The Consensus Principle1
This paper, which was Commended by the
judging panel for the SCL Hudson Prize 2002, is published
by the Society of Construction Law and may be read on the
Society's website, www.scl.org.uk.
Maior privato visus dum privatus fuit, et
omnium consensu capax imperii nisi imperasset. Tacitus,
of the Emperor Galba.
Judges sometimes reach unexpected decisions,
declaring the law to be different from what most lawyers,
and their clients, previously thought it to be. The Court
of Appeal, when construing construction agreements, has seemed
especially prone to doing this.
During a remarkable six-month period in 1994,
the Court of Appeal delivered a series of three decisions
which (to say the least) surprised construction lawyers and
their clients: Trafalgar House Construction (Regions) Ltd
v. General Surety and Guarantee Co Ltd2
on 22 February; Perar BV v. General Surety and Guarantee
Co Ltd3 on 22 April; and
Crown Estate Commissioners v. John Mowlem & Co Ltd4
on 29 July5. The first two cases
related to performance bonds; the last to the final certificate
issued under a building contract. (The House of Lords reversed
Trafalgar House on appeal.)
Construction lawyers' reactions (described
below) to all these decisions suggests that consensus about
the meaning of the agreements construed by the Court of Appeal
already existed in the construction industry and among construction
lawyers: consensus to which the Court of Appeal's decisions
appear to have run counter.
Construction lawyers, and other construction
professionals, may argue about whether the Court of Appeal
did or did not buck the consensus in each of the three 1994
cases. It is however difficult to avoid the conclusion that
it did depart from prevailing opinion in Trafalgar House,
since not only did the Court of Appeal's judgment run counter
to the House of Lords decisions in the twin cases of Trade
Indemnity Co. Ltd. v. Workington Harbour and Dock Board6
and Workington Harbour and Dock Board v. Trade Indemnity
Co. Ltd (No. 2)7 but it
was in due course reversed by the House of Lords8.
The purpose of this essay is not to criticise
any of the judges, whether in the Court of Appeal or among
the former Official Referee judges, though some criticism
is unavoidable. It may not even be especially relevant whether
their decisions were objectively right or wrong. Perhaps the
consensus was right and the judges were wrong, or perhaps
the consensus was wrong and the judges were right9.
The problem is not that the judges took the wrong decisions;
it is that their decisions, whether objectively right or wrong,
differed from the then prevailing consensus, on whose basis
the parties had entered into the agreements in question.
The purpose of this essay is to encourage judges
to adhere to current consensus by proposing sound legal reasons
why they should do so; and to suggest how judges might depart
from a legally unsound consensus in a way which protects the
parties to existing contracts.
Trafalgar House and the temporary
demise of the conditional bond
Before Trafalgar House, construction lawyers and their clients
considered conditional performance bonds to be a guarantee
of the contractor's performance, and proof of the employer's
damages to be a prerequisite of the bond's enforcement10.
But the Court of Appeal held that such bonds are not guarantees
and that their clear purpose is to assure the employer of
immediate funds; effectively, that so-called conditional bonds
are in reality no different from on-demand bonds which require
the surety to pay whatever the employer asserts in good faith
to be his damages11.
Trafalgar House featured a traditionally
worded conditional bond in favour of a main contractor to
secure the performance of groundworks by a subcontractor.
The subcontractor failed to perform properly. The main contractor
asserted in good faith that his damages greatly exceeded the
amount of the bond.
The lawyer on the Fleet Street omnibus might
have expected the Court of Appeal's judgment to have followed
the decisions of the House of Lords in Trade Indemnity
and Workington, which concerned a bond similar to the
one in Trafalgar House but without the customary second
condition as to satisfaction and discharge by the surety12
and held that the claimant under a performance bond must prove
both the breach and the amount of its damages13.
If our lawyer on the No. 11 bus had forgotten
about Trade Indemnity and Workington he might have expected
the Court of Appeal to have based its decision on a proper
construction of the contract. Reading the virtual requirement14
(not included in the Workington bond) for the surety
to satisfy and discharge the damages sustained by the main
contractor up to the amount of the bond, he might have considered
it reasonable for the claimant, like most claimants, to have
to prove his damages.
