Learning purpose

This lesson examines how the burden and standard of proof govern disputed construction facts, contractual entitlement and causation. It explains why a claimant must establish its own positive case, why a respondent's unproved alternative does not automatically establish the claim, and how a tribunal may draw a reasoned inference without converting possibility into probability.

Successful completion requires the learner to distinguish legal and evidential burdens; apply the balance of probabilities to particular facts; separate proof of an event from proof of its consequences; evaluate competing explanations without assuming that they are exhaustive; use Rhesa Shipping Co SA v Edmunds (The Popi M) and Datec Electronics Holdings Ltd v United Parcels Service Ltd correctly; and prepare a proof schedule which connects every pleaded proposition with the records required to establish it.

1. Legal burden, evidential burden and the pleaded case

The legal burden identifies the party which must establish a fact in issue before the tribunal can decide that fact in its favour. In an ordinary contractual claim, the claimant must prove the facts required by the contractual cause of action and the relief sought. A contractor claiming an extension of time for late design information must therefore establish the contractual event, the relevant contractual condition, its own compliance with any applicable procedure, and the causal effect upon completion. Monetary recovery also requires proof of the recoverable loss and its causal connection with the relied-upon breach or contractual event. The exact propositions depend upon the contract and pleadings, so the phrase "the claimant bears the burden" begins the analysis without completing it.

That burden concerns facts which the governing law or contract requires that party to establish. It does not mean that every document must originate from the burdened party, because project records held by either participant may prove the necessary proposition. Nor does it mean that the claimant must disprove every imaginable alternative account before presenting a viable claim. The claimant must adduce evidence which makes its affirmative case more probable than not after the evidence has been evaluated as a whole. Where the evidence leaves the tribunal unable to decide whether the affirmative proposition probably occurred, the party bearing that legal burden loses upon that unresolved proposition.

The evidential burden describes a different and more practical responsibility during the presentation of a case. A party which advances a particular factual explanation must ordinarily produce sufficient material to make that explanation a live issue requiring consideration. Once the claimant produces records capable of establishing late information and resulting delay, the respondent may need evidence supporting its allegation that labour shortage, defective work or resequencing caused the same period. That practical need does not necessarily transfer the claimant's continuing legal burden upon its pleaded entitlement. The distinction prevents an evidential response from becoming an unexamined reversal of the governing legal burden.

Positive defences and counterclaims require their own burden analysis because the respondent may become the party asserting the material proposition. An employer relying upon an express contractual right of set-off must establish the facts and contractual conditions supporting that deduction. A respondent alleging that a release, settlement, waiver or contractual exclusion defeats an otherwise established claim must prove the factual foundation required by that defence, subject always to the particular law and wording governing it. The same proceeding may therefore contain several legal burdens, each attached to a distinct issue rather than allocated permanently to one party for every purpose.

Pleadings, contractual notices and decision documents should expose those separate propositions before the evidence is assembled. A pleading which states only that "late drawings caused delay" obscures at least four matters: the required issue date, the actual issue date, the affected work and the consequence for contractual completion. Similarly, a response stating only that "poor productivity caused the delay" similarly conceals the trades, activities, periods and production records supporting that account. Proper definition matters because the tribunal applies the standard of proof to facts in issue, not to slogans describing an entire dispute.

Construction practitioners should therefore begin with a proposition schedule rather than a document dump. Each row should identify the party bearing the legal burden, the proposition requiring proof, the contractual or legal reason why it matters, the contemporaneous records supporting or contradicting it, and any evidence still missing. Separate rows should address occurrence, contractual qualification, procedural compliance, readiness, factual causation, contractual causation, time consequence and monetary consequence. That structure reveals where a persuasive narrative rests upon an unproved link, while preserving the distinction between evidence which establishes a fact and evidence which merely makes another explanation less attractive.

The House of Lords stated the continuing burden with particular clarity in The Popi M, as reproduced by the Court of Appeal in Howmet Ltd v Economy Devices Ltd [2016] EWCA Civ 847 at paragraphs 110 and 111. Lord Brandon explained that the shipowners retained the burden of proving loss by an insured peril, although the underwriters could suggest another cause without assuming an obligation to prove that alternative. He also recognised that a court may remain unable to identify the true cause after extensive evidence, in which event the party bearing the burden has failed.

