Angela Powell v The Information Commissioner & Anor [2024] UKFTT 234 (GRC)

[2024] UKFTT 00234 (GRC)

Case Reference: EA-2020-0137

First-tier Tribunal

General Regulatory Chamber

Hearing Heard on: 22 February 2024

Date 16 April 2024

Information Rights

Heard: By CVP

Before

TRIBUNAL Judge BUckley

TRIBUNAL MEMBER JO murphy

TRIBUNAL MEMBER david cook

Between

ANGELA POWELLAppellant

The Information commissionerRespondent

EQUALITY AND HUMAN RIGHTS COMMISSIONRespondent

In person for Appellant

for First Respondent: Did not appear

for Second Respondent: Mr. Willis (Solicitor)

Decision: The appeal is allowed in part

Substituted Decision Notice:

Organisation: The Equality and Human Rights Commission

Complainant: Angela Powell

The Substitute Decision-

Signed Sophie Buckley Date: 18 March 2024

[1]For the reasons set out below: 1.1. The public authority was entitled to withhold the following information under section 40(2) of the Freedom of Information Act 2000 (FOIA): 1.1.1. The information requested on 15 March 2019. 1.1.2. The percentage of men and women in each hourly pay quarter. 1.1.3. The median pay of men. 1.1.4. The median pay of women. 1.2. The public authority was not entitled to withhold the following information under section 40(2) FOIA: 1.2.1. The percentage difference between median pay for women and median pay for men. 1.2.2. Mean pay for men. 1.2.3. Mean pay for women. 1.2.4. The percentage difference between mean pay for women and mean pay for men.

[2]The public authority shall disclose the information requested to the requestor within 42 days of the date this decision is sent to the public authority by the tribunal.

[3]Any failure to abide by the terms of the tribunal’s substituted decision notice may amount to contempt which may, on application, be certified to the Upper Tribunal. REASONS

Introduction

[1]This is an appeal against the Commissioner’s decision notice FS50890713 of 3 March 2020 which held that the Equality and Human Rights Commission (‘the EHRC’) was entitled to rely on section 40(2) of the Freedom of Information Act 2000 (FOIA) to withhold the requested information, but that the EHRC was in breach of section 17 FOIA in relation to the second request for information (referred to as ‘the March Request’ in this decision). The Commissioner found that the EHRC complied with its section 16 duty to provide advice and assistance. The Commissioner did not require the EHRC to take any steps.

[2]This is a remitted appeal. The Upper Tribunal in a decision reported at

[2023] UKUT 221 (AAC)

overturned the decision of the First-tier Tribunal promulgated on 16 December 2020 and remitted it to the First-tier Tribunal for a complete re-hearing.

[3]The EHRC was joined as second respondent after the appeal had been remitted.

Factual background to the appeal

[4]The EHRC was established by the Equality Act 2006. It has a statutory duty under section 8 of the Equality Act 2006 to work towards the elimination of unlawful discrimination and to enforce the Equality Act 2010. The Equality Act 2010 contains a right to equal pay for men and women for equal work or work of equal value, unless any difference in pay can be justified.

[5]The gender pay gap is the difference in average pay between men and women in a workforce. As part of its duties in 2017 the EHRC produced a document entitled ‘Fair opportunities for all: a strategy to reduce pay gaps in Britain”. Since April 2017 there has been a legal obligation on most employers with 250 or more employees to publish their gender pay gap information each year. The EHRC has the power to take enforcement action against any employer who does not comply with their reporting duties.

[6]A ‘Gender Pay Gap analysis’ is an analysis conducted in line with the process set out in the Equality Act 2010 (Gender Pay Gap Information) Regulations 2017 S.I. 2017/172. Employers with more than 250 employees are required to publish the results of this analysis, which comprises six pieces of information. The EHRC employs around 160 people, and so is not obliged to publish these results, but does so voluntarily due to its role as the ‘watchdog’ for equality issues in the UK.

[7]The six results required to be published are as follows: 7.1. The percentage of men and women in each hourly pay quarter 7.2. The mean gender pay gap using hourly pay (expressed as a percentage) 7.3. The median gender pay gap using hourly pay (expressed as a percentage) 7.4. The percentage of men and women receiving bonus pay (two figures) 7.5. The mean gender pay gap using bonus pay (expressed as a percentage) 7.6. The median gender pay gap using bonus pay (expressed as a percentage)

[8]This matter relates to the salaries of staff in the Birmingham office of the EHRC. At the relevant time there were 8 employees in the Birmingham office: 3 women, including Ms Powell, and 5 men.

[9]Ms Powell was employed by the EHRC from 2008 to 2017 and was made redundant in 2017. She maintains that the dismissal was unfair and that she was unlawfully underpaid whilst employed by the EHRC. She asserts that she was negligently advised to settle her employment tribunal claim by union solicitors.