The judge at first instance, Mr Recorder Knight
QC15, on 9 February 1993 gave
summary judgment for Trafalgar House for the full amount of
the bond on the grounds that there was no arguable defence
to the claim. The learned recorder's more detailed reasons
are not indicated in the Court of Appeal judgments.
On 22 February 1994 the Court of Appeal dismissed
the appeal by General Surety, holding that a bond is not a
guarantee and that its clear purpose is to assure the claimant
(at least if he is a main contractor)16
of immediate funds; that is to say, that the surety must pay
the sum (up to the amount of the bond) which the claimant
asserts in good faith is his damages.
The author has been unable to discover whether
counsel for General Surety before the Court of Appeal mentioned
the current consensus. But before the House of Lords counsel
did argue that ever since Calvert v. London Dock Co17.
conditional bonds have been regarded as guarantees; and that
the language of these bonds has an established meaning, which
should not be changed.18
The reaction of the legal profession to the
Court of Appeal's decision may be represented by the comments
of Richard Davis, a solicitor and consultant at Masons:
"The industry received a shock in 1994
when the Court of Appeal construed a bond in the ICE (5th
edition) form as a primary obligation …, until the
decision was overruled by the House of Lords." 19
A news report in Building on 18 March
1994,20 probably written with
the benefit of legal advice, described the Court of Appeal's
decision as "a surprise legal judgment [that] could convert
thousands of standard performance guarantees into onerous
on-demand bonds overnight." According to the report,
"industry sources estimate the top four insurance companies
alone have issued about 5000 such bonds worth £950m
to contractors and subcontractors." Frank McCormac, senior
legal adviser of Balfour Beatty, was reported as saying: "This
decision is wrong in my view and must be appealed."21
In the same issue of Building, an article
by Timothy Elliott QC seemed to endorse the Court of Appeal's
"cash flow" approach:
"What an employer wants from a bond is
an immediate cash payment."
But Mr Elliott also appeared to acknowledge
that the decision - though largely based on the "cash
flow" argument - did run counter to prevailing opinion:
"This is worrying stuff for bondsmen.
On the face of it, if the court holds that a claim is made
in good faith, the surety has to stump up. Pleas of set-off,
final accounts and challenges as to proof of loss will not
avail. This surprising decision is at variance with what
many had understood to be the law."
Hudson's Building Contracts (Eleventh Edition),
published before the House of Lords decision in Trafalgar
House, described the Court of Appeal's "on-demand"
interpretation on the wording of the bond as "startling".22
The House of Lords, on 29 June 1995, allowed
General Surety's appeal on the grounds that a conditional
performance bond is a guarantee and is therefore subject to
proof of damage in order to establish liability.23
If the Court of Appeal got it wrong in Trafalgar
House - and the House of Lords said they did - they may
be said to have got it wrong for two reasons. First, because
they didn't follow Trade Indemnity and Workington.
And secondly, because they misunderstood the commercial and
legal purpose of conditional performance bonds.
Perar and the innocent insolvent
Before Perar, our second case on performance bonds,
construction lawyers and their clients considered the contractor's
insolvency under JCT contracts to be a breach of the building
contract, entitling the employer to call any performance bond.
But the Court of Appeal held that the contractor's insolvency
is not of itself a breach of contract, but is rather a no-fault
event covered by an exclusive code under which the parties
contract out of their other remedies.
In Perar the bond was conditioned on
default by the contractor, whose employment under the building
contract was automatically determined on his going into administrative
receivership.
No directly relevant precedent was cited before
the Court of Appeal, but on a literal interpretation of the
bond, and the building contract to which it was collateral,
there appeared to be no breach or default where the contractor's
employment was automatically determined.
Nevertheless, the average construction lawyer,
in the author's submission, would probably have taken the
view that insolvency is a breach of contract and an act of
default. That view would have been correct, except for the
way in which JCT contracts at that time dealt with all forms
of contractors' insolvency, treating them in effect as no-fault
events.
