James Coombs v The Information Commissioner

This appeal is brought under s.57 of the FOIA against the Commissioner’s DN dated 5 August 2022 with reference IC-82675-K0J4 which is a matter of public record.

The Tribunal conducted a CVP hearing at which it heard evidence from Mr Coombs and considered an agreed open bundle, submissions from Mr Coombs in a number of emails and a list of authorities not included in the original bundle and listed in Annex A. The Tribunal took into account all the evidence before it and made findings on the balance of probabilities.

The full details of the background to this appeal, Mr Coombs’ Request for information and the Commissioner’s decision are set out in the DN.

The Public Authority is the Consortium on behalf of 18 schools as set out in the DN. The Consortium was joined as a Second Respondent by case management direction (“CMD”) on 12 January 2023. Mr Coombs applied for a reconsideration of the CMD on 12 January 2023 and the Consortium was removed as Second Respondent on 13 February 2023. In view of its decision to allow the appeal, the Tribunal sets aside the decision of 13 February 2023 and directs that the Consortium should be joined as a party to the appeal.

On 30 October 2020, the Appellant made a FOIA Request to the Consortium for information in the following terms:

“Please provide anonymised data for tests taken in 2019 for entry to grammar schools this September (2020). For each candidate who sat the test please include the following

• Date of birth*

• Verbal reasoning raw score

• Verbal reasoning standardised

• Non-verbal reasoning raw score

• Non-verbal reasoning standardised

• Total age weighted score

• Entitlement to Pupil premium funding (if this is captured)

Please could you also explain what data is returned by the test provider (to avoid Requests like, “Pupil Premium funding if captured”)

*I understand the tests are taken by children applying to fifteen schools so would estimate (500 applicants per school) seven eight thousand results in total with approximately 20 children sharing each date of birth. If you are concerned that this information should be withheld under s.40(2) because it would contravene any of the

data protection principles in the 2018 Data Protection Act, please explain why this is the case and consider instead rounding all values to the nearest week.”

The Consortium responded on 26 November 2020 refusing the Request on the basis of s.43(2) (commercial interests) of the FOIA. The Consortium stated that disclosure would prejudice the interests of the Consortium’s 11 test provider, GL Assessment Limited (“GLA”). The response referred, also, to s.40(2) FOIA (third party personal information). Mr Coombs asked for an internal review of this decision on 27 November 2020. He provided detailed arguments to support his view that the exemption had been incorrectly applied.

The Consortium conducted an internal review and provided the outcome on 7 January 2021. It upheld its decision to refuse the Request under s.43(2) and added some more detail to this although considered s.43(2) to be the primary exemption.

Mr Coombs contacted the Commissioner on 15 January 2021 to complain about the way his Request for information had been handled.

During the course of the Commissioner’s investigation, the Consortium amended its position and stated that it was still relying on s.43(2) but considered that for the date of birth and entitlement to pupil premium funding the Consortium considered this would not, on its own, engage s.43(2) so was now also being withheld under s. 40(2).

The Commissioner considered the scope of his investigation to be to determine if the Consortium has correctly withheld the information requested on the basis of s.43(2) of the FOIA, and if s.43(2) has not been correctly applied he went on to consider whether s.40(2) was engaged in relation to the dates of birth and entitlement to premium funding.

On 5 August 2022 the Commissioner issued the DN as follows:

On 2 September 2022 Mr Coombs appealed against the Commissioner’s DN.

A person requesting information from a public authority has a right to be informed by the public authority in writing whether it holds the information (s.1(1)(a) FOIA) and to have that information communicated to him if the public authority holds it (s.1(1)(b) FOIA).

However, these rights are subject to certain exemptions. The relevant exemptions in this appeal are s.40(2) and s.43(2) FOIA.

S.40(2) – Personal data is defined as “data which relate to a living individual who can be identified (a) from those data, or (b) from those data and other information which is in the possession of, or is likely to come into the possession of, the data controller, and includes any expression of opinion about the individual and any indication of the intentions of the data controller or any other person in respect of the individual.”

The effect of s.40(2) of FIOA is that a request for personal data of a third party may only be disclosed if such disclosure is compatible with the data protection principles enshrined in the Act (s.40(2). It is an absolute exemption, so the public interest balancing test does not apply (FOIA, s.2(3)(f)(ii)). The term “personal data” is broadly defined by section 1 of the Act as “data which relate to a living individual who can be identified”. The first data protection principle (as set out in Part 1 of Schedule 1 to the Act) is that personal data “shall be processed fairly and lawfully” and in particular shall not be processed unless at least one of the conditions in Schedule 2 is met. Aside from the individual’s consent (Schedule 2, condition 1), the most significant provision is as follows:

“The processing is necessary for the purposes of legitimate interests pursued by the data controller or by the third party or parties to whom the data are disclosed, except where the processing is unwarranted in any particular case by reason of prejudice to the rights and freedoms or legitimate interests of the data subject.”

