Part 5 – Why expert witnesses disagree: understanding temporal divergence in arbitration

Introduction

Disagreement between expert witnesses is a familiar and accepted feature of international arbitration.

Indeed, much of the arbitration process surrounding expert witness evidence is designed around the expectation that disagreement will occur. Expert witness reports are exchanged. Joint statements are prepared. Expert witness conferences are convened. Cross-examination is conducted. Tribunals are then invited to determine which expert witness analysis is ultimately more persuasive.

The implicit assumption is straightforward. The expert witnesses are understood to be analysing the same dispute^[1]Parties’ liability positions. See Part 3 article on ‘figures-as-figures’ from the same evidential position and reaching different conclusions through the application of professional judgement.

Often that assumption is justified. Sometimes it is not.

In practice, expert witnesses rarely operate within a static evidential landscape during arbitration proceedings. Documents emerge. Witness evidence develops. Pleaded cases evolve. Legal positions are refined. Assumptions that appeared reasonable at one stage of the proceedings may cease to be sustainable at another.

The dispute itself moves.

When that occurs, disagreement between expert witnesses may not arise because they have analysed the same issue differently. It may arise because they are analysing different liability/evidential versions of the dispute.

That distinction matters.

What appears to be disagreement between witnesses may actually be disagreement between procedural moments (that result in different liability/evidence positions).

A tribunal naturally focuses on the reasoning that produces competing expert witness opinions. That is understandable. expert witness evidence is, after all, intended to assist the tribunal in evaluating technical, quantum, delay or other specialist issues. However, before asking which expert witness’s reasoning is more persuasive, there is sometimes a prior question.

Are the expert witnesses analysing the same version of the dispute?

If they are, the tribunal can proceed to evaluate the competing reasoning on its merits. If they are not, the task is different. The tribunal must first understand whether the disagreement is truly analytical, or whether it is the product of procedural evolution.

This analysis builds upon the “figures-as-figures” assessment considered in Part 3. Before competing figures can be compared numerically, the tribunal must also establish whether they have been produced from a sufficiently common and synchronised procedural platform.

This article examines that problem. It introduces the concept of the moving platform and proposes temporal divergence as a distinct category of expert witness disagreement. It then proposes a tribunal synchronisation question: before preferring one expert witness opinion over another, tribunals should consider whether the expert witness evidence has been procedurally synchronised.

Because if the platform has moved, expert witness comparison should not occur as though it has remained still.

The authors of this paper are both seasoned quantum expert witnesses, with over 50 years of combined experience in the international construction industry. Between them, they have been appointed as lead quantum expert witnesses on more than 100 matters. The views set out herein are grounded in their collective professional experience and a desire to enhance the quality and efficiency of expert witness evidence in complex construction disputes.

 

The assumption that expert witnesses are addressing the same dispute

The comparison of competing expert witness evidence rests upon an assumption that is rarely stated explicitly.

It is assumed that both expert witnesses are answering the same question.

That assumption makes comparison possible. If two expert witnesses are analysing the same dispute, relying upon the same evidential record and operating within the same analytical framework, disagreement can reasonably be understood as a difference in professional judgement. The reality is often more complicated.

Expert witnesses are instructed independently. They receive information through different legal teams. They may work with different assumptions. They may be asked to address different formulations of the issues in dispute. Even where the broad subject matter appears similar, the foundations upon which each analysis is constructed are not necessarily identical.

Differences in data, assumptions, methodology and instructions are all familiar features of expert witness evidence. They are frequently treated as explanations for disagreement.

That is useful, but incomplete. There is a further complication. The foundations themselves may not remain fixed throughout the proceedings. The platform upon which an expert witness opinion is constructed may move.

 

The moving platform in expert witness evidence

Arbitration is not a static process. It is an evolving process.

The evidential landscape that exists when an expert witness prepares an initial report rarely remains unchanged throughout the life of an arbitration. Additional disclosure may occur. Witness evidence may alter the understanding of key events. Legal positions may be narrowed, expanded or abandoned. Assumptions may become more or less sustainable as the case develops.

Sometimes the change is incremental. Sometimes it is fundamental. In either case, the effect can be significant.