That practical consequence imposes a demanding but legally coherent evidential discipline upon both parties. Weakness in the respondent's alternative may strengthen the relative force of the claimant's evidence, yet it does not erase a missing element from the claimant's own case. Conversely, the respondent cannot defeat an otherwise proved claim merely by naming a conceivable competing cause without evidential support. The tribunal must identify the proposition, allocate its burden, examine all relevant evidence and decide whether the required standard has been met. Burden analysis supplies the consequence of unresolved evidence; it should not replace a conscientious evaluation which the available evidence permits.

2. The balance of probabilities and the quality of proof

The ordinary civil standard asks whether the fact in issue more probably occurred than not. Its application concerns the tribunal's evaluative conclusion upon the evidence rather than a mechanical percentage assigned to individual documents. In Jones v Birmingham City Council [2023] UKSC 27, the Supreme Court confirmed at paragraphs 49 to 51 that common law recognises one civil standard, namely proof upon the balance of probabilities. Inherent improbability remains a relevant factor when deciding whether an event occurred, so proof of an improbable event may require more cogent evidence. At paragraph 51 their Lordships rejected any general rule that a serious allegation, or serious consequences for the defendant, requires more cogent evidence before it can be proved. Gravity therefore does not alter the degree of probability which the civil standard requires.

Applying that principle requires close attention to the precise factual proposition requiring determination. A tribunal may find that a drawing arrived on 8 August because the document-control record, transmittal and receipt acknowledgement agree. More difficult evidence is required to decide whether installation would probably have started on 4 August if that drawing had arrived earlier. The second proposition concerns a hypothetical project history and requires evidence about labour, predecessor activities, materials, access and approvals. Both propositions are governed by the same civil standard, although the second requires a broader evidential foundation because more conditions affect the proposed conclusion.

The standard does not permit a tribunal to split a single past event into fractional factual findings merely because the evidence remains uncertain. Where a particular event must be found as a historical fact, the tribunal decides whether that event probably occurred. That does not prevent a later assessment of loss from addressing uncertainty through an accepted valuation method where the law permits estimation. It does prevent a claims analyst from presenting an unproved historical allegation as "sixty per cent true" and then treating the remaining percentage as an automatic discount. Questions concerning historical fact, legal causation and quantification must therefore remain analytically distinct throughout.

Contemporary records frequently carry persuasive weight because they were created near the events and for operational purposes, although their date alone does not make them conclusive. Daily reports may record a drawing hold while omitting that insufficient labour attended the workface. Labour returns may show headcount without establishing competence, allocation or productive hours. Programme updates may display movement without recording whether the logic, calendars and status dates were applied consistently. The tribunal evaluates provenance, purpose, completeness, internal consistency and compatibility with the wider record before deciding what each document proves.

Witness evidence may explain records, identify working practices and address matters which documents did not capture. Its value depends upon the witness's opportunity to observe, the clarity of recollection, consistency with contemporary material and willingness to recognise limitations. Expert evidence may organise programme, technical or accounting material and provide specialist opinion within the expert's discipline. However, the expert does not decide whether the legal burden has been discharged, because that evaluative decision remains for the tribunal upon the total evidence and the governing legal test.

The House of Lords demonstrated that distinction in Datec Electronics Holdings Ltd v United Parcels Service Ltd [2007] UKHL 23. At paragraphs 32 to 40, Lord Mance reviewed several possible explanations for the disappearance of computer processors, including theft by an employee, theft by an outsider and accidental misdelivery. The expert evidence could not independently identify the probable cause, yet the judge remained responsible for evaluating the documentary, circumstantial and factual evidence. Accordingly, the absence of an expert conclusion did not compel a finding that the claimants had failed.

Probability may arise through a combination of evidence which no single item could establish independently. A late drawing, a contemporaneous hold entry, an idle-workface photograph, material availability and a logic-linked update may together support the inference that information prevented planned work. Equally, labour returns, low output on unaffected areas and repeated late starts may weaken the proposition that the drawing alone caused completion delay. The correct evaluation considers whether the whole body of evidence makes the pleaded fact more probable than competing possibilities, without treating correlation, chronology or repetition as a substitute for causal reasoning.

