Out-Law / Your Daily Need-To-Know

An important case where the Court of Appeal  considered exclusions and limitations contained in a systems integration contract in relation to UCTA.

Watford Electronics Limited v Sanderson CFL Limited (TCC)

  • [2000] 2 All E.R. (Comm) 984

Watford Electronics Limited v Sanderson CFL Limited (CA)

  • [2001] EWCA Civ 317
  • [2001] 1 All E.R. (Comm) 696
  • [2001] B.L.R. 143
  • (2001) 3 T.C.L.R. 14
  • [2002] F.S.R. 19
  • (2001) 98(18) L.S.G. 44

Help with citations


This judgment arose out of a hearing concerned with preliminary issues, although the facts are set out in some detail in both judgments.  The relevant background is as follows.

Watford sold electronic components, including computer products (PCs and peripherals) by mail order and also at two retail outlets.  At the time of the contracts in question, its turnover was around £21 million.  The business had been successful, by the time of the judgment turnover was around £25 million and the group of companies had a combined turnover of some £100 million.

Sanderson was part of a group that in 1991 had a combined turnover of over £60 million although Sanderson itself had a turnover of only some £4 million in 1992.  It supplied systems, including systems incorporating a product known as Mailbrain, a marketing package dealing with mail order processing, database marketing, telemarketing and wholesale order processing.  As at 1992, there were 120 customers for Mailbrain.  Furthermore, Mailbrain was marketed as a product that could be linked to other packages such as sales, purchase and nominal ledger and payroll.  It did this by a package called Genasys, which was itself an integrated suite of financial accounting packages.  Sanderson had produced Genasys both as an independent package and one which could be integrated with Mailbrain.

In 1992, Watford moved into new premises to accommodate its growth and quickly realised that it needed an integrated software system to deal with its various areas of business – mail order, retail sales, warehouse stock control and accounting.  This would facilitate the growth of the business and allow Watford to provide a better service.

The two parties came into contact with each other at a roadshow organised by Sanderson in February 1992.  Watford had just tried a different computer system where the implementation had failed after a week, and was therefore keen to ensure that there was no repetition of the experience.  While there was some dispute as to what was said in the meetings (and when), it was clear that the brochure together with pre-contractual correspondence and discussions would have left Watford with the impression that the system as offered would be fully integrated. 

Another issue that was discussed was the hardware to be used.  Again, the content of the conversations was disputed, with Sanderson’s salesman claiming he had only ever advised on the disk space that was required.  Judge Thornton found that Sanderson had not advised on the need for a minicomputer as opposed to PCs, in spite of the rising number of proposed users, and that the supply of the hardware was therefore in the context of Sanderson confirming that it was an appropriate specification for the proposed applications.

At least one reference site visit was made, and a demonstration took place at Watford’s premises.  The demonstration led to the requirement for at least some modifications.

In September 1992, there was a series of quotations from Sanderson, which reflected the discussions between the parties, especially as to price and the required modifications to Mailbrain and Genasys.  Sanderson was to supply Mailbrain, Genasys and its EPOS (electronic point of sale) system, together with 17 specific modifications, and a licence for 100 users.  The hardware was supplied by both parties.

Sanderson had sent through its standard conditions with a letter dated 23 September 1992 and these were discussed at a meeting on 28 September.  Watford was concerned with the limitations of liability, which excluded liability for consequential or indirect loss and limited other liabilities to the contract price.  Watford observed in negotiations that it then had a monthly turnover of £1.5 million, so any problems could cause considerable harm.  However, Sanderson said that the clauses were non-negotiable and in the end only agreed to add a clause to use best endeavours to allocate appropriate resources to the project.  Watford felt it had no option but to agree.

There were initially three contracts, dealing with sale of the hardware, licensing the software and also a “software modification licence”.  The three contract forms all had similar wording in them, and the one for the sale of equipment had these words (this report will refer to this numbering, although the original contracts had different numbers for the otherwise identical clauses):

“Neither [Sanderson] nor [Watford] shall be liable to the other for any claims for indirect or consequential losses whether arising from negligence or otherwise.  In no event shall [Sanderson’s] liability under the Contract exceed the price paid by [Watford] to [Sanderson] for the Equipment connected with any claim.

In addition to [the clause at paragraph 1] Sanderson CFL commit their best endeavours in allocating appropriate resources to the project to minimise any losses that may arise from the Contract.  [this was the addendum agreed in negotiations]

The parties agree that these terms and conditions (together with any other terms and conditions expressly incorporated in the Contract) represent the entire agreement between the parties relating to the sale and purchase of Equipment and that no statement or representations made by either party have been relied upon by the other in agreeing to enter into the Contract.”