An expert witness opinion that was carefully reasoned and entirely appropriate when prepared may no longer sit upon the same analytical platform several months later.

That does not necessarily mean the opinion was wrong. It may simply mean that it addressed a different version of the dispute. This is where much misunderstanding in the comparison and analysis of expert witness evidence can arise. Arbitration often treats expert witness disagreement as evidence of analytical divergence. Two expert witnesses have reached different conclusions, and the tribunal is asked to decide which conclusion should be preferred.

In many cases (i.e. those that are done right), that is precisely the correct approach. But not always.

One expert witness may be analysing the dispute as it presently stands. Another may, through no fault of their own, be analysing the dispute as it stood when an earlier expert witness report was prepared. One may have updated their analysis following disclosure. Another may not yet have had the opportunity to do so. One may be working from a pleaded case that has since narrowed. Another may still be responding to the broader case as originally advanced.

Viewed in that way, the disagreement is not only analytical. It is temporal.

The expert witnesses are not necessarily disagreeing about the same dispute. They may be analysing different procedural versions of it. That is the moving platform problem.

 

How expert witness evidence changes during arbitration

The platform may move in several familiar ways.

New documents may be disclosed after an expert witness has formed an opinion. Those documents may not invalidate the earlier analysis, but they may alter the evidential foundation upon which any current assessment should proceed.

Witness evidence may change the factual landscape. A sequence of events may be clarified. A decision-making process may be explained. A previously assumed fact may become disputed, or a disputed fact may become less controversial.

Legal positions may also evolve. Claims may be refined. Defences may be narrowed. Alternative cases may emerge. Assumptions that were once central may cease to matter. Issues that appeared peripheral may become decisive.

None of this is unusual. It is how arbitration often develops.

The difficulty is that expert witness evidence does not always move at the same speed as the dispute. One expert witness may update promptly. Another may update later. Another may remain tied to an earlier reporting stage until a further round of evidence is permitted.

The resulting disagreement may appear substantive. It may be presented as a clash of professional judgement. It may even become the focus of cross-examination.

Yet the disagreement may have a different source. It may arise because the expert witnesses are standing on different versions of the platform.

 

What is temporal divergence in expert witness evidence?

Not all expert witness disagreement is created in the same way.

There is an important distinction between analytical divergence, temporal divergence and hybrid divergence.

Analytical divergence arises where expert witnesses operate from a substantially common platform but apply different professional judgement to the same underlying material.

This is the form of disagreement most commonly contemplated by tribunals. The expert witnesses are addressing the same issue. They have access to substantially the same material. They are operating from comparable assumptions. This is the setting in which a true “figures-as-figures” comparison becomes possible: the tribunal is evaluating competing numerical conclusions produced from a substantially common platform.

That is true expert witness disagreement. It is the type of disagreement tribunals are well placed to evaluate.

Temporal divergence is different.

Temporal divergence arises where the expert witnesses are effectively analysing different procedural versions of the dispute because the analytical platform has evolved over time.

In those circumstances, each expert witness may have produced a coherent and professionally defensible opinion.^[2]Put simply, this is a different ‘figures-as-figures’ assessment Both analyses may be reasonable. Both may be internally consistent. Both may reflect careful expert witness judgement. The difficulty is that the opinions were not generated from the same procedural moment. What appears to be disagreement between expert witnesses may actually be a difference between procedural moments.

One opinion may belong to an earlier version of the dispute. The other may belong to a later one.

The tribunal is then not simply comparing two answers to the same question. It may be comparing answers to different versions of the question.

That is a different exercise.

Hybrid divergence is more common still.

In many arbitrations, disagreement is partly analytical and partly temporal. The expert witnesses may genuinely disagree on methodology, interpretation or judgement. At the same time, part of the gap between them may be explained by differences in the evidential record, the pleaded case, the assumptions adopted, or the stage at which their respective analyses were prepared.

This matters because tribunals may otherwise attach excessive significance to apparent expert witness disagreement.

An issue may appear to involve a fundamental clash of expert witness opinion when, in reality, part of the disagreement disappears once the platform is aligned. Conversely, a disagreement that survives after synchronisation is likely to be more meaningful, because it is less likely to be an artefact of timing.