Reasoned conclusions should state both the finding and the evidence which crosses the standard. A proper determination might find the drawing late as proved by the transmittal record, but find readiness unproved because labour allocation and predecessor completion remain uncertain. Another determination might find readiness proved through consistent daily records and material receipts, while limiting delay because the accepted programme contained available float. These separated findings show where the burden was discharged and where it was not, enabling the parties to understand the result without disguising evidential uncertainty beneath a single global conclusion.

3. Proving an event is different from proving causation

Proof that a contractual event occurred does not itself prove that the claimed consequence followed from that event. Construction claims frequently establish the first link convincingly because notices, instructions, drawing registers and meeting minutes record the event. The dispute then concerns whether the event affected the relevant activity, whether the claimant was otherwise ready, whether available float absorbed the effect, and whether the activity governed contractual completion. Each proposition carries its own evidential question, even where the same party bears the legal burden across the complete chain.

Every event proposition should define precisely what happened, when it happened and why it mattered contractually. For late information, the claimant must identify the information required, the date by which it was contractually or operationally needed, the date of actual availability and the work which depended upon it. Merely comparing a scheduled issue date with a later transmittal does not establish that the information was both required and usable on the earlier date. The programme, design responsibility matrix, request record, submittal status and physical sequence may all bear upon whether a proven late issue constituted an operative event.

By contrast, the causation proposition asks what difference the event probably made to the relevant contractual consequence. A drawing issued ten working days after its scheduled date does not automatically produce ten working days of completion delay. Work may have been planned to start only four days before actual issue, available float may have absorbed part of the effect, or another constraint may have prevented progress throughout the same period. The claimant must connect the event with the affected work and then connect the affected work with completion or the identified expenditure, applying the contract's required causal language.

Evidence of readiness frequently supplies the disputed bridge between the proved event and its alleged consequence. A contractor asserting that missing information prevented installation should identify available labour, delivered materials, completed predecessors, access, permits, approved methods and plant. Those matters need not be proved through one perfect record, but the combined material must support the proposed counterfactual. Where daily reports reveal insufficient installers and low output on available workfaces, the tribunal must assess whether the information event nevertheless caused some delay, whether both matters operated, or whether the claimant has failed to prove the claimed completion consequence.

Any respondent allegation concerning poor productivity must likewise be defined with factual and contractual precision. If the respondent merely challenges the claimant's proof of readiness and causation, the continuing legal burden may remain with the claimant. However, a respondent advancing poor productivity as the factual basis of a counterclaim, deduction or other affirmative contractual consequence, the respondent must establish the facts required for that separate case. This distinction prevents the same allegation from receiving an automatic burden allocation without examining the purpose for which the allegation is advanced.

Different remedies may require different causal demonstrations from the same project event. An extension-of-time provision may ask whether an employer-risk event delayed completion, while a loss-and-expense provision may require proof that identified expenditure resulted from that event. A claimant might prove a four-day effect upon the critical path yet fail to prove that an aggregated monthly site-cost total resulted from those four days. Conversely, a claimant may prove discrete abortive design expenditure without proving any effect upon contractual completion. The burden schedule should therefore maintain separate rows for time and money rather than allowing one conclusion to control both.

Decision-makers must also distinguish gaps within the evidence from genuine competing causes supported by positive records. Missing resource records may mean that readiness remains unproved, while positive labour returns may establish that reduced labour actually constrained production. An unsupported possibility should not receive the same status as a competing explanation grounded within records. Nevertheless, The Popi M prevents the tribunal from assuming that the only true cause must be one of the parties' pleaded explanations. If neither positive case is proved and the evidence permits an unidentified cause, the burden decides the unresolved issue against the party required to establish it.

A disciplined contractual determination therefore examines every necessary link along the causal chain in sequence. It identifies each necessary proposition, states which party bears its legal burden, records the evidence relied upon, addresses material contrary evidence and reaches a separate finding upon the balance of probabilities. That method may establish the event while rejecting the claimed duration, establish some causal effect while leaving the full completion consequence unproved, or establish time entitlement while rejecting unsupported monetary aggregation. Such differentiated findings reflect the evidence more faithfully than an all-or-nothing response to the general assertion that delay occurred.