The system went live in February 1993, but demonstrated serious problems even before that.  This culminated in Sanderson recommending that Watford acquire a minicomputer.  Watford executed two further contracts on 17 August 1993 for a Bull minicomputer and an increase of the licence to allow 100 users.  Both contracts included clauses (1) and (3) above, but not (2).

Watford claimed for loss of profits (£4,402,694), increased costs of working (£996,063) and the costs of acquiring an alternative system (£119,204).  There was an alternative claim for misrepresentation and negligence comprising sums paid to Sanderson (£104,596) and the increased costs of working (£996,063).


Judgment at first instance

Judge Thornton looked at the five contracts but concluded that in reality there was one contractual relationship.  It was not realistic to break the contracts down into separate obligations to supply software (with modifications), hardware, training and a licence.  The judge therefore thought that all the contracts should be read together, so far as reasonably possible.

The contractual package involved providing goods, services and software and so there was a “hybrid of implied terms”.  For goods and services, these are to be found respectively in section 14 of the Sale of Goods Act 1979 and section 13 of the Supply of Goods and Services Act 1982.  These were incorporated into the various forms of contract in this case.  Thus, there were the following implied terms:

  • The goods supplied would be of merchantable quality, namely as fit for the purpose or purposes for which goods of that kind are commonly bought as is reasonable to expect having regard to any description applied to them, the price (if relevant) and any other relevant circumstances.
  • The services supplied would be carried out with reasonable care and skill.

The contract could also have a further implied term:

  • The goods supplied would be reasonably fit for any particular purpose for which the goods should be acquired.

This last implied term depended on whether there was reliance by Watford on Sanderson’s skill or judgment.

The question raised in particular was whether the software itself had the implied terms of merchantability and fitness for purpose.  These could be implied by virtue of the contract for supplying software being a bailment or by virtue of terms implied at common law.

The definition of “software” in the contract included the physical elements (of the medium) and so it could be regarded as a contract to supply goods.  This being so, the contract fell within section 6(1) of the Supply of Goods and Services Act 1982 dealing with the hire of goods bailed.  Section 9 of the same Act provided for similar implied terms as under a sale of goods to apply (again, dependent on showing reliance).

Turning to the position at common law, implied terms were such as reasonably necessary to give business effect to the parties’ intentions.  Since implied terms of merchantability and fitness for purpose applied to the hardware element, it was a business necessity for the software to have been supplied under similar contractual arrangements.  Judge Thornton did not accept the argument that, under the common law, goods supplied under a contract of hire did not carry with them strict implied terms, but only an implied term that they should be as fit as reasonable care and skill could make them.

In general terms, the implication of implied terms would apply to such issues as compatibility and integration, although the full trial would have to establish the exact scope of these warranties.

Sanderson sought to argue that Watford, as a supplier of PCs and other similar goods and a supplier of some of the hardware in this case, was not relying on Sanderson’s skill and judgment.  This argument was rejected by the judge.  He referred to all the negotiations, including the brochure and the correspondence as showing that Watford was clearly relying on Sanderson in its advice and recommendations.

This was relevant given the judge’s finding that Sanderson’s brochure and various items of correspondence did include representations as to such matters as integration.

The clause at paragraph 3 was a clause that sought to exclude liability for misrepresentation.  Given that there had been representations that induced the contract, the clause was in substance an exclusion clause and subject to section 3 of the Misrepresentation Act 1967.

Sanderson was contracting on its “written standard terms of business” and thus fell within section 3 of UCTA.  The fact that one additional clause (paragraph 2) was included was not sufficient to take the case out of section 3.  The ambit of the additional obligation added by the addendum was narrow and insubstantial when seen against the remaining terms and conditions.

Most of the losses were “indirect or consequential”, and the combined limit of liability under the various agreements was about £140,000.  Watford was claiming direct losses of about £120,000, but the bulk of its contractual claim (nearly £5.5 million) was made up of elements that were “indirect or consequential”.  The reasonableness of the exclusion was therefore very important.

When looking at the question of reasonableness under UCTA, Schedule 2 set out some guidelines which are of use, even though they only strictly apply to sections 6 and 7.  Looking at the relevant guidelines and considering all the factors, Judge Thornton thought the following.