There is therefore a relationship between the type of divergence and the weight that may properly be attached to it.

Analytical divergence will ordinarily carry greater evaluative significance because it survives synchronisation. If the expert witnesses are working from the same evidential record, addressing the same pleaded case, applying comparable assumptions and analysing the dispute from the same procedural moment, then a remaining disagreement is more likely to reflect a true difference of expert witness judgement.

That does not mean the disagreement is necessarily correct, persuasive or decisive. It means only that it is the kind of disagreement the tribunal can more confidently evaluate on its merits.

Temporal divergence should be approached differently.

If disagreement arises because one opinion relates to an earlier procedural version of the dispute and another relates to a later one, the disagreement may be real, but its significance is different. It may not justify the same evaluative weight because it may reduce, disappear or alter materially once the platform is aligned.

Hybrid divergence requires the most care.

Where disagreement is partly analytical and partly temporal, the tribunal should try to separate the two. The analytical component may require determination. The temporal component may require synchronisation. Treating both as a single undifferentiated disagreement risks giving too much weight to a divergence that may be partly procedural rather than substantive.

This is why the distinction matters.

It is not simply a classification exercise. It is a weighting exercise.

Before deciding how much weight to attach to expert witness disagreement, the tribunal should consider what kind of disagreement it is.

The distinction is therefore not academic. It affects:

• weight;
• cross-examination;
• the usefulness of joint statements; and
• the tribunal’s confidence in comparing expert witness conclusions.

In short, tribunals should not treat all disagreement as equal. Before deciding which disagreement matters, they should consider what kind of disagreement they are dealing with.

 

How tribunals should assess competing expert witness opinions

The tribunal is not merely comparing competing expert witness opinions. It is comparing opinions generated at particular points in the procedural lifecycle. That is often overlooked.

Expert witness reports appear in the record as documents. They can therefore seem fixed, complete and comparable. The fact that each report was generated at a particular stage of evidential and procedural development may receive less attention.

But timing matters.

An expert witness opinion is not produced in the abstract. It is produced by reference to the evidence available, the assumptions adopted, the instructions given and the issues then understood to require analysis. If those matters later change, the opinion may remain useful, but its procedural location must be understood.

The tribunal synchronisation problem arises where expert witness opinions generated at different stages of the dispute are treated as though they were generated from the same platform.

That can lead to unfair or unreliable comparison.

It may also distort the assessment of expert witness credibility. An expert witness who has not updated an opinion may appear inflexible. Another who has updated may appear inconsistent. Neither characterisation is necessarily correct. The real question may be whether the procedural platform has changed and whether the expert witness evidence has been synchronised with that change.

The tribunal should therefore ask a synchronisation question before moving to preference.

Before preferring Expert A over Expert B, the tribunal should consider:

• Are the expert witnesses working from the same evidential record?
• Are they addressing the same pleaded case?
• Are they operating from the same assumptions?
• Have both analyses been updated to reflect material procedural developments?
• If not, has the disagreement been caused, wholly or partly, by a lack of synchronisation?

This is not a mechanical checklist. Nor is it an invitation to overcomplicate the expert witness process. It is a discipline.

It requires the tribunal to pause before treating disagreement as substantive. It encourages the tribunal to distinguish between disagreement that reflects professional judgement and disagreement that reflects procedural misalignment.

The point is not that expert witness comparison must wait until every possible evidential movement has ceased. That would be unrealistic. Arbitration cannot function if expert witness evidence is constantly deferred in pursuit of perfect stability.

The point is more modest.

Expert comparison should, so far as reasonably practicable, occur only after material procedural differences have been identified and the expert witness analyses sufficiently synchronised.

The synchronisation question is not solely an expert witness question. It is ultimately a case management question.

Where the tribunal identifies a material risk that competing expert witness analyses have been generated from different procedural versions of the dispute, it may be appropriate to consider whether further clarification, updated analysis, targeted expert witness engagement or revised evidential directions are required before meaningful comparison (and evaluation) can occur.

This does not require tribunals to become advocates, nor does it require continuous intervention in the expert witness process. It simply recognises that meaningful comparison depends upon understanding whether the expert witness evidence is being assessed from a sufficiently aligned procedural platform.