His Honour Judge Michael Rich QC, giving judgment
on 18 December 1992, held that the employer had no cause of
action against the contractor because on determination the
contractor was neither entitled nor obliged to continue with
the works.
Deciding the appeal on 22 April 1994, the Court
of Appeal held that the contractor was not in breach or default
by becoming insolvent. In the absence of any saving for other
rights and remedies of the employer, the determination provisions
formed an exclusive code whereby the parties contracted out
of other remedies.
A distinguished construction lawyer described
the decision as "another judgment on conditional bonds
which produces a surprising if conventional result."24
The decision in Perar received adverse
comment by the learned editors of the Building Law Reports,
who wrote, "The practical effect of this decision is
that in a circumstance in which the employer most needs the
bond, ie the administrative receivership of the contractor,
the bond is unavailable."25
They did not, however, say that the decision was wrong, and
Perar may be a good example of a case where the decision
was right and the consensus wrong.
The editors pointed out that the wording of
such bonds, and possibly the JCT contracts, required reconsideration.
The relevant JCT contracts have subsequently been amended
so that automatic determination of the contractor's employment
occurs only where a provisional liquidator is appointed or
a winding up order is made.26
In the author's view, to complete the protection of beneficiaries
performance bonds should state that such events (or any other
event of insolvency) are deemed to be breaches of the principal
contract.
Crown Estate and the excessively
final certificate
Before Crown Estate, construction lawyers and their
clients considered the final certificate under JCT contracts
to be conclusive as to the absence of defects except where
the contract documents expressly require the architect's approval
of the quality of materials or standards of workmanship. But
the Court of Appeal held that standards and quality are inherently
for the reasonable opinion of the architect, so that the final
certificate is conclusive as to all such matters.
The JCT building contract in Crown Estate
provided that the final certificate (which the architect had
issued) shall have effect in any proceedings as conclusive
evidence that where the quality of materials or the standard
of workmanship are to be to the reasonable satisfaction of
the architect, the same are to such satisfaction.
On a literal interpretation of these words,
the final certificate is not conclusive except as to quality
and standards which the building contract says are to be to
the satisfaction of the architect. Of these there were no,
or virtually no, examples in the Crown Estate building
contract.
His Honour Judge James Fox-Andrews delivered
judgment at first instance on 10 December 1993. He adopted
a literal interpretation, holding that the final certificate
was conclusive only as to matters relating to standard and
quality which were expressed to be for the satisfaction of
the architect.
The Court of Appeal held, on 29 July 1994,
that all matters of standards and quality of work and
materials are for the reasonable opinion of the architect.
Therefore the final certificate was conclusive as to all such
matters.
The Court of Appeal's decision in Crown
Estate led to a firestorm of articles and comment which
persists to this day.
In February 1995 the respected legal columnist
in Building, Tony Bingham, came out firmly in favour
of the Court of Appeal, writing that Crown Estate had
"identified what the industry thought was the true position
all along": with the exception of lawyers, who, "or
at least some of them, do not like the decision at all".
Mr Bingham's belief, if correct, is an indication that clients
don't always agree with their lawyers; which raises the question
considered below: if the courts are required to ascertain
the current consensus on the effect of a particular common
form agreement, whose consensus should they be looking for?
That of construction lawyers, or that of their clients?
A noted construction solicitor has commented:
"The Court of Appeal caused consternation in the construction
industry."27
The eminent commentator on construction law,
Ian Duncan Wallace QC, noted that the Crown Estate
decision "… relates to wording wisely or unwisely
used in hundreds if not thousands of public contracts [and]
involves future cost to the taxpayer of probably hundreds
of millions if not billions of pounds through the uncompensated
cost of repairing defects in public buildings caused by breach
of contract …".28
Mr Duncan Wallace's comments highlight the
damaging effect which a "surprise" judgment on the
meaning of a common form agreement can have on large numbers
of parties who have entered into such agreements in the belief
that they had a different meaning.