  • Both parties were of about the same size, and the representatives who negotiated the contract each had considerable experience.  There was therefore no particular inequality of bargaining power and neither party was under any particular pressure to contract with the other.
  • The market may have been a “buyer’s market” only to the extent that Watford could negotiate the price down.  In other ways, having regard to the terms and conditions, Sanderson was relatively inflexible in negotiating its liability clauses.  There was also evidence that such exclusions and limitations were standard at the time.
  • With regard to the competition, there were other packages available, but Mailbrain was the only one which appeared to fulfil Watford’s needs and there was no evidence to show that there was any other integrated package available elsewhere.
  • There was no evidence that Watford received any particular inducement to accept the liability terms, and the price reduction agreed seemed to have bearing on the matter.  Judge Thornton did not think that the fact of a price reduction reducing the profit margin had any bearing on the reasonableness or otherwise of the clause.
  • Watford was aware of the existence of the clause, though only at a late stage of the negotiations.  The only amendment that Watford could achieve was a “make-weight amendment”.
  • While there were modifications, these were minor in comparison to the standard system that was being supplied.  In any event, such work of modification was fairly standard for this type of software.
  • With respect to insurance, Sanderson had no general policy in place.  The evidence was that where Sanderson was forced to change its liability provisions and accept some limited liability, they took “one-off” insurance where they felt exposed.  In this case, Sanderson obviously assumed that Watford would sign the contract without having to make the amendment.  However, there was no evidence that Watford could have obtained insurance to cover its potential loss of profits resulting from Sanderson’s failure.
  • The fact that a discount was negotiated made no difference, as it was not in any way referable to the exclusion clause.
  • It was irrelevant that Watford’s own standard terms of business excluded consequential loss: the businesses were very different, and in any case Watford’s own exclusion might be held to be unreasonable.
  • It was also irrelevant to the question of reasonableness that Watford had made a reference site visit and received a demonstration.  Neither would have enabled it to assess whether acquiring such a system would be risk-free or not.
  • Sanderson was contracted to supply and configure the system.  It was up to them to use their expertise to assess the requirements.  The fact that a salesman could not verify all the facts from two visits did not mean that it was reasonable for Sanderson to limit its liability for what was in reality something within its own control.
  • There were other factors as well: that Watford was dependent on Sanderson’s expertise, that the exclusion clause effectively left Watford without a remedy, that failure was potentially very serious for Watford.

Taking all the factors together, Judge Thornton “unhesitatingly” found that Sanderson had not established that the clause was reasonable.  It was not justified by any particularly onerous or unusual liabilities that Sanderson might encounter, insurance was available to Sanderson and Watford could not have obtained software elsewhere without having to accept this clause since it was so common at the time.  The effect was to deprive Watford of any damages where there had been significant failures by Sanderson.

The judgment of the Court of Appeal (Chadwick LJ)

The appeal related only to the judge’s finding as to reasonableness of the exclusion and limitation contained in paragraph (1).

The Court followed the guidance given in George Mitchell (Chesterhall) Ltd v Finney Lock Seeds Ltd [1983] 2 AC 803 as to the approach an appellate court should take when reviewing a lower court’s decision as to a view of what was reasonable.  The appellate court should treat the original decision with the utmost respect and refrain from interference with it unless satisfied that it proceeded upon some erroneous principle or was plainly and obviously wrong.

In determining the reasonableness of the term, it is first necessary to determine the scope and effect of that term as a matter of construction. In particular, it is necessary to identify the nature of the liability which the term is seeking to exclude. The reasonableness of the term does not fall to be considered in isolation. It falls to be determined where a person is seeking to rely upon the term in order to exclude or limit his liability in some context to which the provisions of UCTA or section 3 of the Misrepresentation Act 1967 apply.

The two sentences in paragraph (1) are each intended to have their own separate and distinct purpose. The purpose of the first sentence is (at least) to exclude contractual claims for indirect and consequential losses; that is to say, to exclude liability in contract for losses which could be recovered only under the second limb of the rule in Hadley v Baxendale. These are losses which do not result “directly and naturally” from the breach, but which nevertheless, were or must reasonably be supposed to have been in the contemplation of both parties at the time the contract was made.

The second sentence was an attempt to limit the liability for loss directly and naturally resulting from breach of warranty to the price of the equipment/software, so avoiding an enquiry into what would have been the value of the equipment/software if the warranty had been fulfilled.  This corresponded with the warranties given in Sanderson’s terms – that the hardware should “perform in accordance with its specification” and the software would “during normal use perform the functions detailed within the manuals supplied”. 

