Summary
The Technology and Construction Court (TCC) has held that evaluators from NHS England (the Defendant) made a manifest error when assessing the answer given by Braceurself (the Claimant) to a single question, resulting in a scoring correction of one mark for that question. The Claimant failed in its challenges to five other scores and its broader allegations of ‘lack of transparency’ and ‘unequal treatment’. However, this one scoring change would have been enough to change the result, making Braceurself the highest scoring bidder.
Braceurself Ltd v NHS England [2022] EWHC 1532 (TCC) provides key lessons, including:
- For those designing the evaluation, about the benefit (as acknowledged in this case) of having evaluators bringing different skillsets to the evaluation, and evaluating a different allocation of questions, and giving them suitable training on how to apply the scoring criteria.
- For evaluators, about seeking to ensure they have properly understood what is written; have not made a mistake / erroneous assumption about what it means; have robustly challenged each other’s thinking in a way which might reduce the likelihood of the kind of ‘consensus mistake’ in this case; and potentially when to ask a bidder a clarification question to ensure understanding.
- For those assessing the legal risk of challenge at award stage, about recognising the risk of challenge when scores are close (perhaps more so where there is an initial difference of approach between evaluators to a key score under challenge); and spotting early the distinction between potentially legitimate scoring differences falling within the evaluators’ ‘margin of discretion’ (which are harder to challenge) and scoring differences resulting from straight error / mistake (for which no margin of discretion exists).
These and further observations of the Court are explored further below.
One final plot twist?
It is worth noting that there may yet be one final twist to this story. Several years ago in a case involving EnergySolutions and the Nuclear Decommission Authority, the Supreme Court held that for a breach of procurement law to warrant an award of damages to an aggrieved bidder, there must not only be a breach of procurement law by the purchasing authority, but that the breach(es) must be ‘sufficiently serious’ to warrant an award of damages – applying the ‘Francovich’ test for damages against a state entity.
As noted above, Braceurself is a case of a single mistake changing the outcome of a procurement. If what is sufficiently serious is to be judged by the impact – in this case, fundamentally changing the outcome of the procurement – it is easier to see how the breach could be held to be sufficiently serious to warrant damages. This is the approach that has previously been adopted in first instance judgments on the point. However, there is a question of whether this test should instead be judged by reference to the nature of the breach – i.e. in this case an error of competence in one requirement.
The Court declined to make a determination in this judgment as to whether the breach was ‘sufficiently serious’ to justify awarding damages to Braceurself. The Court ordered that a further hearing should address this point. This creates the interesting possibility that the Court might yet decide that although the breach changed the outcome of the procurement, it is not sufficiently serious to merit an award of damages. Bidders and authorities shall have to wait to see. If that is the outcome, in a case where a suspension has been lifted – or never applied – the outcome in practice might be that a challenging bidder is left without any effective remedy. That in turn will feed potentially into the question of adequacy of damages in suspension lift applications in future cases.
Postscript: After publication of this article, the Court issued its further judgment on the issue of whether the Defendant’s breach was “sufficiently serious” to warrant an award of damages. It found that the breach was not sufficiently serious. As such, even though the Claimant succeeded in establishing that the Defendant breached procurement law and that the Claimant should have been awarded the contract, it has not been awarded any damages and the contract is still being performed by the bidder which ought to have lost. We analyse this follow-up judgment in Braceurself Part 2: Successful Claimant denied damages as breaches were not “sufficiently serious”.
Further comments from the Braceurself judgment
This judgment raises several further points relevant for both purchasing authorities and bidders to consider for future procurements:
It was noted that the evaluators were not given training on how to provide a single score when considering multiple bullet point sub-criteria. This was potentially relevant because the requirement under challenge was broken down into a number of sub-bullets, raising the question of how scores should be allocated to those sub-bullets. The Court rejected the Claimant’s assertion that this resulted in a lack of transparency as to how the scoring would be applied:
‘….it was observed that the evaluators were not given training or guidance as to how they should provide a single score where the criterion entailed consideration of multiple bullet points. In particular, there was no sub-weighting given to one bullet point over another. This point had no merit. [One evaluator] expressed it well in her evidence when she said what was required was a composite score to a multi-faceted question.’
The Court approved this composite scoring approach taken by the Defendant where each bullet point was not the subject of sub-weighted scores, but instead resulted in a single overall score, because the bullet points were not an exhaustive list of factors.