4. Competing explanations and the rule in The Popi M

Competing explanations create a familiar temptation to select whichever account appears less unsatisfactory. That approach becomes legally unsafe when the available explanations are not exhaustive or when each remains improbable upon the evidence. The Popi M concerned a vessel which sank in calm weather and deep water, leaving important physical evidence unavailable. Shipowners alleged loss through a peril of the sea, while the underwriters suggested ordinary wear and tear. Ultimately, the House of Lords rejected the lower courts' selection of the shipowners' account where the evidence did not establish that explanation upon the balance of probabilities.

Lord Brandon's reasoning, reproduced in Howmet at paragraphs 110 and 111, preserves three connected propositions. First, the legal burden remains with the party required to prove the relevant affirmative case. Secondly, the opposing party may suggest an alternative without assuming a legal obligation to prove it, where the alternative merely challenges the claimant's case. Thirdly, the tribunal may conclude that the true cause remains unknown, even after prolonged investigation, if the evidence does not make the burdened party's account more probable than not.

The reasoning rejects an unqualified process which eliminates the impossible and accepts whatever explanation remains, however improbable. Such reasoning assumes that every possible explanation has been identified and that the underlying facts are sufficiently known to eliminate all but one. Construction records rarely justify such an exhaustive assumption without sustained examination of their completeness and reliability. Incomplete daily reports, changing sequences, unrecorded informal directions, unavailable operatives and retrospective programmes may leave additional explanations open. A tribunal should not transform a closed list drafted by advocates into a complete description of physical reality.

The rule does not mean that elimination and comparative probability are forbidden. Evidence which convincingly excludes realistic alternatives may support a positive inference, particularly where the selected explanation is consistent with the remaining records. The error arises when relative weakness becomes the sole foundation for accepting a case which remains improbable or unsupported. Instead, the claimant must establish its explanation through affirmative and circumstantial evidence considered together, rather than succeed because the respondent presented an even weaker account.

Disputed construction delays demonstrate the practical force of that doctrinal distinction with particular clarity. Suppose a claimant attributes ten days of delay to late drawings, while the respondent attributes the same period entirely to poor productivity. Evidence may establish that the drawings were unavailable for four working days after planned installation, while low staffing continued for another six days. Upon those circumstances, the proper conclusion need not select one global explanation for the complete period. Available evidence may establish a limited information effect, a later productivity effect, temporal interaction, available float, or insufficient proof of completion delay. Accordingly, the tribunal should follow the recorded chronology rather than force the facts into either pleaded total.

An alternative allegation can still matter substantially even when the respondent bears no legal burden to prove it. Credible evidence of insufficient labour may weaken the claimant's readiness case and prevent the claimant from establishing the proposed counterfactual. Conversely, the respondent's failure to support its allegation may remove a substantial challenge and strengthen the relative force of consistent claimant records. Neither consequence changes the governing rule: the tribunal asks whether the party bearing the legal burden has established the required proposition upon all the evidence.

The same reasoning applies to valuation and loss, although the material propositions differ. A contractor claiming prolonged preliminaries must prove the expenditure, the recoverable period and the causal connection with the relevant event. The employer may suggest that part of the expenditure arose from contractor delay without proving a separate counterclaim. If the claimant's accounts remain aggregated and cannot connect the sum with the proved event, rejection may follow from the claimant's own evidential failure. The employer's inability to quantify its alternative does not supply the missing causal allocation.

Practitioners should record the possible outcome "not proved" within every proof schedule. Removing that available outcome encourages false choices and produces recommendations which exceed the evidence. The schedule should ask whether the proposed explanations are genuinely exhaustive, which evidence supports the selected account, what evidence excludes realistic alternatives, and whether the conclusion remains probable without relying solely upon the weakness of the opposing case. Those questions preserve the common-sense application of the civil standard required by The Popi M while allowing properly supported inferences to be drawn.

5. Case focus: Datec Electronics Holdings Ltd v United Parcels Service Ltd

Datec Electronics Holdings Ltd v United Parcels Service Ltd concerned computer processors which disappeared during carriage from the United Kingdom to the Netherlands. The contractual and Convention context made the cause important because the ordinary limitation of liability under the Convention on the Contract for the International Carriage of Goods by Road could be displaced where the loss resulted from conduct treated as wilful misconduct under the applicable law. After trial, the judge concluded at paragraph 41 that the probable cause of the loss was accidental rather than theft. The Court of Appeal majority reversed that evaluation, and the House of Lords agreed with it at paragraphs 50 and 51, holding that theft involving a United Parcels Service employee had been shown upon a strong balance of probability and dismissing the carrier's appeal.