This should be compared with the rule in section 53(3) of the Sale of Goods Act 1979, which says that in cases of breach of warranty, the prima facie measure of loss is the difference between the value of the goods at the time of delivery and the value they would have had if they had fulfilled the warranty.

Thus, Sanderson’s terms sought to limit liability for breach of warranty to the price of the equipment or software, that is to say, they tried, in the context of section 53(3) to peg the value which the goods would have had if the warranty had been fulfilled to the price paid by the buyer.

The purpose, therefore, of the first sentence of paragraph (1) was to exclude contractual claims for indirect and consequential losses i.e. those things that could only be claimed under the second limb in Hadley v Baxendale.  These would comprise losses not flowing directly and naturally from the breach, but which nevertheless were or must reasonably be supposed to have been in the contemplation of the both parties when contracting.

The entire agreement clause at paragraph (3) coloured the meaning and effect which the parties must be taken to have intended should be given to paragraph (1).

Chadwick LJ observed how entire agreement clauses and the effect of an acknowledgement of non-reliance had been considered by the Court of Appeal in Grimstead v McGarrigan (unreported 27 October 1999). The acknowledgement of non-reliance seeks to prevent the person to whom the representation was made, not from asserting a representation was made or was false, but from asserting that he relied on it. This is important in the context of commercial certainty and the parties being able to measure the risks and agree the price of the contract on the basis of the warranties given.

In the present case, the importance of the paragraph (3), is that the first sentence in paragraph (1) has to be construed on the basis that the parties intended that the whole agreement was to be contained or incorporated in the document which they signed and on the basis that neither party had relied on any pre-contractual representation when signing that document.  The first sentence of paragraph (1) therefore did not apply to negligent pre-contractual misrepresentations.

Judge Thornton made three errors which vitiated his conclusion.

The true scope and effect of the limit of liability in paragraph (1) was more limited than Judge Thornton seemed to recognise.  Paragraph (1) did not exclude or restrict liability for pre-contract misrepresentation, whether under statute or common law. 

Given the purpose and effect of the term in paragraph (1), the relevant questions were:

  1. was it fair and reasonable, having regard to the circumstances which were, or ought reasonably to have been in the contemplation of the parties when the contract was made, to include a term which sought to exclude contractual claims for indirect and consequential losses? and
  2. was it fair and reasonable, having regard to those circumstances, to include a term which sought to restrict loss directly and naturally resulting, in the ordinary course of things, from breach of warranty to the price paid for the equipment or software?

The obligation to use best endeavours to allocate appropriate resources to minimise any losses that might arise under the contract was confined to losses resulting from breach of warranty.  However, it plainly extended to all losses arising from breach of warranty, including indirect or consequential losses. If Sanderson did not use its best endeavours to allocate appropriate resources to ensure the software performs the specified functions, it could not rely on the provision excluding claims for indirect or consequential loss. If they had used best endeavours, then such losses would fall on Watford.

The judge’s view that this addendum was a “make-weight amendment” was not an apt description of its effect.

Watford had its own standard terms and conditions of sale with a limitation of liability clause similar to the first sentence of paragraph 1. The relevance of this was to show that Watford was well aware of the commercial considerations which led a supplier to include a provision restricting liability for indirect or consequential loss.  In particular, Watford was well aware that a supplier would be likely to determine the price at which it was prepared to sell its products by reference (among other things) to its exposure to the risk of unquantifiable claims for indirect or consequential losses which might be suffered by the customer if things went wrong. This had a direct bearing on reasonableness.

The Court considered reasonableness in relation to each limb of paragraph (1).

As for the term excluding indirect loss, it is important that the term did not exclude loss resulting from pre-contractual statements in relation to which a claim lies (if at all) in tort or under the Misrepresentation Act and that the addendum meant that it did not exclude indirect or consequential loss resulting from breach of warranty unless Sanderson has used its best endeavours to ensure compliance with the obligation to provide appropriate resources.