[10]The appellant asks the EHRC to apply its ‘pay gap method’ to the requested data. This is a reference to a Gender Pay Gap analysis and an Equal Pay analysis.

[11]There was no bonus pay in the office in question, and therefore the latter three parts of the Gender Pay Gap analysis do not apply in this appeal.

[12]An ‘Equal Pay analysis’ as conducted by the EHRC, involves publishing the EHRC’s mean and median pay figures for the whole organisation. Requests, decision notice and appeal The request

Requests, decision notice and appeal

[13]This appeal concerns two requests for information made by Ms Powell.

[14]The following request (‘the March Request’) was made on 15 March 2019:

“Please find attached a table which requests details from the Birmingham EHRC office, Victoria Square House, namely by, employees in post by grade/name, job title and salary as at 9th February 2017. Please note that, the existing names in this table are only a guide. Please add, delete or change as appropriate so as to make it an accurate and transparent picture.”

[15]The following request (‘the May Request’) was made on 3 May 2019:

“Can you apply EHRC’s pay gap method to the data and provide me with your results”

The response

[16]On 25 March 2019 the EHRC responded to the March Request. It provided the number of employees at each pay grade and the details of the width of each pay grade. It refused to provide more detailed information, relying on section 40(2) FOIA. The EHRC upheld its position on internal review.

[17]On 28 August 2019 the EHRC responded to the May Request. It withheld the requested information under section 40(2). The EHRC upheld its position on internal review.

[18]Ms. Powell referred the matter to the Information Commissioner on 14 November 2019.

The Decision Notice

[19]The Commissioner noted that the March Request sought precise details about individuals, their salaries and their grade or job title. It also sought this information in tabular form, whereby each employee would have their information recorded on a separate row. In the Commissioner’s view, this amounted to the creation of a unique job “profile” for each employee. The Commissioner concluded it made little difference whether an employee’s name was redacted or not, as individuals could be identified from this job profile alone.

[20]The Commissioner concluded that the low number of employees (8) would make it even easier for a motivated individual to link some, if not all, of those employees with a profile. That individual would then have each employee linked to a job title and an individual salary (as opposed to a salary band).

[21]The Commissioner observed that it appeared that the complainant had at some point been employed by the EHRC and that this would make it easier for her to identify individuals. Further, even if the complainant herself were unable to identify individuals, there is a distinct possibility that it would come into the hands of people who would be able to “deanonymise” the data.

[22]The Commissioner concluded that how much an employee is being paid for doing their job was information that related to them. Accordingly he was satisfied that the information was personal data.

[23]In relation to the May Request, the Commissioner considered that it sought the same information in a slightly different format. He considered that, because of the numbers of employees involved, the statistical analysis could be unravelled sufficiently to reveal the raw data. The Commissioner considered that a person who was aware of the gender split in the Birmingham Office would be able to combine that information with the pay band information the EHRC has already disclosed to work out at least some of the individual salaries. The Commissioner considered that there was a risk that individuals could be identified and the information was therefore personal data for the same reasons set out earlier.

[24]The Commissioner considered that there was a legitimate interest in public bodies being held accountable for the way they spend that money, and a further legitimate interest in ensuring that the body responsible for policing equal pay law is an exemplar of best practice.

[25]The Commissioner concluded that disclosure was not necessary to achieve those legitimate interests. The Commissioner considered that the legitimate interest in ensuring that the EHRC is complying with the law was met by its publication of gender pay gap data.

[26]The Commissioner found that the EHRC breached section 17 of the FOIA in its delay in responding to the May Request.

[27]The Commissioner found that the EHRC complied with its duties to provide advice and assistance because he was not convinced that there was any further advice and assistance that the EHRC could reasonably have been expected to offer.

Notice of Appeal

[28]In summary, the grounds of appeal are that disclosure would not contravene the data protection principles.

The Commissioner’s response

[29]The Commissioner submits that the information plainly falls within the definition of personal data and that this is not disputed by the appellant. The Commissioner maintains that disclosure is not necessary for the purposes of the legitimate interests. Even if disclosure was held to be necessary the Commissioner submits that the third party individuals would not have had a reasonable expectation that their personal data would not be disclosed to the world at large as a result of such a request. The concerns of the appellant do not outweigh a number of third party individuals’ fundamental rights, freedoms and reasonable expectations of privacy. To process and disclose the information would be unlawful and it would contravene GDPR article 5.

[30]The Commissioner repeats his findings on section 16 FOIA.

The EHRC’s response

[31]The EHRC submit that the March Request is for personal data for the reasons set out in the Decision Notice at paragraphs 22-27. The appellant was an employee of the EHRC for 9 years which would make the ease of identifying individuals greater than suggested in the Decision Notice.

