, been widely followed elsewhere. While the earlier cases involved
random accidents, the problem is destined for a more prominent role in products
liability claims involving design defects, as already shown by the DES, Agent
Orange and asbestos cases.

Alternative liability
The earliest, so-called "alternative liability", theory originated in the case of

Summers v. Tic? 1 where two hunters, using shotguns, fired simultaneously in the
direction of the plaintiff, one shot putting out his eye. The court reversed the
conventional burden of proof, holding that where a single injury has been inflicted
by one or the other of two negligent defendants, but the plaintiff cannot prove
which one, it was for each of them to exculpate himself by establishing on a
balance of probabilities that he was not the one. The rationale of this decision was
that the equities between an innocent plaintiff and two negligent defendants, each
one of whom could have oaused his injury, favour placing the risk of proof
uncertainty on the latter.32

It has been questioned whether this principle should be confined to two
defendants, in which case the odds on either one's being the culprit are 50:50.
Contribution could ensure that each bore 50 per cent of the loss, so that the extent
of each one's liabilitywould in effect reflect the probability of his having caused the
injury. While we are in general reluctant to accept statistical proof of culpability,
particularly on the question of identification, those concerns have much less
weight in application to defendants whose negligence has once been established.
Moreover, matching the extent ofliability to the degree of probable causation is an
accepted rule for assessing damages for future contingencies.33 Thus the chance
of future arthritis or epilpesy, even ifless than "more probable than not" (51 per
cent?), justifies an award, not for 100 per cent, but for the discounted value of its
probability (which may be more or less than 50 per cent). Applying the same

31. 33 Cal.2d 80, 199 P.2d I (I948);followed by the Supreme Court of Canada in Cooky. Lewis [1951]
S.C.R.830.
32. Restatement, Torts, Sec011d 54338 (1965). This principle has been repeatedly applied in chain
collision and water pollution cases. It is also behind cases which shift the burden of proof to a negligent
tortfeasor to show for how much of the damage he is not responsible.
33. This is well established in English law: see Fleming, The Law of Torts 7th ed. (1987), at p.206. Not
so well in American Law: see King, "Causation, Valuation and Chance", 90 Yale L.J. 1353 (1981).

46

http://www.ingentaconnect.com/content/external-references?article=0044-0094(1981)90L.1353[aid=2967649]


MASS TORTS

rationale to proof, uncertainty on causation - both with respect to the question of
whether it would have made a difference had the defendant been careful34 and to
the present question of which one of several negligent actors caused the injury - is
therefore not as great a departure from conventional premises as might have first
appeared.

On the other hand, the principle would not reach to cases where the alternative
cause or causes are of innocent origin as in Wilsher v. Essex A rea Health Authority.35
That was the case of a prematurely born infant suffering from various illnesses
including oxygen deficiency. While in intensive care he was negligently given
excessive oxygen and later discovered to be suffering from damaged retinas, a
condition that could have had other causes. The House of Lords refused to shift
the burden of proof to the defendant on the issue of causation as the lower courts
had done, merely because the defendant's negligence had entailed a substantial
risk of being responsible. That this decision, though couched in rather
conservative terms, does not foreclose a different result when the alternative
causes are all of negligent origin could look for support to another pair of cases36

whose compatibility rests on just such a distinction in the context of "superseding
causation". Whether it will is a question raised in a pending appeal to the Lords in
a case where the plaintiff suffered tetraplegia after being struck successively by two
negligent drivers.37

Market share
The most innovative theory to date was launched by the California Court in

Sindell v. Abbott Laboratories.38 Having rejected all the preceding precedents,
including that of joint enterprise, as unsuitable for application against the more
than 300 manufacturers of DES because they would have exposed each of them to
joint and several liability for every injury caused by a "defective"genetic drug, the
Court discerned a more equitable solution in limiting each manufacturer's liability
merely to its market share. That way, when all claims had been satisfied, no one
defendant would have had to pay for more injuries than were statistically
attributable to him.