Looking at UCTA’s Schedule 2 guidelines, factors in favour of reasonableness were:

  • The parties were of equal bargaining strength
  • The inclusion of the term was likely to affect Sanderson’s pricing decision
  • Watford must be taken to appreciate that
  • Watford knew of the term, and must be taken to understand its effect
  • The product was to some extent modified to meet Watford’s precise needs

Other factors pointed in the opposite direction:

  • Mailbrain was the only package which appeared to meet Watford’s needs
  • Watford could not reasonably have expected to have been able to acquire a similar software package on better terms

When considering whether an exclusion of indirect loss was fair and reasonable, the following factors were relevant:

  • There is a significant risk that a non-standard software product that was modified to Watford’s precise requirements might not perform to the Watford’s satisfaction
  • There is a significant risk that the customer might not make the profits or savings which it hoped to make (and might incur consequential losses arising from the product not performing correctly)
  • Those risks were or must reasonably have been in the parties’ contemplation when contracting
  • Sanderson was in a better position to assess the risk that the product would fail to perform
  • Watford was in a better position to assess the amount of the potential loss if the product failed to perform
  • The risk of loss was likely to be capable of being covered by insurance, but at a cost
  • Both parties would have known, or ought reasonably to have known, when contracting that the party that would bear the risk of loss (or would bear the cost of insurance) was one of the factors influencing price

It was reasonable to expect that the contract would deal with the risk of where consequential or indirect loss should fall.  Where experienced businessmen representing substantial companies of equal bargaining power negotiate an agreement, they may be taken to have had regard to the matters known to them.  They should be taken to be the best judges of the commercial fairness of the agreement which they have made, including the fairness of each of the terms in that agreement and whether the terms are reasonable.  The court should not assume that either is likely to commit his company to an agreement which he thinks is unfair, or which he thinks includes unreasonable terms.  Unless satisfied that one party has, in effect, taken unfair advantage of the other – or that a term is so unreasonable that it cannot properly have been understood or considered – the court should not interfere.

The parties did negotiate as to price and Watford did obtain substantial concessions on price from Sanderson.  There was some negotiation about the liabilities, and although Watford did not get what it wanted, it did get something of value.  It would be impossible to hold that Sanderson took unfair advantage of Watford or that Watford did not understand and consider the effect of the term excluding indirect loss.

The term excluding indirect loss was therefore a fair and reasonable one.

As to the second sentence of paragraph (1), all it does is substitute a value equal to the price paid by the buyer for the goods for the “value which the goods would have had if they had fulfilled the warranty” under section 53(3) under the Sale of Goods Act 1979 (or the equivalent rule at common law).  It would be impossible to hold that this was an unfair or unreasonable substitution to make in a case like this.


This judgment at first instance was given on 27 July 2000 although the case did not seem to receive much attention.  It is the last of four cases that had sent shock waves through the computer industry, with suppliers questioning whether they could limit or exclude any of their liability without it falling foul of UCTA.  Given the importance of the case to the IT industry, an expanded commentary is provided here.

The process had started with South West Water v ICL, had been confirmed in Pegler v Wang, and was repeated in Horace Holman v Sherwood International.  The same has also been seen at first instance in Watford Electronics v Sanderson.  It appeared that the judges of the Technology and Construction Court seemed determined to use UCTA to undo limitations and exclusions agreed by parties in contracts for IT systems implementation or integration.  The result: four cases from the TCC, all decisively in favour of the user against the supplier.

The judgment of the Court of Appeal in Watford Electronics represents nothing less than a seismic shift in judicial attitudes and its lessons must be carefully learned and incorporated into drafting and sales/procurement processes.  First, it is necessary to put all these decisions in context.

Where does this judicial mistrust of limitations and exclusions derive from?  It is certainly nothing new: every law student learns how judges since the last century have applied a rule of construction contra proferentem so as to construe a clause of limitation or exclusion against the interests of a person relying on it.  However, the latest contest between a court of first instance and a higher court can perhaps be seen as the latest in a line of contests over the power of courts to review the substantive fairness of contracts – in this case by seeking to overturn the effect of limitation and exclusion clauses.

Some years ago, some judges realised that there was a natural limit to the contra proferentem rule of interpretation, since as draftsmen got better and wording became less ambiguous, it became harder and harder to find an interpretation that would help the claimant circumvent the clause in question. 

It may be recalled how the courts invented the doctrine of “fundamental breach” to get around this problem.  This doctrine was to the effect that some breaches were so serious, so fundamental, that the contract was destroyed by them, and with the contract went any clauses of limitation or exclusion.  Thus, after a fundamental breach, there could be no question of reliance on a limitation or exclusion clause.

There were many problems with this approach, for example, at what point did a breach become fundamental with the effects described above rather than just really serious?  Also, what if a party wanted to affirm the contract even after a breach that was indubitably fundamental?  According to the strict doctrine, there would be no contract left to affirm.