[32]In relation to the May Request the EHRC concur with the submissions by the Commissioner dated 12 December 2022 (in the Upper Tribunal proceedings) that personal data relating to all or some of the individuals known to the appellant is likely to be identifiable from the data if disclosed, particularly by the appellant and any one of the seven other individuals.

[33]It is useful to summarise those submissions from the Commissioner here. In essence the Commissioner submits that the information in the scope of the May Request includes the median salary of male employees and the median salary of female employees. The median salary would be the salaries of the ‘middle’ female and male employees who, given the small number of staff involved and the appellant’s familiarity with the organisational structure of her former employer, could easily be identified.

[34]If the information is limited to a Gender Pay Gap analysis the Commissioner submits that the appellant, with her knowledge of the EHRC Birmingham office (including broad job ranking within the office), would be able to calculate which individuals correspond to which pay quartiles and/or pay bands. Further, it is submitted that if the appellant knew approximately what salary at least one of her two female colleagues was being paid, she would be able to calculate the salary of the other women employees, and given her knowledge of the organisational structure of the EHRC’s Birmingham office, would be able to match those salaries to individuals and would then know the median pay for women employees and be able to calculate the median pay for male employees.

[35]The EHRC agree with the reasons set out in the Decision Notice at paragraphs 36 to 54 that there is no lawful basis on which to process the personal data.

[36]The Appellant has indicated that she proposes to use the requested data for a potential negligence claim. The EHRC submit that Data Protection Act 2018, schedule 2, paragraph 5 sets out the process and relevant considerations which makes it permissible for the data controller to disclose information in connection with legal proceedings; but this is outside of the FOIA.

Evidence

[37]We read an open and a closed bundle and a copy of the Upper Tribunal decision in this appeal. We also read and took account of: 37.1. The Commissioner’s response to the appeal to the Upper Tribunal dated 12 December 2022 (at page 62 of the Upper Tribunal Bundle). 37.2. Ms Powell’s 9 page ‘response’ in the Upper Tribunal proceedings.

[38]The closed bundle contains the withheld information within the scope of the March request. It is necessary to withhold this information from Ms Powell, otherwise the purpose of the appeal would be defeated.

[39]The closed bundle also contains two items of correspondence which the tribunal considers do not need to be in the closed bundle. One is a letter from the Commissioner to the EHRC following the Upper Tribunal decision asking the EHRC to provide the withheld information. That letter is irrelevant to the issues we have to determine and therefore does not need to be added to the open bundle. The other letter is the response to the March request and already appears in the open bundle.

[40]We also took into account an additional internal EHRC email dated 1 February 2024 provided by the EHRC during the course of the hearing, which sets out some calculations. A copy was provided to the appellant with personal data redacted and that redacted version is open. Legal framework Personal data

Legal framework

[41]The relevant parts of section 40 of FOIA provide:(1) Any information to which a request for information relates is exempt information if it constitutes personal data of which the applicant is the data subject.

(2) Any information to which a request for information relates is also exempt information if – (a) It constitutes personal data which does not fall within subsection (1), and (b) either the first, second or the third condition below is satisfied. (3A) The first condition is that the disclosure of the information to a member of the public otherwise than under this Act - (a) would contravene any of the data protection principles... …

[42]Personal data is defined in section 3 of the Data Protection Act 2018 (DPA): (2) ‘Personal data’ means any information relating to an identified or identifiable living individual (subject to subsection (14)(c)). (3) ‘Identifiable living individual’ means a living individual who can be identified, directly or indirectly, in particular by reference to—(a) an identifier such as a name, an identification number, location data or an online identifier, or

(b) one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of the individual.

[43]I adopt the following summary of the law on personal data from the Upper Tribunal decision in this appeal.

[44]Recital 26 of the preamble to the General Data Protection Regulation (EU) 2016/679 states: (26) The principles of data protection should apply to any information concerning an identified or identifiable natural person. Personal data which have undergone pseudonymisation, which could be attributed to a natural person by the use of additional information should be considered to be information on an identifiable natural person. To determine whether a natural person is identifiable, account should be taken of all the means reasonably likely to be used, such as singling out, either by the controller or by another person to identify the natural person directly or indirectly. To ascertain whether means are reasonably likely to be used to identify the natural person, account should be taken of all objective factors, such as the costs of and the amount of time required for identification, taking into consideration the available technology at the time of the processing and technological developments. The principles of data protection should therefore not apply to anonymous information, namely information which does not relate to an identified or identifiable natural person or to personal data rendered anonymous in such a manner that the data subject is not or no longer identifiable. This Regulation does not therefore concern the processing of such anonymous information, including for statistical or research purposes.