Reliance on the data protection principles under s.40(3)(a) FOIA constitutes an absolute exemption (s.2(3)(f) FOIA). The first data protection principle is that:

“Personal data shall be processed fairly and lawfully and, in particular, shall not be processed unless—

(a)

at least one of the conditions in Schedule 2 is met, and

(b)

in the case of sensitive personal data, at least one of the conditions in Schedule 3 is also met.”

Processing includes “disclosure of the information or data by transmission, dissemination or otherwise making available” and therefore includes disclosure under FOIA.

s.43(2) – commercial interests

This exemption is subject to the public interest test. Information likely to prejudice the commercial interests of any party must still be disclosed under the FOIA unless the balance of the public interest favours maintaining the exemption. The weight to be given to the public interest varies depending on the likelihood and severity of the prejudice.

The Tribunal stands in the shoes of the Commissioner and takes a fresh decision on the evidence. The Tribunal does not undertake a review of the way in which the Commissioner’s decision was made.

Mr Coombs in his grounds of appeal dated 2 September 2022 submitted the following points:

The Commissioner submits the following:

The Consortium submitted the following points:

On an appeal under s.57 FOIA the task of the Tribunal is to decide whether the DN is in accordance with the law (s.58(1)(a) FOIA). The Tribunal is entitled to review any finding of fact on which the DN was based (s.58(2) FOIA).

In reaching its decision the Tribunal took account of all the evidence before it whether or not specifically referred to in this Decision. The Tribunal applied the legislation and case.

The Consortium is itself not a public authority, although each school within the Consortium is a public authority. The Consortium acts as a single point of contact and processes the Request on behalf of the schools. The Consortium has been regarded as referring to the public authorities.

In relation to pupil premium the Tribunal found that the information was held individually about each pupil who applied for a place at each individual school and this information was not collected by the Consortium and not used in the calculation of standardisation scores. Mr Coombs told the Tribunal that he accepted that the pupil premium information was not held by the Consortium and he did not wish to pursue this matter which was, therefore not in issue between the parties.

Lincolnshire is for the most part a selective county with children attending either a grammar school, secondary modern school or comprehensive school. The Consortium commissions bespoke 11 tests from GLA so that pupils applying can take the same set of tests. GLA produces and marks the 11 tests used by the Consortium.

Pupils who sit the 11 tests in Lincolnshire are assessed on their verbal and non-verbal reasoning. Pupils’ papers are marked and the raw scores are the number of questions a pupil gets correct in any given paper. These assessments are then converted to age-standardised scores. Each pupil will have a verbal reasoning raw score, an age-standardisation score, a non-verbal reasoning raw score and an age-standardised score.

The 11 test involves multiple choice answers. The answers are scanned into a machine and automatically marked. One raw mark is awarded for each correct question. No marks are awarded for incorrect or not attempted answers.

Standardised scores are used to determine admissions to grammar schools. Standardisation is a statistical process that rescales scores from different tests so they can be compared. Standardisation puts the results onto a common scale.

The guide to parents explains that “After each test is marked, the number of correct answers achieved, along with the age of the child (in years and months at the time of taking the test) is used to calculate a ‘standardised’ score.”

GLA is a test provider with a role to decide how examinations should be set and marked and the outcomes translated into results that can then be used to allocate or deny school places.

Age-weighting is used to correct the small but statistically significant differences in older children’s test scores. The chronological age of the oldest and youngest candidates differs by 365 days. Age standardisation is applied to raw results to adjust for the relative disadvantage of younger candidates.

The Tribunal found that s.43(2) FOIA was not engaged. In reaching its decision the Tribunal has borne in mind three steps involved in the application of the ‘prejudice test’ set out in Hogan and Oxford City Council v ICO (EA/2005/0026). The Tribunal has borne in mind, also, that in Guardian Newspapers Ltd v Brooke v ICO and BBC (EA/2006/0011) the Tribunal held that “would be likely to” meant that prejudice “there would be a “very significant and weighty chance” that it would occur. A “real risk” is not enough; the degree of risk must be such that there “may very well be” such inhibition, even if the risk falls short of being more probable than not.”