- Recognition of the benefit of different evaluation skillsets:
Whilst the distribution of multiple evaluators across questions can create differences of approach, the Court held that this was a good thing:
‘The Claimant has suggested that the distribution of multiple evaluators in different permutations across the questions created a difference of approach and, therefore, inconsistency. I disagree. The whole point was that there should be different evaluators with appropriate skill sets and experience necessary to address the particular question. This was a good thing, not something about which there should be complaint. I do not agree there was an appreciable risk of inconsistency.’
This follows a similar finding in Bechtel v HS2 that having evaluators that bring a different range of experiences to an evaluation of the same requirement can, in fact, be beneficial, provided they all possess the necessary skills and experience to evaluate effectively.
- Very limited moderation notes do not necessarily result in a lack of transparency:
It has been an issue in the Court in procurement cases on the extent of detail required in evaluation and moderation notes in order for a purchasing authority to meet the core procurement principle of transparency. In this case, not only were such notes sparse but they also did not entirely align with the rationales given in the award letter. However, the Court held that:
‘Having heard the evidence, I am satisfied that, in general terms, there was sufficient correspondence between the notes and the feedback provided in the debrief letter. The notes did not purport to be a verbatim record and the purpose of both the notes and the spreadsheet compiled by the moderator were intended to capture the essence of having arrived at the moderated score. The Claimant has not begun to satisfy me that there was any error of process here, still less the respects in which any such error might have led to the Defendant having made a manifest error, or that it amounted to breach of the requirements for transparency or equal treatment.’
This is a divergence from earlier authorities which suggested a more granular level of evaluation and moderation recording was required.
- There is no margin of appreciation for mistakes:
The Court gives weight and deference to the expertise of the evaluators, allowing for a wide margin of discretion in the evaluation of bids. The Court will only interfere with scoring in cases of manifest error (i.e. where scores are obviously wrong).
In this case, however, on the one decisive point the evaluators had simply misunderstood what the Claimant had written, such that the score awarded was a result of error rather than evaluative judgment. As such, there was no margin of appreciation on that point.
Factual case summary: further detail of the case
The procurement at the centre of the dispute was for a seven-year contract for the provision of dental and orthodontic services in the south of England, sub-divided into geographic areas (the 'Procurement'). Two bidders were in competition for the contract award for the East Hampshire lot: Braceurself Ltd was the incumbent provider. Orthodontics by Eva Petersfield & Alton Ltd (‘PAL’) was the other bidder. NHS England was the tendering authority.
PAL won by a margin of 2.25%. The gap between the bidders’ overall scores was such that a score uplift of a single point on one question could turn the outcome of the Procurement.
Court Proceedings were issued by Braceurself in Autumn 2019 which invoked the automatic suspension on contract signature for the NHS’ contract with PAL, preventing award. The automatic suspension was lifted by the Court in November 2019 and the contract was then signed with PAL leaving Braceurself with mainly a claim for damages against the NHS. An in-person trial on liability took place over five days from 28 February 2022, and judgment was handed down in late June 2022.
Among the six allegations of manifest error, it was the challenge to the patient accessibility question that proved to be the focal point of the case, resulting in the Court re-scoring the bid in favour of Braceurself.
In summary, the evaluators of this question had misunderstood part of Braceurself’s answer: confusing a stair climber with a stair lift (which are not the same) and misunderstanding a reference to the circumstances in which Braceurself would offer ‘alternative premises’. The Court concluded that the evaluators could have clarified Braceurself’s proposed solution in a ‘matter of moments’. Instead, the evaluators made a collective mistake and therefore moderated on a factually incorrect basis, leading them to downgrade the Claimant’s score for its answer to this question from a 4 to a 3.
Two of the evaluators had given Braceurself’s response to this question a score of 4, but then, when in moderation with the third evaluator, they together decided to moderate this down to a 3, based on misinformed discussions on the accessibility component of Braceurself’s response to this question. The Court also found that these two evaluators could not account for their individual rationales during the moderation.
Importantly, accessibility was the sole area given in the feedback as a potential basis for improvement (indicating this was the key reason for the award of the score of 3, rather than 4). Since this feedback was based on the mistake explained above, the Court held that Braceurself’s score should not have been moderated down; it should have been a 4 (the highest available score).
The score uplift in favour of Braceurself increased its overall score by 2.5% and resulted in a score that was 0.25% higher than PAL’s original winning bid, therefore changing the outcome of the Procurement in favour of Braceurself.
This analysis is authored by two members of the Burges Salmon procurement team, Richard Binns (Partner) and Mark Gwilt (Associate).
Case: Braceurself Ltd v NHS England [2022] EWHC 1532 (TCC) (20 June 2022) (bailii.org)