[45]In NHS Business Services Authority v IC and Spivack

[2021] UKUT 192 (AAC)

, Upper Tribunal Judge Jacobs said as follows: 4. I have to decide whether any person was identifiable from the data withheld by [the public authority] when taken together with other information by someone who was motivated to identify one or more of the persons within the data using all the means reasonably likely to be used … … Just looking at the legislation12. Section 3 of the 2018 Act creates a binary test: can a living individual be identified, directly or indirectly? If the answer is ‘yes’, the data is personal data. Otherwise, it is not. That is what the Act says, and it is consistent with the Regulation. There is no mention of any test of remoteness or likelihood.

13. The test has to be applied on the basis of all the information that is reasonably likely to be used, including information that would be sought out by a motivated inquirer, as in this case. That derives from Recital 26.

[46]Common Services Agency v Scottish Information Commissioner

[2008] UKHL 47

concerned a process of anonymisation of data, called “barnardisation”. The House of Lords said that it was a “question of fact” (for the Scottish information commissioner, in that case) “on which he must make a finding” as to whether, through barnardisation, the information became data from which a living individual could no longer be identified. “If barnardisation can achieve this, the way will then be open for the information to be released in that form because it will no longer be personal data” within the meaning of s1(1) of the Data Protection Act 1998 (paragraph 27 of the judgement).

[47]The data protection principles are set out article 5(1) of the GDPR. Article 5(1)(a) GDPR provides: that personal data shall be processed lawfully, fairly and in a transparent manner in relation to the data subject. Article 6(1) GDPR provides that processing shall be lawful only if and to the extent that at least one of the lawful bases for processing listed in the article applies.

[48]The only potentially relevant basis here is article 6(1)(f):

“Processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party, except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which requires protection of personal data, in particular where the data subject is a child.”

[49]The case law on article 6(1)(f)’s predecessor established that it required three questions to be answered, which we consider are still appropriate if reworded as follows: 1. Is the data controller or a third party pursuing a legitimate interest or interests? 2. Is the processing involved necessary for the purposes of those interests? 3. Are the above interests overridden by the interests or fundamental rights and freedoms of the data subject?

[50]Lady Hale said the following in South Lanarkshire Council v Scottish Information Commissioner

[2013] 1 WLR 2421

about article 6(f)’s slightly differently worded predecessor: “27. ... It is well established in community law that, at least in the context of justification rather than derogation, ‘necessary’ means ‘reasonably’ rather than absolutely or strictly necessary .... The proposition advanced by Advocate General Poiares Maduro in Huber is uncontroversial: necessity is well established in community law as part of the proportionality test. A measure which interferes with a right protected by community law must be the least restrictive for the achievement of a legitimate aim. Indeed, in ordinary language we would understand that a measure would not be necessary if the legitimate aim could be achieved by something less. ... “

[51]Section 40(2) is an absolute exemption and therefore the separate public interest balancing test under FOIA does not apply.

The role of the tribunal

[52]The tribunal’s remit is governed by section 58 FOIA. This requires the tribunal to consider whether the decision made by the Commissioner is in accordance with the law or, where the Commissioner’s decision involved exercising discretion, whether he should have exercised it differently. The tribunal may receive evidence that was not before the Commissioner and may make different findings of fact from the Commissioner.

Issues

[53]The issues for the tribunal to determine are: 53.1. Is the requested information personal data? 53.2. Is the data controller or a third party pursuing a legitimate interest or interests? 53.3. Is the processing involved necessary for the purposes of those interests? 53.4. Are the above interests overridden by the interests or fundamental rights and freedoms of the data subject? 53.5. Did the EHRC comply with section 16? Discussion and conclusions Scope of the May Request

Discussion and conclusions

[54]Whilst we are considering the matter afresh, we note that the Upper Tribunal in this appeal concluded, at paragraph 31, that it was ‘reasonably clear’ that the information requested in the May Request included the following information: 54.1. Median pay for men and women, 54.2. Mean pay for men and women, and 54.3. The ‘gender pay gap’ analysis on both median and mean pay bases (expressed in terms of percentages).

[55]Given the matters set out in paragraphs 7-9 above we agree. We find that the request was for the results of the gender pay gap analysis and equal pay analysis, i.e. the end data that is required to be published, rather than for all the information used to perform that calculation. We have considered each element of the information separately. If the information released relates to an identifiable individual, then it is personal data.

[56]Broken down into each element, we find that that the information within the scope of the May request consists of: 56.1. Median pay for men 56.2. Median pay for women 56.3. Mean pay for men 56.4. Mean pay for women 56.5. The percentage difference between median pay for women and median pay for men. 56.6. The percentage difference between mean pay for women and mean pay for men.

Findings of fact

[57]We heard evidence from Ms Powell and make the following findings of fact on the balance of probabilities.