The Tribunal found that the commercial interests of GLA is an applicable interest in relation to the s.43(2) exemption.

The Tribunal is not persuaded that a causal relationship exists between the potential disclosure and the prejudice or that the prejudice would be a real and significant risk.

In reaching its decision, the Tribunal attached weight to the opinion of Mr Alan Parker dated 2 July 2021 (pages D84 to D91). He was a former Director of Education and Schools Adjudicator and is likely to be in a knowledgeable position to offer an informed and reliable opinion.

The Tribunal found that the knowledge of what happens to raw scores could be of no benefit to any tutor and, therefore, there could be no prejudice.

The Tribunal found that the disclosed information would not help anyone to understand how examinations are constructed and assessed because the raw scores are produced on the basis of correct answers to questions.

The Tribunal was not persuaded on the basis of the evidence that there was a significant and weighty chance that GLA would be put at a competitive disadvantage in future bids. The Tribunal was not persuaded that the raw and standardised score tables were in nature “extremely sensitive and confidential” as claimed by the Consortium (C56) or that disclosure would lead to competitors attempting to undermine their scoring/standardisation approach (D104). The evidence does not support the assertion that disclosure would affect GLA’s ability to participate competitively in a commercial activity. GLA’s method of calculating scores cannot be strictly confidential (D104). The questions are marked depending on the number of correct answers.

The Tribunal was not persuaded that disclosure would undermine the integrity of the tests on the grounds asserted, namely “… it would allow a tutor to deliver more targeted tutoring, undermining the core purpose of the test” (page C56). The integrity of the exam would not be undermined and it was more likely that the integrity would be enhanced as would have public confidence.

The Tribunal has considered carefully Mr Hilton’s witness statement but attached little weight to it because it contained a number of inaccuracies making it unreliable. Mr Hilton is the Head of Admissions Testing for GLA which produces and marks the 11 tests used by the Consortium. He stated (paragraph 8 of his witness statement) that candidates were “not privy” to their raw scores. This is not correct as all candidates are entitled to obtain their own personal data and have a right to request information about their performance. Mr Hilton asserted that GLA’s proprietary methodology for age standardising raw scores was a “trade secret” (paragraph 10). He asserted that a competitor could gain insight into the methods used and by “reverse engineering the methodology” thereby identifying those “… sections of the tests are most optimal in terms of maximising a candidate’s score.” For the reasons as submitted by Mr Coombs this would not be possible because raw scores are produced on the basis of correct answers to questions. No sections of the test could maximise a candidate’s score. GLA’s method was not its intellectual property as asserted by Mr Hilton. The process of standardisation is in the public domain.

The Tribunal is not persuaded that disclosure would give any advantage to an individual tutor or a tutoring organisation. Any tutoring organisation would be interested in the areas covered by questions and what knowledge and skills a child would need to achieve the highest possible score but this is not relevant to how the raw score is used thereafter. Disclosure would not provide inside knowledge about which topics or subjects carried the most marks because scoring is based on the number of correct answers achieved and the papers are equally weighted.

The Tribunal is not persuaded that disclosure would have any potential to identify in the test which sections were most optimal in relation to maximising a candidate’s score.

The Tribunal found that there was not a real and significant risk of prejudice by the disclosure of the information.

In reaching its decision the Tribunal attached weight to the fact that other public bodies have disclosed this information. The Essex Consortium publishes raw and standardised marks from its test every year along with an explanation of exactly how the calculations are done.

The Tribunal accepted and adopted the comprehensive and compelling submissions by Mr Coombs without rehearsing the detailed evidence and submission and the Tribunal accepted and adopted the substantive submissions on the material issues.

In any event, information likely to prejudice any party’s commercial interests must still be disclosed under FOIA unless the balance of the public interest favours maintaining the exemption.

The weight to be given to the public interest varies depending on the likelihood and severity of the prejudice. The Tribunal found that severity of any prejudice would be small. The Tribunal found that disclosing the information better served the public interest than maintaining the exemption. The Tribunal considered the need for transparency of standards and procedures and that disclosure was significant for the general public as it concerned the use of public money. The Tribunal found that the disclosure would assist the public in understanding the process by which grammar schools decide which pupils will be admitted and the fact that similar information about other grammar schools was already in the public domain was a compelling argument for disclosure.