A number of objections have been raised against this solution. Perhaps the most
formidable is that it departs from the prior art not merely by lacking all precedent
but by being incompatible with the traditional notion of tort as a system of
individual responsibility. This was not corrective but distributive justice.
Allocation of responsibility was based no longer on proof of particular but of
statistical causation. Despite the Court's disavowal, this was indeed an

34. However, the House of Lords have recently declined to do so: Wilsher v. Essex A.H.A. [19881 2
W.L.R. 557; see also Kay v. Ayrshire (5Arrall H.B. [1987J 2 All E.R. 417 (H.L.).
35. Supra n.34.
36. Joblillg v. Associated Dain'es [1982] A.C.794; Baker v. Willoughby [1970J A.C.467.
37. 26 Ca1.3d 588, 607 P.2d 924 (1980).
38. See Filzgerald v. Lalle [1987] QB.781 (C.A.).

47



THE DENNING LAW JOURNAL

industry-wide liability. Even if defensible in terms of economic efficiency,39it did
not conform to basic notions of individual justice.

Secondly, the assumption that all would work out at the end of the day was
wishful thinking. The Court was content with the plaintiffs joining the
manufacturers of a "substantial share" of the DES market, apparently viewing
even 70/80 per cent as too ambitious. Moreover, it was left uncertain whether a
plaintiff could still collect the whole of his judgment from anyone defendant or
was limited to the latter's market share. In the first eventuality, the cost of securing
contribution and the risk of insolvency beyond his own share would still be borne
by each defendant. How does this really differ from solitary liability except insofar
as contribution will be regulated by reference to market share rather than other
possible criteria of responsibility?

The Indeterminate Plaintiff
In the preceding situations the plaintiff knew that he had suffered injury as a result
of another's tort but did not know precisely whose. This, the problem of the
indeterminate defendant, has its converse image in situations where the claimant is
one of several victims, only some of whom have been injured by a single tortfeasor,
but who are unable to say which one among them. To illustrate this, the problem
of the indeterminate plaintiff, suppose she is one of a group of persons exposed to a
toxic emission from the defendant, but the same symptoms also emanate from
independent "background risks". For example, in the Nevada Nuclear Explosion
case the plaintiffs could point to a strong positive association between their cancer
and exposure to ionizing radiation, but their cancer was indistinguishable from
that also prevalent and attributable to unknown causes.40 Similarly, in the Agent
Orange case, dioxin was present in the Vietnam countryside besides the amount in
the defoliant used by the US forces, procured from seven identified American
chemical companies.41

Typically, the association of the injury with the defendant's activity rests on
statistical rather than specific (anecdotal) evidence. Thus the evidence may show
that, after the defendant's emission, the incidence of the particular disease rose
from 100 to 190 for a given population. Here, doubts about statistical proof are
compounded by the fact that it does not even tip the balance of probabilities, i.e. 50
per cent plus. In the wake of Sindell, proposals have been made to apply a
mirror-image solution to the instant problem so that the defendant would be held
responsible for, and the plaintiffs as a group could recover, 9/19 of their injuries.42

39. See Calabresi, "Concerning Cause and the Law of Torts", 43 U.'Chi.L.Rev. 69 (1973), especially
at pp.84-91.
40. Allen v. U.S. 588 F.Supp. 247 (D.Utah 1984). The same applies to adverse reactions from many
drugs, without any clear distinction between iatrogenic and spontaneous illness: see Newdick, "Strict
Liability for Defective Drugs in the Pharmaceutical Industry", (1985) 101 L.QR. 405, at pp.420-30.
41. In re "Agent Orange" Product Liability Litigation 597 F.Supp.740 (E.D.N.Y. 1984).
42. Delgado, "Beyond Sindell: Relaxation of Cause-in-Fact Rules for Indeterminate Plaintiffs", 70
Calif L. Rev. 881 (1982).