Writing some years later, another leading commentator,
differing from Tony Bingham, described the Joint Contracts
Tribunal's subsequent amendment to its standard contracts,
limiting the "conclusiveness" of the final certificate
to very narrow grounds, as being "what most of the industry
thought the clause had meant in the first place".29
The Joint Contracts Tribunal did not rush to
adapt to the Crown Estate decision. But in March 1995
it issued a "Notice to Users", bringing the case
to the attention of users of its forms, and later issued formal
Amendments to its standard forms which inserted the word "expressly"
in the relevant place in clause 30.9.1.1.30
The Notice to Users records that the Court of Appeal's judgment
provided an interpretation of the effect of a Final Certificate
under JCT 80 which was much wider than the Tribunal had intended.
The surprise and criticism which the Court
of Appeal's decision met with reflected the fact, in the author's
respectful view, that despite the absence of any ambiguity
the court unnecessarily departed from a literal interpretation
consistent with the contract's purpose.
The hazards of construction
Judges, when construing agreements, either see themselves
as having no clear authority to follow a consensus among specialist
lawyers and their clients which they perceive to be wrong
in law; or are departing from a consensus because they are
unaware of its existence or are not especially interested
in it; or believe they know and are following the consensus
when in fact they do not know it and are not following it.
True to classical principles of construction,
judges infer the purpose of an instrument from what it says,
and they usually think it wrong to consider external evidence.31
However, judges - like the rest of us - cannot help forming
their own opinions about the industries with whose legal problems
they have become familiar.
The ex cathedra expression of such opinions
can lead judges into difficulty. When (for example) Saville
LJ bases his judgment in Trafalgar House on his own
understanding of the purpose of a performance bond, based
on his own knowledge of the construction industry - a question
of fact, or at any rate an external, or subjective factor
- he invites the obvious criticism that his knowledge might
be wrong.
It seems to the author that expert evidence
on such matters as the commercial and legal purpose of an
instrument is essential; and that it is for counsel to introduce
such evidence at first instance, and for the judge at first
instance to rule on it as a question of fact, binding on the
appeal judges except where it is wrong on the face of it.
The introduction of expert evidence as to the
commercial and legal purpose of a class of agreements lies
at the heart of the author's proposal for a "consensus
principle", which would require judges to ascertain that
purpose with the assistance of expert witnesses assembled
by the parties on counsel's advice. The commercial and legal
purpose of a class of instruments, so ascertained, constitutes
a "consensus", and judges should be loth to interpret
any agreement in such a way as to disturb that consensus.
Difficult questions arise which this essay
can make little attempt to answer. First, is the consensus,
which the judges should be seeking to ascertain, consensus
as to the commercial and legal purpose of a document or consensus
as to its legal effect? They are nearly, but not quite, the
same thing.
Secondly, is the appropriate consensus that
in the construction industry as a whole or the consensus among
construction lawyers in particular? As Tony Bingham's comments
about Crown Estate suggest, the consensus among lawyers and
the consensus among their clients can differ. On the one hand,
it is construction industry clients who tell construction
lawyers what they want and who take the risks involved in
using standard and common form contracts. On the other hand,
judges might consider the ascertainment of consensus to be
an extension of their existing practice of referring to academic
legal opinion, in which event it would be more appropriate
for the courts to consult the opinions of practising lawyers.
Canons of construction
In the author's opinion, each of the three 1994 decisions,
both at first instance and on appeal, turns on the canons
of construction of contracts adopted by the judges.
There is clear authority for the proposition
that the courts will uphold an established construction where
a contract is based upon a standard form of commercial agreement.
Lord Esher MR stated, in Dunlop & Sons v. Balfour
Williamson & Co.:32
"It is a wholesome rule that has often
been laid down that when a well-known document has been
in constant use for a number of years, the Court, in construing
it, should not break away from previous decisions, even
if in the first instance they would have taken a different
view, because all documents made after the meaning of one
has been judicially determined are taken to have been made
on the faith of the rule so laid down."
What is less clear is whether the courts should
uphold an existing consensus about the meaning of "a
well-known document [which] has been in constant use for a
number of years" where that consensus is not supported
by any previous decision.
In construing agreements, judges must consider
numerous matters, among them: the literal meaning of the words
used; the expressed intention of the parties; the legal background
against which the contract was made;33
the desirability of certainty in cases where the contract
is based upon a standard form of commercial agreement;34
terms implied by custom of the relevant trade; a construction
which does not permit one party to the contract to take advantage
of his own wrong; and the reasonableness of the result.