[58]We find that Ms Powell had a good idea of which pay grade her fellow employees were in. She stated that she believed she knew which pay grade they were in. She had a ‘vague understanding’ of where other employees sat in their pay grade. Ms Powell did not know the actual salary of any of her fellow employees.

[59]Ms Powell knew, because the EHRC had told her, the total spending on salaries and the number of employees in the office.

[60]Ms Powell knew that there was one other female in her pay grade. In relation to the other female in her pay grade, her understanding was that the other female was probably at the top of the pay grade, based on her understanding of a previous pay harmonisation exercise. Ms Powell was at the bottom of her pay grade.

[61]She knew that the median female salary might not be her salary.

Is the requested information personal data?

[62]It is clear that the information requested in the March Request is personal data. It is the salaries of named individuals.

[63]In relation to the May Request we find that the request was for the results of the gender pay gap analysis and equal pay analysis, i.e. the end data that is required to be published, rather than for all the information used to perform that calculation. We have considered each element of the information separately. If the information released relates to an identifiable individual, then it is personal data.

Female median pay

[64]A median salary is worked out by arranging the salaries for a group of employees into a list in descending order. If there are an odd number of employees, the median salary is the salary of the employee in the middle of that list. If there are an even number of employees, the median salary is the average of the two middle salaries.

[65]In the Birmingham Office, there is an odd number of women. The median female salary will therefore the actual salary of the middle woman.

[66]The appellant knows that she is one of two women in her pay grade. It would be clear to her, from her knowledge of people’s pay grades that either her salary or the salary of the other female in her pay grade was the median salary. Given the small number of women we infer that other woman in Ms Powell’s pay grade would also know this information.

[67]Because an individual knows their own salary, if the median salary was provided to the world, then both individuals would know, by a process of elimination, which one of them was paid the median salary. Whoever was not paid the median salary would therefore know the exact salary of the other woman.

[68]On that basis we find that the median female salary is personal data. It relates to an individual and that individual is identifiable by certainly one other person. Given that there are only 3 women in the office, it is quite likely that other individuals in the office would also be able to identify the female who received the median salary, but we do not need to make this finding for the purposes of our decision.

Male median pay

[69]Part of the requested information is the percentage difference between male median pay and female median pay. If male median pay was released then female median pay could be calculated using this percentage. The female median pay reveals personal information about an identifiable female employee in the way explained above. On that basis we conclude that this is also personal data.

[70]If the male median pay is released on its own, we do not accept that this would reveal personal data. There is no evidence that anyone knew who the man receiving the median salary was. Given that there are only 5 men in the office, it is quite likely that the employees would be able to speculate, possibly accurately, as to which man received the median salary, but there is no evidence that anyone knew who this man was and the speculation might be wrong.

[71]Applying the principles set out by the Upper Tribunal in Spivackin paragraph 22,we do not accept that it is possible to identify a specific individual male on the basis of the median male salary alone. It is not sufficient to say that there is a pool that contains or may contain a person covered by the data, nor is it sufficient that it is reasonably likely that someone is covered by the data. Linking any specific male to the data does not rely solely on the data disclosed and other data available by reasonable means. It involves speculation.

Mean pay for women

[72]Ms Powell already has a good idea which pay grade her fellow employees are in. She therefore already knows what range their salary falls in. For example, she knows that the other female in her pay grade has a salary somewhere between £29,980-£34274.28.

[73]If Ms Powell knew the exact salary of one of the other female employees, she would be able to use the mean salary and her own salary to work out the exact salary of the third female employee. However she does not know the exact salary of any other employees.

[74]The EHRC demonstrated how one of the female employees who knew which pay grade the other female employees were in and using their own salary, could calculate a salary range for a fellow female employee that was narrower than the pay grade salary range.

[75]Assume that Mrs A works with Mrs B and Mrs C. They are the only females in the office. Mrs A and Mrs B are in the same pay grade which has a salary scale of £10,000-£15,000. Mrs C is in a lower pay grade with a salary scale of £5,000-£10,000. Mrs A finds out that the mean female pay for that office is £10,500 The total salary of the three females must therefore be £31,500.

[76]Mrs A knows that she is paid £11,000. She knows that Mrs C is paid between £5000 and £10000. She can therefore work out the lowest possible salary for Mrs B as follows: 11000 + 10000 + Mrs B’s salary = 31500 31500 - 21000 = 10,500

[77]Mrs A can also work out the highest possible salary for Mrs B as follows: 11000 + 5000 + Mrs B’s salary = 31500 31500 – 16000 = 15,500 (capped at 15,000 because Mrs A knows the pay grade)

[78]Thus Mrs A finds out that Mrs C’s salary is between £10,500 and £15,000. The calculation produces a salary range for an identifiable individual that is narrower than the pay scale.