Lord Mance reviewed the evidential possibilities at paragraphs 32 to 40, including theft by an employee, theft by an outsider and accidental delivery elsewhere. The experts could not determine the probable cause from their specialist material alone. Nevertheless, their inability to select the cause did not prevent the trial judge from considering the whole factual and circumstantial record. The judicial task was to decide probability upon all admissible evidence rather than delegate the ultimate conclusion to experts or treat their uncertainty as legally decisive.

On appeal, the Court of Appeal accepted that the employee-theft explanation was more probable than the other explanations considered together, rather than merely more probable than each alternative viewed separately. Lord Mance examined that appellate reasoning carefully across paragraphs 41 to 50 of his opinion. At paragraph 48, he distinguished The Popi M and rejected the submission that the Court of Appeal had simply chosen the least unlikely possibility. His analysis at paragraphs 49 and 50 concluded that the reasoning depended upon the positive and negative evidence considered as a whole, which supported the inference of employee theft upon the required standard.

This distinction governs the proper use of comparative reasoning whenever several factual explanations remain available. A tribunal may compare explanations, examine whether records exclude or weaken alternatives, and infer the probable cause from circumstantial evidence. The selected explanation must still cross the balance of probabilities when compared with the remaining possibilities collectively. It is insufficient that the selected explanation appears marginally stronger than each of several alternatives separately, because the combined probability of the alternatives may still exceed the selected case. Nor may the tribunal assume that the parties have identified every possible explanation when the evidence leaves that assumption unsafe.

The decision also shows why expert neutrality about the ultimate cause does not produce automatic failure. Experts may establish matters which make one explanation more or less plausible, including system access, handling procedures, programme logic, productivity capability or cost allocation. The tribunal then evaluates those specialist conclusions alongside witness testimony and contemporaneous records. In a delay dispute, an expert may demonstrate that an update contains a logic path from late information to completion, while factual evidence may show that the relevant crew was unavailable. A proper final finding requires both evidential strands to be evaluated together by the tribunal.

Principles governing appellate restraint upon factual findings formed another material part of the decision. Lord Mance addressed the approach to factual findings at paragraphs 46 and 47 before examining whether the inference could properly stand. A claims specialist preparing an adjudication referral, arbitration submission or internal determination should therefore expose the evidential route to every important inference. Consequently, a conclusion identifying the records, contrary material and comparative reasoning is more capable of review than an assertion that one explanation simply "fits better". Transparent reasoning also permits later correction where an assumed factual premise lacks proper evidential support.

Applied to construction claims, Datec permits a reasoned inference from distributed records where direct evidence is unavailable. A drawing hold may be established through transmittals, requests, daily entries, photographs and later sequence changes, even though no single witness observed every consequence. Poor productivity may likewise be inferred from labour allocation, hours, quantities and output on unaffected areas. The analyst must then test whether one explanation is more probable than all realistic alternatives together, preserving the possibility that the burdened party has not proved its case.

Read together, Datec and The Popi M provide a balanced method for evaluating uncertain causes. Datec prevents evidential uncertainty from becoming a refusal to reason where the whole record supports a probable inference. The Popi M prevents comparative reasoning from becoming an obligation to select one unproved explanation. Construction practitioners should therefore build affirmative evidential chains, test realistic alternatives collectively, identify missing information and retain "not proved" as an available conclusion whenever the civil standard remains unsatisfied.

6. A proof schedule for construction claims

A useful proof schedule begins with the contractual result sought and works backwards through every required proposition. The schedule should contain separate columns for the proposition, burdened party, governing contract or legal test, supporting records, contrary records, present assessment and missing evidence. Each proposition should be sufficiently precise for a tribunal to answer it upon the balance of probabilities. Broad entries such as "employer delay" or "contractor fault" should be replaced with dated events, affected activities and identified consequences.

Each proof schedule should separate occurrence, contractual qualification and causation before addressing any claimed quantum. For a late-information claim, distinct rows should address the required information date, actual issue date, contractual responsibility, notice compliance, claimant readiness, effect upon the activity, available float, effect upon completion and resulting cost. Respondent allegations should appear against the propositions they challenge, while affirmative deductions or counterclaims should receive separate rows identifying the respondent's burden. This structure prevents the existence of a breach from silently proving every later element.