The Tribunal found that there is a strong public interest in transparency particularly in the 11 process. Although no precise figures are before the Tribunal, Mr Coombs asserted that each of the grammar schools in the Consortium costs approximately £5 million to run each year and there is a belief that grammar schools offer a superior level of public funded education. There is strong public interest in understanding the process by which grammar schools decide which pupils they admit. The Tribunal found this a compelling argument.

The Tribunal found that similar datasets have been disclosed revealing mistakes in the processing. Disclosure would enable 7,500 sets of parents, guardians and carers to scrutinise the calculations. There is strong public interest in an external and objective assessment of the quality of the 11 tests.

There is a strong public interest in ensuring the Consortium is accountable for the way it spends public money and that any academic selection process can be understood by those involved in it. There is public interest in understanding the reliability of the tests used by grammar schools when determining which children are admitted.

The Tribunal found that the public interest in disclosure outweighs the information being withheld. In reaching this decision, the Tribunal considered that transparency and accountability promote public understanding, uphold standards of integrity, ensure justice and fair treatment for all, and secure the best use of public resources. The Tribunal considered, also, that that this information is not available in the public domain by other means.

There is no suggestion that the Consortium has acted improperly. However, there is public interest in test providers being open and transparent about how raw marks are processed to produce the standardised scores used in the admissions determination and how state funded selective schools operate.

S.40 FOIA provides an exemption from the right to information if it is personal data as defined by the Act. ‘Personal data’ means any information relating to an identifiable living individual.

In reaching its decision the Tribunal has considered that in order to determine whether a person is identifiable, account should be taken of all the means that are reasonably likely to be used by any person to identify the individual. The ‘motivated intruder’ test has been approved by the Upper Tribunal in Information Commissioner v Magherafelt District Council [2013] AACR 14 at 37-40 and 87. The ‘motivated intruder’ is taken to be a person who starts without any prior knowledge but who wants to identify the individual from whose personal data the anonymised data has been derived. This approach assumes that the ‘motivated intruder’ is reasonably competent, has access to the internet, libraries, and all public documents, and would employ investigative techniques such as making enquiries of people who may have additional knowledge of the identity of the data individual or advertising for information.

Mr Coombs estimated that approximately 20 children would share each date of birth and by aggregating date of birth data to the nearest week this would result in approximately 140 candidates sharing each birth week.

The Tribunal found that individuals could not be identified if the disclosure is in the form directed and would not be personal data as defined.

To identify an individual’s test results from the requested information a motivated intruder would need to establish the individual’s date of birth and eligibility for pupil premium funding. The motivated intruder would need to identify the scores of all the other data subjects with the week of birth and same pupil premium funding and then persuade them to disclose their individual scores in order for their records to be eliminated.

The Tribunal found that pupil premium funding is held in confidence and would not be disclosed.

The Tribunal found that an individual could not be identified from the disclosed information and other information which is in the possession of or likely to come into the possession of any person after it has been disclosed.

The Tribunal was satisfied that the information could be sufficiently anonymised by not providing dates of birth but birth weeks. The likelihood of identifying an individual was too remote to satisfy the test.

Although the Consortium holds information which would identify the pupils to whom the requested information relates, it does not follow that the sufficiently anonymised information would still be personal data when disclosed. Anonymised data which does not lead to the identification of a living individual does not constitute personal data. It is personal data in the hands of the Consortium but is not personal data to the requester because the public cannot identify any individual from it.

In reaching its decision the Tribunal has considered all the means likely and reasonably to be used to identify an individual.

The Tribunal considered that the likelihood of the motivated intruder identifying 20 children out of 7500 with a specific date of birth and getting them to disclose their results is very remote and too remote to satisfy the test in Dept of Health v ICO EWHC 1430 and ICO v Magherafelt UKUT 263.

The Tribunal found that disclosure of the information as stated in the Substituted Decision Notice would not contravene the data protection principles and the Consortium was not entitled to withhold it under s.40 FOIA.

The Tribunal considered whether disclosure of the personal data under FOIA would be lawful. The Tribunal found that Mr Coombs had a legitimate interests in this information for the reasons put forward by him. There is public interest in ensuring scrutiny and transparency. The Tribunal found that the disclosure of the information is reasonably necessary for the purposes of the identified interests. The Tribunal has balanced these interests against the privacy right and expectations of the data subjects and finds that the interests in disclosure are not overridden by the interest or fundamental rights and freedoms of the data subject which require protection of personal data.

The Tribunal found an error of law in the Commissioner’s DN in deciding that the information was correctly withheld and that s.43(2) was engaged; it was not necessary to consider s.40(2) FOIA. Accordingly, the appeal is allowed.