[79]Different salaries and different pay scales produce a narrower or a broader range. For example, the hypothetical salaries and actual pay scales used by the EHRC in their worked example produced a salary range for an identifiable individual of between £31,711.82 and £34,274.28. This is a narrower range of £2,562.46 when compared to the salary bracket of £4,294.28.

[80]As Mr. Willis pointed out in submissions, if the person carrying out the calculation was either the appellant or the other female employee in her pay grade, this might allow them to work out whether the other woman in their pay grade was paid more or less than them.

[81]We accept that the fact that an individual’s salary falls within a particular bracket, is information that relates to them. That individual is identifiable by the other two females in the office.

Mean pay for men

[82]It was not argued that a similar calculation could be carried out in relation to the men, presumably because there are more than 3 men. However if the mean salary for men is released along with the percentage difference between female and male mean salaries, that could be used to calculate the mean female salary and would accordingly reveal identifiable personal data in relation to one of the female employees to the two other female employees in the office in the way set out above. Mean pay gap percentage

[83]If the mean salary for men and not women was released it would enable the mean salary for women to be calculated which does reveal the personal data of an identifiable individual to the two other female employees. If the mean salary for men and women is released as well, the mean pay gap percentage would not add anything to the personal data released by those other figures.

[84]If the mean pay gap percentage is released without the mean salary of men and women, it would be possible to calculate the mean salary for women and the mean salary for men because the EHRC have released the total salary figure and the number of male and female employees are known. This would accordingly release the personal data of a female employee to the two other female employees in the way described above.

Median pay gap percentage

[85]It has not been explained to us how the median pay gap percentage, if released without any other information, would reveal personal data. We cannot see any mechanism by which this would be possible even in a small office. The appellant and the other female in her grade do not know which one of them is paid the median salary. The appellant might assume that hers was the median salary, because her understanding was that the other female was probably at the top of the pay grade, but she would not know for certain that she was being paid the median salary. It is speculation. The other female in her pay bracket might make the same assumption.

[86]Making an assumption or speculating that she was being paid the median female salary, the appellant could then apply the median pay gap percentage to her salary and provide a speculative answer as to what the median male pay was.

[87]Given that there are only 5 men in the office, it is quite likely that she would also be able to speculate, possibly accurately, as to which man received the median salary, but there is no evidence that she knew who this man was and she might be wrong.

[88]Applying the principles set out by the Upper Tribunal at paragraph 22 of Spivack,we do not accept that it is possible to identify a specific individual male relying on the data available. It is not sufficient to say that there is a pool of individuals that contains or may contain a person covered by the data, nor is it sufficient that it is reasonably likely that someone is covered by the data. Linking any specific male to the data does not rely solely on the data disclosed and other data available by reasonable means. It involves speculation.

[89]The median pay gap percentage, if released with the male median pay and not the female median would enable the female median pay to be calculated, which would release personal data for the reasons set out above. The percentage of men and women in each hourly pay quarter.

[90]Mr. Willis submitted that it was not possible to carry out this calculation in such a small office.

[91]The method for calculating the percentage of men and women in each hourly pay quarter is as follows: 91.1. Sort your full-pay relevant employees in a list according to hourly rate of pay, in order of highest to lowest paid. 91.2. Divide this into four quarters, with an equal number of employees in each section. These quarters will be the: upper hourly pay quarter, upper-middle hourly pay quarter, lower-middle hourly pay quarter and lower hourly pay quarter. If there are employees on the same hourly pay that overlap between hourly pay quarters, make sure that men and women are split as evenly as possible across the hourly pay quarters either side of the overlap. 91.3. Work out the percentage of men and women in each of the four parts.

[92]Whilst in our view it is not impossible to work out this percentage for a group of 8 employees, there will only be two employees in each pay quarter. Each pay quarter would therefore either be made up of(a) 50% female and 50% male employees; or

(b) 100% female or male employees.

[93]There is only a requirement to publish the percentage of women and men in each pay quartile for the purpose of gender pay gap reporting. There is no requirement to publish any information on what the hourly pay is of any employee or group.

[94]The Commissioner argues in his response to the Upper Tribunal proceedings that the appellant would be able to calculate which individual corresponds to which pay quarter. We agree. The appellant knows that one of the women was in a lower pay scale. She was at the bottom of her pay grade, and she understood that the other woman in her pay grade was at the top of the grade. She would therefore at least be able to identify the hourly pay quarter of the other women. We find that the pay quarter information is accordingly information relating to an identifiable individual.

Legitimate interest

[95]The EHRC has a statutory duty under section 8 of the Equality Act 2006 to work towards the elimination of unlawful discrimination and to enforce the Equality Act 2010, including the equal pay provisions. The EHRC’s statutory role includes working towards reducing the gender pay gap and enforcing employers’ obligations to publish gender pay gap information.

