I have linked the two topics of global claims and concurrency for two reasons. First, because the linkage was made in City Inn [2008] BLR 269 and [2010] BLR 473 between apportionment of losses in certain circumstances in global claims, in accordance with the decision in John Doyle Construction [2004] BLR 295, and entitlement to prolongation costs; secondly, because, in practice, questions of concurrency and global claims frequently arise from the same claim.

Two matters need emphasis: First, as ever, the express words of the contract must be examined with care. They may, but usually do not, provide an answer to the problems considered below. Secondly, the issue raised here is one of causation.

It is hardly surprising that it gives rise to difficulties, given that the Courts have consistently struggled with this concept. Thus the general view is that ultimately a decision on whether X caused Y depends on the court's common sense interpretation of the facts.

Part of the unspoken background to both global claims and questions of concurrency is the question of fairness. It is hard to avoid the general conclusion that if losses or delay have in part been caused by the employer's default, then to that extent the contractor should be recompensed or relieved from payment of liquidated damages, even if one cannot quite put one's finger on the precise extent to which the delay or losses were caused by that default.

Such an approach might be said to be fair and allows the decision-maker to weigh matters in the balance and arrive at a just result. If such an approach were followed, it might well reduce the time and cost of litigation, since critical-path analysis would be of less relevance and long, tedious explanations of what occurred on site would be avoided.

Of course, such an approach would make it difficult for an adviser to predict the outcome, since much would be left to the decisionmaker's discretion, but this is no different from an apportionment in tort or delict where loss is caused in part by the defendant and in part by the claimant.

In such cases, the court can reduce the claimant's recovery to the extent that it thinks it 'just and equitable' to do so as under the Law Reform (Contributory Negligence) Act 1945. Just as adjudication may provide rough justice, such an approach to global claims and concurrency may be welcome and appropriate, if it provides speedy and relatively inexpensive justice.

One can therefore understand the attraction of the City Inn solution, which enables the architect to grant an extension of time as he 'estimates to be fair and reasonable', which entitles him to apportion periods of concurrent delay as between the contractor and employer.

The law has been quite willing to adapt the strict requirements for proof of causation to particular circumstances. Thus a claimant can recover against either or both of two contract breakers, provided only that each of them was an effective cause of the loss: Heskell v Continental Express Ltd [1950] 1 All ER 1033. But for such a rule, both the contract breakers could blame the other as being the 'true' cause of the loss, thereby depriving the claimant of any remedy. Indeed, in Scots law, it is arguable that this exception is more extensive. Where two separate contract breakers cause 'one common result', they are jointly and severally liable to the pursuer: Grunwald v Hughes and Others (1965) SLT 209.

This is no different from the law in England. However, in the Scottish case of McGillivray v Davidson (1993) SLT 693, it appears to have been held that two such contract breakers could be sued on the basis of joint and several liability even though they did not both cause the whole damage claimed. The judge held that:

"The essence of joint and several liability is that each defender should be liable for the whole damage, but it does not follow from that that each defender's breach must have been a material cause of the whole damage."

In the context of tort, the rules as to causation appear to be even more flexible. Lord Hoffmann said in Kuwait Airways Corp v Iraq Airways Co. [2002] UKHL 19:

"There is therefore no uniform causal requirement for liability in tort. Instead, there are varying causal requirements, depending upon the basis and purpose of liability. One cannot separate questions of liability from questions of causation. They are inextricably connected.

One is never simply liable; one is always liable for something and the rules which determine what one is liable for are as much part of the substantive law as the rules which determine which acts give rise to liability. It is often said that causation is a question of fact. So it is, but so is the question of liability.

Liability involves applying the rules which determine whether an act is tortious to the facts of the case. Likewise, the question of causation is decided by applying the rules which lay down the causal requirements for that form of liability to the facts of the case."

Thus the present rules of causation as applied to global claims or questions of concurrency are not immutable; the law can adapt those rules (and has) to the needs of society.

However desirable such a shift in thinking might be, the present law is reasonably clear. In general terms, in a claim under or for breach of contract, the claimant must prove, on a balance of probabilities, that 'but for' the defendant's wrongful act, the relevant damage would not have occurred. In other words, if the damage would have occurred in any event, the defendant's wrongful act is not a 'but for' cause. It is important to be clear as to how demanding this test is.