In the event, the doctrine was rejected by the House of Lords in Suisse Atlantique Société d’Armement Maritime S.A. v N.V. Rotterdamsche Koken Centrale [1967] 1 AC 361, a case which reaffirmed the traditional position that the court was only concerned with interpreting contracts, albeit using canons of construction such as contra proferentem.  However, this was not the end of the story, since the doctrine of fundamental breach was revived by Lord Denning MR in Harbutt’s “Platicine” Ltd. V Wayne Tank and Pump Co. Ltd [1970] 1 QB 447.

It was finally put to rest in Photo Production v Securcor Ltd [1980] AC 827, but the history of the doctrine and the way it was killed off are interesting for an examination and proper understanding of Watford Electronics.  In them, we can see a contest between judges of lower courts and those of the courts above them about how best to do “justice” in individual cases while paying due regard to legal principle.  It is this last matter which seems most to exercise the judges who hear the appeals – and quite rightly too.  The present state of the law is nothing short of scandalous, with lawyers quite unable to advise their clients as to what is the best way of limiting or excluding liabilities.  Lord Wilberforce put the issue well in his judgment in Photo Production at page 843:

"The doctrine of “fundamental breach” in spite of its imperfections and doubtful parentage has served a useful purpose.  There was a large number of problems, productive of injustice, in which it was worse than unsatisfactory to leave exception clauses to operate. …. But since then Parliament has taken a hand: it has passed the Unfair Contract Terms Act 1977.  This Act applies to consumer contracts and those based on standard terms and enables exception clauses to be applied with regard to what is just and reasonable.”

In other words, what we are witnessing is the courts over the course of several decades wrestling with how they can resolve a dispute and arrive at a just conclusion, given the existence of limitation and exclusion clauses.  At times, the courts have invented a doctrine such as fundamental breach, more recently, they have espoused the Unfair Contract Terms Act 1977 and used that to resolve the dilemma.  What remains unresolved is what, in any case, is the just result: only when that is resolved, can the means to achieve the “just” result be agreed on.

There is a sense in which we can see a number of different stages in this development:

  1. The early cases, where the courts simply used canons of construction (contra proferentem) to circumvent some of the unpleasant consequences of upholding limitation and exclusion clauses.
  2. The attempt by courts of first instance and the Court of Appeal to use the doctrine of fundamental breach as a rule of substantive law to circumvent limitation and exclusion clauses.
  3. The emergence of statutory consumer-protection and other “welfarist” statutes to protect consumers and others against the abuse of limitation and exclusion clauses.
  4. The post-statutory stage, where the courts have to work out the ambits of where they should or should not use the new powers given to them by statute.

The most important thing to realise coming out of the Court of Appeal’s judgment in Watford Electronics is therefore not the immediate impact of the result – there is no doubt but that users and suppliers will be analysing their contract terms and practices anyway – but that it should be seen in a wider context of overall hostility to limitations and exclusions by the courts who are working at the coal-face of trying to achieve just results.  This requires that we should try to assess what the weaknesses are in the Court of Appeal’s judgment, so that the possible lines of attack on it are properly understood, for, as has been seen from the above history, it may be safely anticipated that attempts will be made to distinguish it, or to limit it as closely as possible to its own facts.  This will be looked at below.

First, it may be as well to look at some of the principles that practitioners rarely consider when seeking to review laws on limitation and exclusion clauses.  What is a “just” result when applying such laws?  One’s first response, on hearing of a user deprived of the benefits of a system for which it had contracted, is to sympathise, to determine that proper compensation (by which is normally meant full compensation) should be made.  Is that “just”?  Maybe.  However, many in the IT industry feel an uneasiness about such a sweeping view of “justice”.  Take the facts in Watford Electronics itself: contracts for the supply of an integrated system worth just over £100,000 led to a contractual claim for £5.5 million.  The very grossness of the disparity must surely give rise to some unease.  Why should anyone in business undertake such serious risks against such a minimal return?

When one looks at the books, there is surprisingly little on the subject of limitation and exclusion clauses as such.  Indeed, the trend in modern European academic literature is rather to the effect of promoting judicial intervention in private contracts – the recent book by Hugh Collins “Regulating Contracts” (OUP 1999) being a good example (see especially chapter 11 “Unfair Contracts”).  Academic literature supporting principles derived from freedom of contract is to be found almost exclusively in writings from the USA.  Often it is said that overturning limitations and exclusions leads to increased insurance premiums, which leads to an inflationary rise generally: this may well be true, but there is no readily available empirical evidence to support this, at least in the context of IT disputes.