That is the normative proposition.

Where the platform has moved materially, tribunals should identify whether the expert witness evidence has moved with it. If it has not, the tribunal should be careful before treating the resulting disagreement as a true contest of expert witness judgement.

 

Platform stabilty

This is where the concept of platform stability becomes useful. Platform stability is not concerned with whether expert witnesses agree. Nor is it concerned with which expert witness is ultimately correct.

It is concerned with whether the foundations upon which competing opinions are constructed are sufficiently stable, current and comparable to allow meaningful comparison.

Those foundations typically include:

• the question being addressed;
• the evidential record available;
• the assumptions adopted;
• the instructions provided;
• the pleaded case being analysed;
• the methodology applied; and
• the procedural moment at which the opinion was formed.

Methodology often becomes the most visible focus of debate. It is easy to identify. It is relatively easy to challenge. It provides a neat structure for cross-examination and tribunal questioning. However, methodology sits at the top of a larger analytical structure.

Where the underlying platform is unstable, methodological disagreement may explain far less than is commonly assumed. The more significant issue may be that the expert witnesses are not standing on the same ground or not standing on the same version of the ground.

Platform stability therefore provides the framework through which tribunals can determine whether apparent disagreement is analytical, temporal, or hybrid. That classification matters because it affects weight.

If the platform is stable and synchronised, disagreement is more likely to be meaningful. It is more likely to represent a true difference of expert witness judgement and can be evaluated accordingly.

If the platform is unstable or unsynchronised, disagreement must be interpreted with caution. It may be less a contest between competing expert witness opinions and more a consequence of comparing opinions generated from different procedural versions of the dispute.

The tribunal’s task is therefore not merely to identify disagreement. It is to understand what kind of disagreement it is, and how much weight that disagreement should properly bear.

 

Reframing the tribunal’s task

The practical implications are significant. Tribunals frequently ask which expert witness is correct. In many cases, that is entirely the right question. Where the possibility of a moving platform exists, however, another question should come first.

Have the expert witness opinions been procedurally synchronised?

That question reframes the tribunal’s task.

The tribunal is not being asked to avoid deciding between expert witnesses. Nor is it being asked to indulge endless refinement of expert witness evidence. It is being asked to ensure that the comparison it undertakes is a fair and meaningful one.

If the expert witnesses are operating from stable, current and comparable platforms, differences between them are more likely to reflect genuine differences of professional judgement. Those are the disagreements tribunals are well placed to determine.

If the platforms have diverged, the tribunal must first understand the source and significance of that divergence.

• Has new evidence emerged?
• Have assumptions changed?
• Has the pleaded case evolved?
• Has one expert witness updated while another has not?
• Has the disagreement been created by timing rather than reasoning?

Only once those questions have been addressed can the tribunal assess the true significance of the competing conclusions.

That assessment is not only relevant to admissibility or process. It is relevant to weight.

A disagreement that survives synchronisation may deserve substantial attention because it reflects a disagreement that remains after the platform has been aligned. A disagreement that exists only because the platform has moved may deserve less weight, or a different form of procedural response, because it may not reflect a true contest of expert witness judgement.

The tribunal should therefore be cautious before treating unsynchronised disagreement as decisive. Otherwise, there is a risk that the tribunal is not choosing between competing analyses of the same dispute.

It is choosing between analyses of different procedural versions of the dispute. That is not an ideal basis on which to decide anything. Though arbitration does occasionally expect people to decide highly consequential matters on precisely that sort of basis.

 

Practical implications for expert witness evidence and tribunals

The moving platform problem has consequences for the design of the expert witness process.

First, early engagement between expert witnesses becomes more valuable because it allows instability to be identified before it becomes entrenched. Early engagement will not eliminate disagreement. Nor should it. But it can help identify whether disagreement is likely to be analytical, temporal, or hybrid.

Second, joint statements should do more than record areas of agreement and disagreement. They should expose the source of disagreement. Where the disagreement arises from different assumptions, evidential bases, instructions, pleaded cases or procedural timing, that should be made clear.