It means that if the defendant can show that the claimant's loss resulted from a cause for which the claimant is responsible, even if that cause is less than the dominant cause (but is nevertheless of significance) the claimant's case will fail.

Put simply, the claimant must prove that the defendant's breach was the sole effective cause of the loss, and any competing cause which is more than trivial is likely to defeat the 'but for' test. Expressed in this way, it is immediately apparent why global claims and concurrent causes give rise to such problems.

Global Claims

Cases on global claims can take two forms. In the more extreme form, a claim is made for the total cost, less the tendered cost, and that additional cost is said to be attributable to a number of causes. This is sometimes referred to as a 'total' cost claim. In the less extreme form, a cost is identified, for prolongation or disruption, and a number of causes are said to have been responsible for the identified cost.

What they have in common is usually an assertion that there is a complex inter-reaction of causes which, it is asserted, makes it difficult or impossible to identify the precise loss resulting from each cause but that, in combination (the precise combination being extremely difficult or impossible to specify) the causes relied upon gave rise to the loss claimed. Thus, conceptually, this departs from the 'but for' test, since it is not possible to say that 'but for' cause A the same loss would not have resulted from causes B to X i.e. the 'but for' test is not satisfied as regards any one of the causes.

Such a claim will, however, succeed, provided two criteria are satisfied. First, the claimant must demonstrate that 'but for' causes A to X, the loss in question would not have been suffered. Secondly, each of causes A to X must be matters for which the employer bears financial responsibility.

In the latest case, Walter Lilly [2012] BLR 503, the claim was not a total cost or global claim but there are obiter dicta of relevance:

"There is nothing in principle 'wrong' with a 'total' or 'global' cost claim. However, there are added evidential difficulties (in many but not necessarily all cases) which a claimant contractor has to overcome. It will generally have to establish (on a balance of probabilities) that the loss which it has incurred (namely the difference between what it has cost the contractor and what it has been paid) would not have been incurred in any event. Thus, it will need to demonstrate that its accepted tender was sufficiently well priced that it would have made some net return.

It will need to demonstrate in effect that there are no other matters which actually occurred (other than those relied upon in its pleaded case and which it has proved are likely to have caused the loss)."

Although using different language, this is an entirely orthodox analysis. But for the events relied upon, would the loss have been incurred in any event? Note the use of the phrase: "...other matters which actually occurred (other than those relied upon in its pleaded case and which it has proved are likely to have caused the loss)". Thus if it is 'likely' that other matters would have caused the loss, the claim will fail. Given the context, I do not think 'likely' simply means that on a balance of probabilities other matters caused the loss; it must refer to a significant cause or, more obliquely, a not insignificant cause of the loss as being something else.

This is reinforced by the examples given by the judge. Thus he assumes a case where there is a global claim for £1m and, 'but for' one underpriced item in the preliminaries of £50,000, the contractor would have made a net return.

In that case, "the consequence would simply be that the global loss is reduced by £50,000 because the claimant contractor has not been able to prove that £50,000 of the global loss would not have been incurred in any event".

Further, the judge gives the example of "there being events during the course of the contract which are the fault or risk of the claimant contractor which caused or cannot be demonstrated not to cause some loss, the overall claim will not be rejected save to theextent that those events caused some loss...assuming that this time can be quantified either precisely or at least by way of assessment, that amount would be deducted from the global loss".

This again is adopting the language of the 'but for' test, although two points are to be noted.

First, both in this passage and elsewhere, the judge makes the point that it is for the contractor to prove that 'but for' the events relied upon it would not have incurred the costs claimed, it is not for the employer to prove this; in other words the burden is on the contractor.

Secondly, the use of the word 'assessment', in the context of quantifying the effect of an event that does not entitle the contractor to recovery, might suggest some degree of latitude.

Once again, given the context, I do not think this is what the judge intended to indicate. There are many circumstances in which damages are 'assessed' (see, for example, Chaplin v Hicks [1911] 2 KB 786); it will always be a matter of fact and degree as to the extent to which a contractor must be able to identify and prove the sum to be disallowed, in order to recover the balance by way of a global claim.