What support can be found for limitations and exclusions of liability?  There are probably the following arguments in favour of upholding limitation and exclusion clauses:


It is not only the parties to a contract or a dispute that want certainty – any practising lawyer knows already that clients want the lawyer to provide positive advice, not simply throw their hands in the air and pronounce that the court will do whatever is “reasonable” but no-one knows in advance what this is.  Fighting litigation in this way becomes an expensive business – the user arguing for unlimited liability, the supplier sticking to the contractual measure, the high stakes involved in the difference often ensuring that litigation does not reach a compromise.  Other interested parties also want certainty – insurers of claims, investors, a supplier’s other customers, all of whom have a vested interest in limiting the amount of any claim against the supplier.

2.Possible harm to other affected parties

Thus, in Pegler v Wang, the consequence of the judgment (albeit based on a construction of the relevant clauses contra proferentem) was that Wang went into voluntary liquidation, probably providing an “unjust” situation not only for Pegler, but also for Wang’s other customers for which it may have been providing a perfectly good service.

There is the harm done by the interference in the negotiation process – introducing the spectre of unlimited liability in fact radically alters the balance of power in settlement negotiations.  Of course, it is true that many settlements are in excess of the relevant contractual limitation (George Mitchell (Chesterhall) Ltd. v Finney Lock Seeds Ltd [1983] 2 AC 803 is an example of an industry where this happened) but the routine overturning of such limitations potentially introduces an element of imbalance in such negotiations.

3.Intervention is harmful to the parties

One of the results that was coming out of the use of UCTA by the Technology and Construction Court was that users were in some cases reluctant to negotiate limitation clauses.  The reason was simple: negotiation might take the case outside UCTA, whereas simply taking the clause as it stood left the user with the possibility of arguing UCTA applied and forcing the supplier to accept unlimited liability.  This radically alters the negotiating balance between the parties and ends up with a situation where it actually pays a user to be careless of its own welfare.

Armed with these thoughts, it is easier to understand critically the division in opinion between the judges of the Technology and Construction Court and those of the Court of Appeal.  “Justice” in the TCC, following the four cases referred to at the beginning of this comment, apparently meant giving the user an entitlement to full compensation, regardless of whether the user could or should have bargained for a different result in the contract, or in fact actually did so.  This is not the case according to the Court of Appeal.

As was said above, it is also necessary to be aware of the possible arguments that will be used to distinguish the Court of Appeal in Watford Electronics and some comments need to be made about this.  There are probably at least three ways in which a subsequent court will try distinguish Watford Electronics.

  1. The case was decided very much on its own facts and depended to a large extent on the fact that an addendum had been specifically negotiated which, as the Court of Appeal found, substantially altered the thrust of the limitations and exclusions.  If that clause had not been added, the result might have been different.
  2. The Court of Appeal accepted that the parties were on an equal footing, but it only takes a little imagination for a court in a different case to find that an IT supplier was the dominant party and could impose its “written standard terms of business” on the other.  This happened in Pegler v Wang where one of the facts found by the judge was to the effect that, as the parties had started work, this gave the supplier a dominant position in negotiations.
  3. If some special warning is given of “consequential” or “indirect” losses by the user pre-contract, it may be in future held to be unreasonable to exclude this type of loss.   A court might simply say that this type of loss was direct, not consequential, loss.

It is this last item which is probably going to cause most trouble.  The Court of Appeal’s reasoning is a little hard to detect.  Chadwick LJ equated the limitation of liability to the measure of damages provided by section 53(3) of the Sale of Goods Act 1979, which is the prima facie measure of damages in the event of delivery of goods not conforming with a warranty of quality.  This is, of course, correct but it is not the whole story.  It is only a prima facie rule and there are plenty of exceptions to it.  In particular, the cases distinguish between situations where goods are sold for resale (in which case the rule in section 53(3) is perfectly adequate) and those situations where goods are sold for use.  It is in this latter situation, goods sold for use, where damages go far beyond the rule in section 53(3).  A reference to just some of the cases will illustrate this.

In Parsons (Livestock) Ltd v Uttley Ingham & Co Ltd [1978] QB 791 the defendants sold a hopper for storing nuts for feeding pigs.  At the time of installation, the defendants forgot to open the top ventilator, with the result that the nuts went mouldy.  Some pigs fell ill, and the result of this was that more pigs died from a further disease.  The loss of the pigs was worth £10,000, and the plaintiffs also claimed for loss of profits.  The claim for the loss of the pigs was allowed, but the further claim was not allowed as being too remote.  A similar reasoning can be seen in Amstrad plc v Seagate Technology Inc (1997) 86 BLR 34, where the defendant supplied defective hard disk drives.  The plaintiff recovered damages for lost sales for lost and delayed sales, but not for lost sales of the plaintiff’s following range of computers, which were allegedly damaged by the low reputation of the plaintiff’s computers caused by the faulty hard disk drives in the previous range as this could not have been in the contemplation of the parties at the time of contracting.