Every assessment should use controlled language which reflects the present evidential position. "Proved" means the available evidence presently makes the proposition more probable than not. The assessment "not proved" means that the burdened party has not crossed that evidential standard. "Further evidence required" identifies an open question without treating the gap as proof for either participant. The schedule should never assign invented percentages to unresolved historical facts, although programme sensitivity or valuation scenarios may be presented separately where they assist the contractual assessment.

7. Construction workshop: late drawings or poor productivity

Project record

Northbank Science Estates Ltd employed Orion Facade Systems Ltd to design and install the facade for a laboratory development. The fictional contract grants an extension of time where employer-provided design information delays completion, while the monetary provision requires proof of direct loss resulting from the relevant event.

• L10-R01 accepted baseline: Zone 3 bracket installation was planned between 4 and 22 August, the watertight milestone was planned for 29 August, and the displayed logic contained four working days of total float.
• L10-R02 information record: The coordinated bracket drawing was required by 25 July and transmitted on 8 August, which placed actual issue ten working days after the recorded requirement.
• L10-R03 labour record: Orion planned sixteen installers but recorded between nine and eleven installers from 4 until 15 August, with internal emails referring to absenteeism and redeployment elsewhere.
• L10-R04 daily reports: Several entries between 4 and 7 August record a Zone 3 drawing hold, while other entries record late starts and reduced production upon available Zones 1 and 2.
• L10-R05 programme updates: The 1 August update presents Zone 3 upon the longest path, while the 15 August update moves watertight completion to 13 September without resource loading or an inserted delay fragment.
• L10-R06 productivity records: Planned output was forty panels daily, while recorded output upon available Zones 1 and 2 ranged between twenty-two and twenty-eight panels during the reduced staffing period.
• L10-R07 contractor claim: Orion claims ten working days and GBP 145,000, alleging that the late drawing was the sole cause while presenting aggregated site expenditure for August.
• L10-R08 employer response: Northbank attributes the entire period to poor productivity but provides no independent programme analysis or quantified counterclaim.

Learner assignment

1. Identify every proposition which Orion must establish for its extension-of-time claim under the stated contractual wording.
2. Distinguish the date upon which the drawing became late from the period during which its absence could have affected installation.
3. Assess whether records L10-R01, L10-R02 and L10-R04 prove an operative information event upon the balance of probabilities.
4. Identify the evidence required to prove Orion's readiness to install within Zone 3 from 4 August.
5. State how the four working days of displayed float affect the pleaded ten-day completion claim.
6. Explain whether Northbank's failure to prove poor productivity would automatically establish Orion's causal case.
7. Apply The Popi M where late information and labour shortage remain possible explanations for the same period.
8. Apply Datec to the combined information, labour, productivity and programme records without selecting the least unlikely account.
9. State the strongest presently supportable conclusion concerning Orion's extension-of-time entitlement.
10. Assess whether the aggregated GBP 145,000 claim presently establishes direct loss caused by the proved information event.
11. Identify any affirmative proposition for which Northbank would carry a separate legal burden upon the present record.
12. Prepare a proof schedule identifying each proposition, burdened party, supporting record, contrary record and further evidence required.

Tutor observations

Within this exercise, the workshop deliberately separates drawing lateness from operative delay and completion delay. Although the drawing arrived ten working days after its recorded requirement, but Zone 3 installation was planned only from 4 August, leaving four working days before actual issue during which absence could directly prevent that planned work. Meanwhile, the displayed float, reduced labour and weak programme update require separate findings rather than an automatic ten-day award or rejection.

Northbank's unsupported global response does not relieve Orion from proving readiness and completion effect. Equally, poor productivity should not be accepted merely because output fell, since available workface, labour allocation, planned productivity and the contractual consequence require evaluation. The presently responsible conclusion should remain provisional until native programme files, detailed labour allocation, predecessor completion, daily allocation sheets and disaggregated cost records are produced.