[96]We find that there is a legitimate interest in knowing the gender pay gap analysis and the equal pay analysis of the EHRC’s workforce. The EHRC is the statutory body with responsibility for encouraging equality and diversity. One of its purposes is eliminating discrimination. It is responsible for enforcing equal pay legislation and the requirement to publish gender pay gap information.

[97]Whilst there is a stronger legitimate interest in the pay gap of the entire workforce, given the EHRC’s statutory function we accept that there is also a legitimate interest in transparency in relation to the gender pay gap analysis and the equal pay analysis in relation to any particular regional office of the EHRC, including the Birmingham office.

[98]We accept that the appellant, as an employee of the relevant ‘watchdog’ has a private legitimate interest in knowing what the gender pay gap analysis and the equal pay analysis was in the office in which she worked. We accept that she also has a private legitimate interest in finding out whether she might have had a potential equal pay claim against her employer, in order to explore the possibility of an action against her union solicitor.

Reasonable necessity

[99]We have considered whether the disclosure of the requested information is reasonably necessary for the purposes of the identified legitimate interests. Disclosure must be more than desirable, but less than indispensable or an absolute necessity. Disclosure must be the least intrusive means of achieving the legitimate aim in question, because it would not be reasonably necessary if it could be achieved by anything less. We must consider whether the legitimate aim could be achieved by means that interfere less with the privacy of the data subjects.

[100]We deal firstly with the appellant’s private legitimate interest in finding out if she had a potential equal pay claim against her employer. The appellant believes that she was negligently advised to settle her claim and wishes to explore the possibility of an action against the union solicitor.

[101]We do not accept that disclosure under FOIA to the world is reasonably necessary for the purposes of exploring the possibilities of bringing a legal action. Civil proceedings contain their own provisions for disclosure to a party, including pre-action disclosure, and this is a much less intrusive means of achieving the legitimate aim in question.

[102]Save for the hourly pay quarter information, dealt with below, we accept that disclosure of the information requested in the May Request, i.e. the gender pay gap analysis and the equal pay analysis, is reasonably necessary for the purposes of the public and private legitimate interests in transparency in relation to the gender pay gap analysis and the equal pay analysis in relation to the Birmingham office of the EHRC. We accept that the median and mean figures are more meaningful when applied to a larger body of employees, and we have built that into our balancing exercise below, but we find that these figures still provide some insight into the gender pay gap in the Birmingham Office. The identified legitimate interest cannot be achieved by other means.

[103]In relation to the hourly pay quarter percentage figures, in our view these figures are meaningless when applied to only 8 employees. This method provides no real insight into the gender pay gap when each quarter only contains two employees. For this reason we do not accept that disclosure of this particular part of the information is reasonably necessary for the purposes of the above information. This part of the information is therefore exempt under section 40(2).

[104]We do not accept that disclosure of the information requested in the March request, i.e. details from the Birmingham EHRC office of employees in post by grade/name, job title and salary at 9th February 2017, is necessary for the purposes of any of the legitimate interests that we have identified. The release of the gender pay gap analysis and equal pay analysis is a less intrusive way of achieving those legitimate aims. This information is therefore exempt under section 40(2). Are the above interests overridden by the interests or fundamental rights and freedoms of the data subject? Median pay for women and median pay for men

[105]An individual’s exact salary from 2017 is the personal data that would be revealed by release of the median pay for women or the median pay for men combined with the percentage median pay gap. In our view disclosure of an employee’s exact salary would not have been within their reasonable expectations. There is no evidence on harm or distress but we accept that an employee might be upset if their exact salary was released without their consent even 2 years after the event. In our view, if the median pay gap percentage is released the median pay itself does not add significantly to the legitimate interest in transparency. For those reasons we conclude in relation to the median pay for men and women that the legitimate interests are overridden by the fundamental rights or freedoms of the data subject.

[106]For those reasons we find that the respondent was entitled to withhold the median pay for men and the median pay for women under section 40(2).

Median percentage pay gap

[107]Without the median pay itself, we have found that the median percentage pay gap reveals no personal data of an identifiable individual. Accordingly, we find that the respondent is not entitled to withhold this information under section 40(2). Mean pay for men/mean pay for women/

Mean pay gap percentage

[108]This information reveals the same personal data: a narrower salary bracket than is already known for one of the female employees. Only the two other female employees who already know the salary range of the other employees would be able to carry out the calculation set out above to reach a narrower salary range. We have accepted that in certain circumstances this might allow one of the female employees to identify if the other employee was paid more or less than them.

[109]We accept that it would probably be outside the reasonable expectation of an individual that a narrower bracket than the usual salary bracket would be made public. Further we accept that the other female employee would not reasonably expect the appellant to be, in effect, told whether she was being paid more or less than the appellant. However, there are a number of factors that reduce the impact of the release in our view.