However, it is well established that a claim can be made for loss of profits for the delivery and installation of a defective machine that was intended to be profit-making.  Cullinane v British Rema Manufacturing Co Ltd [1954] 1 QB 292 is a complex case, but it seems clear at least that it is possible to claim for the loss of profit resulting from the delivery of a machine (for clay pulverising) that did not comply with a contractual warranty.  Such a claim would far exceed the difference in value measure.

The result of these cases is that it is by no means clear that, in the case of goods sold for use, the prima facie rule contained in section 53(3) should apply.  Indeed, it might almost be said that such a rule would make no sense, since damages in these cases are calculated on a wholly different basis.

In the case of Watford Electronics, it seems that the loss of profits and increased costs of working could well be categorised as direct losses, but the Court of Appeal seems to assume that these losses would be “consequential”.  See, for example, paragraph 56 of the Court of Appeal’s judgment, where Chadwick LJ said,

“The parties negotiated, also, as to which of them should bear the risk (or the cost of insurance against the risk) of making the loss of profits, and other indirect or consequential loss, which Watford might suffer if the product failed to perform as intended.”[emphasis added]

It is curious to find that loss of profits is so equated to indirect or consequential loss.  Following three cases recently (British Sugar Plc v. NEI Power Projects Ltd (1998) 14 Const. L.J. 365;  Chiemgauer Membran und Zeltbrau GmbH v New Millennium Experience Co. Ltd, The Times 16 January 2001; and Hilton International Hotels UK Ltd v Hotel Services Ltd (2000) BLR 235) it is quite clear law that the expression “consequential or indirect loss” refers to the second limb of Hadley v Baxendale, and does not include every instance of a claim for loss of profits (which might well be a claim for direct loss under the first limb of Hadley v Baxendale).  Chadwick LJ also defines consequential loss as being those losses within the second limb of Hadley v Baxendale (without citing any of the three authorities just referred to), but there seems to be an assumption that the majority of Watford’s claim, being for such items as loss of profits, is to be regarded as consequential loss.

The transcripts do not give much information about the nature of the losses claimed other than their bare descriptions.  However, it is by no means clear that Watford’s claim for loss of profits, in circumstances where Watford informed Sanderson prior to contract of the problems it might face, necessarily falls into the second limb of Hadley v Baxendale rather than the first.  Indeed, it could be strongly argued that the loss of profits and the increased costs of working were within the first limb of Hadley v Baxendale (direct loss).

If Watford’s claim for loss of profits was treated as direct loss, there is the outcome that liability would be limited to the contract price (rather over £100,000), against loss of profits claimed of £4.5 million.  It would surely become harder in that case to argue that the limit was reasonable.

There are a number of questions that every IT supplier will want to ask and want answered definitively.  For example, what is that status of complete exclusions of liability for loss of profits or failure to make anticipated savings?  Watford Electronics in the Court of Appeal gives no definitive answer to this question.  Nor is it authority for saying that every limit of liability pegged to the contract price is from now on valid.  It is good authority for arguing before any court of first instance that UCTA has no application to a situation involving parties of equal bargaining power – and this at least must be regarded as progress on the previous situation.

Thus, in practical terms, where does this leave IT suppliers and procurers of IT systems?  Regrettably, it will probably take yet more litigation (including appeals) before we can tell if the courts generally will adopt the freedom of contract approach or the interventionist approach.  This is indeed regrettable in every sense: it is easy to see that legal fees (and all the other costs associated with litigation) will be wasted trying to prove the point one way or another.  This is one area where certainty could actually save costs in the long run – from either the supplier’s or the user’s side.  Unfortunately, as was seen in the case of the doctrine of fundamental breach, it took more than one outing in the House of Lords before that particular bogus legal principle was laid to rest.

I can do no better than end with the words of Lord Wilberforce in Photo Productions:

“At the judicial stage there still more to be said for leaving cases to be decided straightforwardly on what the parties have bargained for rather than upon analysis, which becomes progressively more refined, of decisions in other cases leading to inevitable appeals.”  [1980] AC at 827 at 843G

From many perspectives, let us hope this view prevails.

We are processing your request. \n Thank you for your patience. An error occurred. This could be due to inactivity on the page - please try again.