48



MASS TORTS

The proposed formula would exact from the defendant an amount precisely
proportioned to his share of responsibility for the total incidence of the disease in
the area. Besides loss spreading, it would promote deterrence and economic
efficiency by internalizing the accident cost to the enterprise that is in the best
position to reduce accidents and pass on the cost to its beneficiaries by means of
insurance and higher prices.

Rather less satisfactory is the solution at the plaintiffs' end. Proportional
recovery, by which each member of thL class is compensated in proportion to the
damages sustained by the class as a whole, undercompensates some (90 in the
preceding example) and overcompensates others (100). But this is still better, so it
is contended, than either to compensate none or to compensate all for the full
amount of their injuries.

The first to adopt this rationale was the ever innovative Judge Weinstein, in
certifYingAgent Orange for class action treatment. 43 The departure from traditional
concepts, propounded in Agent Orange, is indeed manifold and startling. On the
basis of mere statistical evidence of a product's propensity for injury, it sanctions a
cause of action by unidentified plaintiffs against unidentified defendants without
specific proof of the defective nature of the product or of its having caused injury
to a particular plaintiff. In short, most elements of products liability have been
collapsed into mere statistical proof of causation.44 While it is true that, strictly
speaking, the Court's reasoning related only to the fairness of the settlement, it
sought approval for an approach to liability that would sever most links to
traditional tort principles.

Afterthought
The prospect of enlisting class actions for the radical solution of social problems

envisaged by these proposals enjoys far from universal support. For even if the
goals are worthy, to entrust such drastic legal change to a selection of activist
judges instead of to the traditional venue for political decision in a democracy
challenges accepted constitutional understandings. The distance separating these
perspectives is nowhere more strikingly illustrated than by the justification for the
Agent Orange settlement enforced by Judge Weinstein:

"Even though the evidence presented to the court to date suggests that the
case is without merit, the testimony of almost 500 witnesses undoubtedly did
serve once again to bring to the public's attention how unfairly Vietnam
veterans have been treated. They have been abused, rejected and humiliated
after serving bravely. Their voices should be heeded by the government and
public for whom they fought ...

43. Supra n.40.
44. Sherman, "Agent Orange Problem and the Problem of the Indeterminate PlaintiW', 52 Brook/JIll L.
Rev. 369 (1986), at p.390. .

49



THE DENNING LAW JOURNAL

Whether or not that pain was caused by Agent Orange, it is shared by a
disproportionately large number of Vietnam veterans. They and their families
should receive recognition, medical treatment and financial support ... The
public received the 'benefit' of combat service and should help to defray the
cost ...

Our country is rich in public and private resources of every kind. Those
resources should be made available to members of the class.,,45

This statement not only turns its back on any dichotomy between principle and
policy in judicial decision making, but invokes a goal of social psychology as
justification for wealth re-distribution, far removed from any conventional
objective of accident compensation policy, let alone of tort law. One might be
tempted to dismiss this scenario as but illustrating the jazzy strain of American
jurisprudence, but should be chastened by the memory that the thalidomide
settlement also fell well beyond the "shadow of the law", though lacking so
indiscreet a spokesman.

Mass litigation is not the only solution for mass accidents. Whether the
procedure is individualized or aggregative, the tort system reveals its inefficiencies
in starkest colour in dealing with mass claims. The funds available for
compensation are limited by the resources of the defendants, including liability
insurance cover. That even industrial giants can be driven into bankruptcy has
been translated from rhetoric to reality in the wake of the asbestos litigation. The
most depressing feature is that the exorbitant cost of administering the tort system
not only threatens the survival of industries peculiarly exposed to risk of mass
claims, like the pharmaceutical and chemical industries, but depletes the available
funds for compensating victims by staggering litigation costs. To the extent that
traditional rules are already being modified in order to facilitate recovery by
victims, the tort system is being distorted, even superseded. If the conventional tort
law is thus proving itself inadequate to the task, should we not, instead of merely
tinkering with it, consider the more radical solution of entirely replacing it by a
compensation scheme?

45. Supra n.39, at pp.857-58.

50