8. Common errors in burden analysis

• Treating proof of a late instruction as proof of an identical period of completion delay collapses separate factual propositions.
• Assuming the respondent must prove every suggested alternative reverses the continuing burden upon the claimant's affirmative case.
• Selecting the least unattractive explanation without testing whether realistic alternatives are collectively less probable misuses comparative reasoning.
• Treating expert uncertainty as determinative overlooks the tribunal's responsibility to evaluate the complete factual and circumstantial record.
• Combining time and monetary consequences within one conclusion conceals different contractual tests and different evidential requirements.
• Ignoring the possible conclusion "not proved" forces a false choice where the available evidence does not cross the civil standard.

9. Practitioner checklist

• Define every fact in issue with a date, activity, contractual consequence and burdened party.
• Separate occurrence, contractual qualification, procedure, readiness, causation, completion effect and monetary consequence.
• Identify whether a respondent allegation merely challenges proof or supports an affirmative deduction, defence or counterclaim.
• Test every proposed inference against all realistic alternatives considered together, including an unidentified cause where appropriate.
• Record supporting and contrary material beside each proposition rather than organising evidence only by document type.
• Preserve missing evidence as an identified gap without filling the gap through assumption or rhetorical certainty.
• Explain why the complete evidential record makes each finding more probable than not.
• Maintain separate proof schedules for extension-of-time entitlement and monetary recovery where their contractual requirements differ.

10. Authorities and further reading

• Rhesa Shipping Co SA v Edmunds (The Popi M) [1985] 1 WLR 948, as reproduced and applied in Howmet Ltd v Economy Devices Ltd [2016] EWCA Civ 847 at paragraphs 110 and 111.
• Datec Electronics Holdings Ltd v United Parcels Service Ltd [2007] UKHL 23 at paragraphs 32 to 51.
• Jones v Birmingham City Council [2023] UKSC 27 at paragraphs 49 to 51.

11. Suggested answers to the workshop

Answer 1: Orion's propositions

Orion bears the legal burden of establishing the information requirement, actual lateness, contractual qualification, procedural compliance, readiness, causal effect upon the relevant activity, effect upon completion and any resulting direct loss. Records L10-R01 and L10-R02 address timing, but neither proves readiness or the claimed completion consequence. The test is proof upon the balance of probabilities, confirmed in Jones at paragraphs 49 to 51. Orion should obtain the contract terms, notices, predecessor completion records, labour allocation, native programme files and cost ledgers before asserting the complete entitlement.

Answer 2: Lateness and operative period

Record L10-R02 proves that the drawing arrived ten working days after the recorded requirement, but record L10-R01 shows installation beginning only on 4 August. The presently identifiable period during which absence could directly prevent planned Zone 3 installation is therefore 4 until 7 August, subject to readiness and calendar verification. Proof of drawing lateness does not establish an equivalent period of contractual completion delay. Orion should produce the accepted information schedule, working calendars, drawing status history and evidence establishing that the transmitted revision was usable immediately upon receipt.

Answer 3: Operative information event

Records L10-R01, L10-R02 and L10-R04 support a probable finding that missing information prevented planned Zone 3 work between 4 and 7 August. The baseline identifies planned work, the transmittal establishes availability only on 8 August, and daily entries record the drawing hold. That combined circumstantial reasoning is consistent with Datec at paragraphs 38 and 48 to 50. Further evidence remains necessary concerning predecessor completion, material availability, drawing approval status and whether alternative productive Zone 3 work remained accessible.

Answer 4: Readiness evidence

Orion must prove that Zone 3 installation would probably have proceeded if the drawing had been available. Records L10-R03 and L10-R06 raise contrary questions because staffing remained below plan and output was reduced elsewhere. The continuing burden principle in The Popi M, reproduced in Howmet at paragraphs 110 and 111, prevents readiness from being assumed merely because Northbank's response is weak. Orion requires trade allocation sheets, competence records, material receipts, access records, completed predecessor inspections, plant allocation and supervisor evidence addressing the proposed Zone 3 crew.

Answer 5: Effect of float

Record L10-R01 displays four working days of total float, which may absorb the presently identified four-day information effect before contractual completion moves. The display is evidence rather than a final finding because calendars, logic integrity, concurrent progress and subsequent recovery remain untested. Orion must establish completion delay under the contractual wording, not merely activity delay. Native programme files, update narratives, status records, longest-path calculations and evidence explaining consumption or ownership of float are required before the ten-day extension can be determined responsibly.