Third, Scott Schedules and issue schedules should be structured in a way that assists synchronisation. They should not simply present competing figures or conclusions. They should identify whether the expert witnesses are addressing the same question, by reference to the same assumptions and the same evidential platform. Frequently, apparent expert witness disagreement results not from professional judgement alone, but from a combination of differences in evidence, assumptions, instructions, pleaded cases and procedural timing. A well-designed Scott Schedule should expose those differences and distinguish them from the matters upon which the expert witnesses truly disagree.

Fourth, tribunals should consider whether material procedural developments require expert witness evidence to be updated before it is compared. An opinion that remains unchanged despite a materially changed evidential landscape is not necessarily a sign of consistency. It may simply be a sign that the platform has moved while the analysis has remained where it was.

Fifth, cross-examination should be directed with care. If the real issue is temporal divergence, cross-examination directed solely at methodology may miss the point. The question is not only why the expert witness reached a particular conclusion. It is whether the conclusion was reached by reference to the same version of the dispute as the competing opinion.

Finally, tribunals should be cautious before drawing credibility conclusions from apparent inconsistency or rigidity where the platform has moved. An expert witness who changes an opinion after the evidential landscape changes may not be unreliable. They may simply be responding properly to the movement of the platform. Conversely, an expert witness who does not change an opinion may not be consistent in any meaningful sense. They may simply be fixed to an earlier version of the dispute.

These are practical points. But they all flow from the same principle. Expert witness comparison should only occur after reasonable procedural synchronisation has been achieved.

 

Conclusion

Expert witness disagreement is an inevitable feature of arbitration. What is not inevitable is the way in which that disagreement is understood.

Arbitration often treats competing expert witness opinions as competing answers to the same question. In many cases, that assumption is justified. Sometimes it is not.

The platform upon which expert witness evidence is constructed does not always remain fixed throughout the life of the proceedings. Facts evolve. Evidence emerges. Assumptions change. Legal cases develop. The dispute evolves.

When that occurs, expert witness disagreement may not be a disagreement of opinion at all. It may be a disagreement created by time. What appears to be disagreement between expert witnesses may actually be disagreement between procedural moments.

That is why temporal divergence requires recognition as a distinct category of expert witness disagreement. Analytical divergence involves different reasoning from a common platform. Temporal divergence involves different procedural versions of the platform. Hybrid divergence involves both.

The distinction matters because it affects how expert witness evidence should be assessed.

Before deciding which expert witness is correct, tribunals should first consider whether the expert witness opinions have been procedurally synchronised.

• Are the expert witnesses working from the same evidential record?
• Are they addressing the same pleaded case?
• Are they operating from the same assumptions?
• Have both analyses been updated to reflect material developments?
• If not, has the disagreement been caused by a lack of synchronisation?

These questions do not complicate the tribunal’s task. They clarify it.

They help distinguish disagreement that should be decided from disagreement that should first be understood.

The point has consequences for weight. Analytical divergence that survives synchronisation is more likely to call for determination. Temporal divergence may first call for alignment. Hybrid divergence requires the tribunal to separate what is truly analytical from what is merely procedural.

The central proposition is therefore simple. Because the platform moves, tribunals should determine whether expert witness opinions have been synchronised before comparing and evaluating them. If they have, disagreement can be assessed on its merits. If they have not, understanding the source of that instability becomes the first step in understanding the disagreement itself.

Before deciding which expert witness is right, tribunals should first ask whether the expert witnesses are standing on the same version of the dispute.

Viewed in this way, the management of expert witness evidence does not end with receiving competing opinions. It includes ensuring, so far as reasonably possible, that those opinions are being compared from a sufficiently synchronised procedural platform.

Because sometimes the most significant difference between two expert witness opinions is not how the expert witnesses think.

It is where, and when, they are standing.

 

To discuss expert witness strategy, quantum evidence or arbitration support, contact our team.

 

About the authors

Michael Tonkin is a Chartered Quantity Surveyor, Chartered Arbitrator, RICS Accredited Expert Witness and CEDR Accredited Mediator with over 35 years of construction industry experience. He is an effective dispute resolver and has acted as quantum expert on around 80 occasions.

Igor Corelj is a Chartered Quantity Surveyor and structural engineer with nearly 20 years of international construction industry experience. He has acted as a quantum expert six times and has assisted the named expert on numerous occasions.