Thus Walter Lilly adopted an orthodox analysis. The same cannot be said of John Doyle, which contains two points of importance for these purposes. First, it was said:

"In this connection, it is frequently possible to say that an item of loss has been caused by a particular event notwithstanding that other events played a part in its occurrence. In such cases, if an event or events for which the employer is responsible can be described as the dominant cause of an item of loss, that will be sufficient to establish liability, notwithstanding the existence of other causes that are to some degree at least concurrent."

Clearly, this is not an application of the 'but for' test but some lesser requirement. Thus loss will be recoverable under this rubric if the primary or major cause of the loss is the employer's responsibility, even if the loss was also caused 'to some degree' by another matter.

The Inner House justified this approach on the basis that it was similar to the test adopted in the Leyland Shipping [1918] AC 350 case.

This may be true, but the issue in Leyland was whether the loss was covered by the terms of an insurance policy or by an intervening event.

In insurance law, it is well established that it must be shown that an insured event was the 'proximate cause' of the loss in order to found recovery. It is doubtful whether this line of authority can be said to be applicable, even by analogy, to global claims.

However, John Doyle can be said to be truly heretical in contemplating apportionment in global claims:

"... it may be possible to apportion the loss between the causes for which the employer is responsible and other causes. In such a case it is obviously necessary that the event or events for which the employer is responsible should be a material cause of the loss.

Provided that condition is met, however, we are of the opinion that apportionment of loss between the different causes is possible in an appropriate case.

Such a procedure may be appropriate in a case where the causes of the loss are truly concurrent, in the sense that both operate together at the same time to produce a single consequence."

No English or Scottish authority was cited for such an approach. Indeed, it is reasonably clear that the Inner House thought, for policy reasons, that this methodology might be acceptable, since the alternative would be to deny the contractor any recovery even if the conduct of the employer or architect was plainly culpable.

Thus John Doyle contemplates downgrading the evidential burden in global claims from the 'but for' test, to the 'dominant cause' test, to, ultimately, apportionment. It is unlikely that apportionment of losses in the case of a global claim would be adopted in an English court. First, the court in Walter Lilly rejected the City Inn solution of apportionment in extension of time claims (considered further below). And secondly in De Beer [2011] BLR 274 it was said, in the context of a discussion about concurrent delay,:

"...the contractor cannot recover damages for delay in circumstances where he would have suffered exactly the same loss as a result of causes within his control or for which he is contractually responsible."

Concurrent Delay

True concurrent delay is often defined as a period of project overrun caused by two or more effective causes of delay that are of approximately equal causative potency (see John Marrin QC, Society of Construction Law Paper No. 179, 2013).

It is therefore said that such a situation is likely to arise rarely. Defined in this way, this is probably correct. But experience suggests that concurrency problems arise with surprising frequency.

This is because it is not easy to identify the converse of an effective cause of delay and, secondly, because there are often overlapping concurrent events where it may be difficult to identify the precise period of overlap.

For example, bad weather may cause a 6-week delay, some of which was also caused by a variation. There is clearly some element of concurrency in this example but it may be difficult to identify the extent.

Of course, concurrency can only arise if the events in question cause the same (or possibly an overlapping) period of delay. It is this concurrency that the debate is, or should be, concerned with.

Thus if an event occurring in January 2012 causes a 5-week delay to completion, an event in September 2012 causing delay cannot cause delay concurrent with that caused by the event in January 2012.

The simplest way of viewing this is that the programme is updated by extending the completion date (assuming this is appropriate) by five weeks to take account of the January 2012 event.

As regards the event in September 2012, the question is whether it caused further delay, i.e. did it cause completion to be delayed by more than the 5 weeks delay caused by the event in January 2012?

It is, therefore, somewhat ironic that the case of City Inn was not a case of concurrent delay at all (or at least not in part). In the Outer House, most causes of delay that caused completion to be delayed were treated as being concurrent; in other words, each cause of delay was not sequentially impacted on completion so as to update the programme for each such cause of delay. In the Inner House, the discussion of this point is somewhat unclear, both as to how the members of the court interpreted the opinion of Lord Drummond Young and in expressing their own opinion.

Nevertheless, whether or not there was such an error in City Inn, it is clear that the events in question must cause the same period of delay, or at least there must be some element of overlap for questions of concurrency to be relevant.

Unlike the global claim analysis considered above, it is clear that the justification for apportionment in City Inn arose from Lord Drummond Young's interpretation of the words 'fair and reasonable' in clause 25.3.1 of the JCT Standard Form.