Answer 6: Northbank's weak alternative

Northbank's failure to prove its global productivity explanation would not automatically establish Orion's causal case. Under The Popi M, as recorded in Howmet at paragraphs 110 and 111, Orion retains the burden of proving the affirmative extension entitlement, while Northbank may challenge causation without proving a complete alternative history. Records L10-R03 and L10-R06 nevertheless provide some evidential foundation for the challenge. Orion must still prove readiness, critical effect and duration through labour allocation, programme logic, daily progress and predecessor evidence.

Answer 7: Applying The Popi M

The decision-maker should not choose late information merely because Northbank's poor-productivity case appears less developed. Records L10-R02 and L10-R04 support an information constraint, while records L10-R03 and L10-R06 support a simultaneous resource concern. The Popi M, reproduced in Howmet at paragraphs 110 and 111, permits the conclusion that Orion has not proved the claimed ten-day cause where the complete evidence remains doubtful. Further records should test separate periods, interaction, float, crew allocation and any additional unrecorded constraint.

Answer 8: Applying Datec

Datec at paragraphs 32 to 40 and 48 to 50 permits an inference from the complete factual and circumstantial record despite expert or documentary gaps. The analysis should compare late information, reduced labour and any remaining realistic explanations collectively, rather than select whichever account seems least unlikely. Records L10-R01 to L10-R06 presently support a four-day information hold alongside a longer staffing problem, but do not establish the completion result. Native programme analysis, allocation sheets, photographs, supervisor evidence and later progress records should complete the evidential comparison.

Answer 9: Present time conclusion

The strongest present finding is that an operative information constraint probably affected planned Zone 3 installation between 4 and 7 August, while the claimed ten-day completion extension remains unproved. Records L10-R01, L10-R02 and L10-R04 support the limited event, whereas L10-R01, L10-R03, L10-R05 and L10-R06 leave float, readiness and later productivity unresolved. Applying the balance of probabilities and The Popi M, Orion should receive no final ten-day determination until native programmes, resource allocation and subsequent critical-path evidence are examined.

Answer 10: Monetary claim

The aggregated GBP 145,000 claim in L10-R07 does not presently establish direct loss caused by the proved information event. Orion must separate actual expenditure, recoverable period, contractual category and causal connection from ordinary August project cost. The distinction between event and consequence requires an independent burden analysis for money, while Datec permits circumstantial inference only where the whole record makes that inference probable. Orion should provide ledgers, invoices, payroll, plant records, cost codes, baseline allowances and a period-specific explanation excluding unrelated labour inefficiency.

Answer 11: Northbank's affirmative burden

Northbank presently uses poor productivity principally to challenge Orion's proof, so Orion's legal burden upon its claim remains unchanged. If Northbank advances a contractual deduction or counterclaim based upon Orion's inefficiency, Northbank must establish the contractual right, breach, causal loss and valuation supporting that affirmative relief. Records L10-R03 and L10-R06 may support investigation but do not alone establish a quantified entitlement. Northbank should identify the contractual provision, pleaded consequence, measured output basis, causally related cost and any notices required for its separate case.

Answer 12: Proof schedule and directions

The proof schedule should create separate rows for information requirement, actual issue, usability, notice, readiness, activity effect, float, completion effect and direct loss. Orion bears those burdens for its pleaded claim, while any Northbank deduction receives separate affirmative rows. Records L10-R01 to L10-R08 should be entered as supporting or contrary material without treating either narrative as proved globally. Applying Jones, The Popi M and Datec, the decision-maker should direct production of native programmes, allocation sheets, predecessor approvals, detailed costs and witness explanations before final determination.

12. Knowledge check

The accompanying ten-question Knowledge Check tests burden allocation, the balance of probabilities, proof of causation, competing explanations, The Popi M, Datec and application of the fictional Northbank project record.

Closing point

Burden analysis identifies the consequence of unresolved evidence without excusing incomplete factual evaluation. A persuasive construction claim establishes every necessary proposition through an affirmative evidential chain, addresses realistic alternatives collectively and accepts "not proved" whenever the required standard remains unsatisfied.

Scope of the lesson

This lesson provides general professional education concerning English contract-law principles and construction-claims analysis. It does not provide legal advice upon any project, replace the governing contract, or determine how a court, tribunal, adjudicator or contract administrator would decide particular facts.