If the judges decide that they must ascertain
the relevant consensus and construe every contract in accordance
with it, this means not only that the judges must take judicial
note of the consensus, perhaps from their own knowledge (though
the Court of Appeal's reasoning in Trafalgar House
suggests otherwise) or from counsel for the parties, but preferably
(in the author's view) from expert evidence; but also that,
if the consensus represents what would otherwise be bad law,
the judges may find themselves unable to change it for fear
of disturbing existing agreements. And though it may be correct
to say that this is a matter for Parliament, in practice law
reform (especially of the ad hoc variety) has a low priority
on the Parliamentary agenda.
Resolving the paradox
Is there, then, any way in which judges can declare the law
to be different from the consensus without affecting existing
contracts, including the contract in the case before the judges?
Such an approach would appear to contravene
the doctrine whereby the judges declare what the law is but
do not change it; and that once the judges have declared what
the law is, that law must apply in the case before them, and
in future cases involving similar facts and agreements which
are not distinguishable.
The approach - which the author describes as
a precedent direction (the substantive equivalent of a practice
direction) - would enable the court in question, or a separate
judicial authority, to say, in effect: "We consider the
current consensus among construction industry clients and
their lawyers, as to the meaning of this class of agreements,
to be wrong in law; but the court must follow the consensus
in cases relating to existing agreements, since otherwise
the parties to those agreements will lose rights which they
thought they would have when they entered into them. However,
our own view of the law will apply to any agreements made
after (say) one month from the date of this direction."
Yet one only has to describe the problem, or
to endeavour to draft a legislative rule for enabling the
court to act in this way, to understand the inherent contradiction.
The prevailing consensus represents, in effect, the parties'
intentions. While the courts are satisfied that parties have
entered into an agreement in reliance on an understanding
of the law which accords with a prevailing consensus, then
under any intention-based rule of construction the courts
must continue to apply that consensus.
If however the court is unaware of the consensus,
perhaps because counsel did not draw it to the court's attention
or because no evidence was introduced to support it, then
the court's decision is per incuriam and should not
be considered legally binding.
Having argued the matter thus far, this essay
might itself be in danger of bucking the consensus if it tried
to suggest a definitive solution.
Perhaps the judiciary, if persuaded of the
merits, might consider itself already authorised to adopt
some of the approaches suggested in this essay.
In addition to the arguments already put forward,
Rule 1.1 of the Civil Procedure Rules lays down the overriding
objective of enabling the court to deal with cases justly,
which includes ensuring that the parties are on an equal footing.
A decision which overturns a prevailing consensus may be unjust;
and such a decision does not treat the parties on an equal
footing since both contracting parties must have entered into
their agreement in the knowledge and on the basis of the current
consensus, which the court has overturned to the unexpected
detriment of one party and to the uncovenanted benefit of
the other. Essentially, such a decision has the same potentially
unjust effect as retrospective legislation.
Moreover, article 1 of the First Protocol to
the European Convention on Human Rights provides: "Every
natural or legal person is entitled to the peaceful enjoyment
of his possessions. No one shall be deprived of his possessions
except in the public interest and subject to the conditions
provided for by law and by the general principles of international
law." It is not fanciful to suggest that "possessions"
includes rights in action or, perhaps, that to deprive a contracting
party judicially of contractual rights which both he and the
other parties thought they were granting or purchasing, might
be a breach of this article.
The author confesses himself unable to suggest
any way of allowing judges, without disturbing existing agreements,
to correct legal errors in the current consensus, except the
path of the precedent direction. The court would decide the
case in hand according to the current consensus; but would,
by means of a precedent direction give public notice that,
when construing agreements of the same kind entered into after
a date stated in the direction, it will rule according to
the view it would have taken had it not considered itself
obliged to follow the current consensus.
At all events, it is for the judiciary to recognise
the serious economic and human problems which can be caused
by its periodical departures from the prevailing consensus,
and to use its collective ingenuity to resolve those problems.
|