As set out above, the general principles of causation must give way to the express terms of the contract and there is no reason in principle why an architect should not be given the power to make an apportionment in the event of concurrent delay caused by relevant events and events not entitling the contractor to an extension of time. The view in Walter Lilly is that those words do not have that effect.

Indeed, it is now broadly accepted in England that where a relevant event and an event for which the contractor is responsible causes concurrent delay, the contractor is entitled to an extension of time (Steria v Sigma [2008] BLR 79, De Beer and Walter Lilly).

Once again, the justification for this is as a matter of construction of the contract in question. In this case, it is said that where the parties have expressly provided for an extension of time for certain events, the parties must be taken to have contemplated that there could be more than one effective cause of delay, one of which would not qualify for an extension of time, but nevertheless by their express words agreed that a contractor is entitled to an extension of time for an effective cause of delay.

A consideration in favour of the English approach is the impact of the doctrine of prevention. This holds that if the employer, by his acts or omissions, prevents the contractor completing by the due date, he is not entitled to claim liquidated damages in respect of any delay caused by those acts or omissions.

If one of the relevant events would otherwise be an act of prevention on the part of the employer and if such an event is a cause of delay, but the contractor does not obtain a full extension of time for such delay, on the basis that the delay is concurrent with delay which is at the contractor's risk, then it offends against the prevention principle.

Whilst this English approach to the construction of extension of time avoids this problem, it is evident that if a contractor only obtains a proportion of the extension of time caused by such a delay, as per the City Inn approach, there is a clash with the prevention principle.

What happens if the act of prevention is not a relevant event but it causes delay concurrently with a matter for which the contractor is responsible?

This cannot be resolved as a matter of construction, at least under the JCT Standard Form, and must be dealt with as a matter of principle. In Jerram Falkus v Fenice [2011] EWHC 1935, the court held:

"... that, for the prevention principle to apply, the contractor must be able to demonstrate that the employer's acts or omissions have prevented the contractor from achieving an earlier completion date and that, if that earlier completion date would not have been achieved anyway, because of concurrent delays caused by the contractor's own default, the prevention principle will not apply."

In other words, the prevention principle only applies if the employer's acts or omissions are the sole (or possibly, dominant) cause of delay but not if it is one of a number of concurrent causes of delay.

This, it is submitted, is inconsistent with the prevention principle and the proposition that a contractor is entitled to an extension of time (or a reasonable time within which to complete) where an act of prevention is a cause of delay, concurrent with a delay caused by the contractor's default.


Absent an express contractual provision, a claimant must show that, 'but for' the event in question, he would not have suffered the loss in question. Claims under or for breach of contracts concerning construction or engineering projects have no separate set of rules. Thus any claim for loss, whether it is a global or total cost claim, must satisfy the 'but for' causation test.

In this respect, Walter Lilly followed wellknown and accepted principles.

The same 'but for' test applies in claims for extension of time but here it is frequently modified by the terms of the contract or the application of the prevention principle, so that the relevant event or the act of prevention trumps another concurrent cause of delay.

Apportionment, either in relation to global claims or extension of time claims, as outlined in John Doyle and City Inn, does not follow established authority and is unlikely to be followed south of the border.

Where are we going?

The history of the development of the law is littered with cases from the upper courts that destroyed the then prevailing orthodoxy. Until Gilbert Ash [1974] AC 698, it was thought that an architect's certificate was like a letter of credit and no set off could be asserted against a certified sum.

Until Beaufort Developments [1999] 1 AC 266, it was thought that only arbitrators could open up, review and revise certificates. In other words, it only takes one case to alter fundamentally our perception of the law. Global claims and concurrency would benefit from thorough analysis in the highest courts.

I do not want to appear to criticise the Scottish courts. They have, seemingly deliberately, departed from the orthodoxy. Whilst there are attractions to permitting tribunals to arrive at a 'just' result, whether it be by apportionment or some other discretionary tool, in the end the 'but for' test, as modified by the express terms of the contract, and the prevention principle, can probably provide a reasoned, just and predictable result in most cases, without introducing apportionment.

The articles and papers published by Keating Chambers are for the purpose of raising general awareness of issues and stimulating discussion. The contents must not be relied upon or applied in any given situation. There is no substitute for taking appropriate professional advice.