[110]First, this is not exact salary information. In our view there is a greater expectation of privacy in relation to a person’s exact salary compared with a range. Particularly in public sector employment, a person’s salary range is usually evident from their job title or job role and therefore publicly known at least amongst fellow employees. The narrowest pay scale at the EHRC in Birmingham at the relevant time was about £2,700, the broadest about £6,800. In the worked example given by the EHRC, the revealed range was £2,732.52 (once capped by the know salary scale bracket). This is about the same salary range as the narrowest pay scale, albeit that is only an example, so the actual range might be smaller.

[111]Second, the calculation put forward by the EHRC can only be carried out by one of the three females in the office and only if they already know which pay grade the other female employees are in. In our view there is unlikely to be any significant upset or distress caused by revealing a narrower bracket to two fellow employees who are already aware of the individual’s pay scale. Although Freedom of Information disclosure is to the world, when considering the level of potential distress or harassment we take account of the fact the only individuals who can, in practice, calculate the personal information are those two other employees. We do accept that there is nothing in theory to prevent those employees from circulating or publishing those pay ranges, or indeed the already known pay grades more widely, but in our view that is, in reality, unlikely.

[112]Third this information relates to a salary range from February 2017. The Birmingham office was closed in 2017. The request was responded to in August 2019. By that date the relevant individuals no longer worked in the Birmingham office and the information related to salaries from 2.5 years earlier. The impact of its release is therefore likely to be less than if it had been released in 2017.

[113]We do accept that the appellant and the other female employee in her pay grade might be able to work out if the other was paid more or less than they were. Given that the appellant has publicly stated where she sits in her pay grade and has set out her and, we infer, others’ understanding of the effect of the previous pay harmonisation exercise, we would not expect this to be news to anyone, and therefore would not expect its release to cause any significant upset, taking into account in addition the age of the information and the fact that the Birmingham Office has closed.

[114]For all those reasons, we do not accept that release has the potential for causing any significant harm or distress and we find that the prejudice to the rights and interests of the individual(s) is fairly limited.

[115]In our view, the specific statutory duties and the role of the EHRC means there is an extremely strong legitimate interest in transparency in relation to the gender pay analysis and equal pay analysis in any particular regional office of the EHRC, including the Birmingham office. This legitimate interest is reduced in this appeal for a number of reasons.

[116]First, there were a very small number of employees in the Birmingham office. This reduces the value of the results of a gender pay analysis and an equal pay analysis because they are intended to operate on large figures. Individual employees’ salaries can skew the results in such a small sample. Given the EHRCs pay structure, we think that there is unlikely to be an unusually high or low individual salary. Overall, we think there remains a reasonably strong legitimate interest in knowing the mean pay gap percentage even in such a small office.

[117]Second, the data is historic. At the relevant data, the Birmingham office had closed and the data was 2.5 years old. That reduces the legitimate interest in transparency even in relation to the EHRC. In our view although the legitimate interest is reduced to some extent by the age of the data, the relevant legitimate interest remains in relation to what the EHRC was paying men and women in its Birmingham office approximately two years earlier.

[118]Taking those matters into account, we find that there remains a moderate legitimate interest in transparency in relation to the gender pay gap analysis and the equal pay analysis in relation to the Birmingham office of the EHRC.

[119]Looked at as a whole, we find that this moderate legitimate interest is not overridden by the fairly limited prejudice to the rights of freedoms of the individual(s) concerned.

[120]In those circumstances we find that the EHRC was not entitled to rely on section 40(2) to withhold the mean pay for mean, the mean pay for women or the mean pay gap percentage.

Section 16

[121]The Upper Tribunal identified at para 41 that our findings in relation to whether any piece of the requested information was not personal data; and so, not exempt under s40(2) might affect consideration of section 16.

[122]In our view, without the need to suggest how the appellant could reframe her request, the public authority was obliged to release any information that fell within the scope of the request that it was not entitled to withhold, and it could withhold any parts which were exempt under section 40(2).

[123]In the light of our findings above, the public authority should accordingly have withheld the information requested in the March request, released the median pay gap percentage, the mean pay for men and women and the mean pay gap percentage and withheld the median pay and the quartile analysis.

[124]In those circumstances we do not think there was any need to provide any further advice or assistance under section 16.

Summary of decision

[125]For the above reasons we conclude that the EHRC was entitled to withhold the following information under section 40(2) FOIA: 125.1. The withheld information requested in the March request. 125.2. The percentage of men and women in each hourly pay quarter. 125.3. The median pay of men. 125.4. The median pay of women.

[126]We conclude that the EHRC was not entitled to withhold the following information under section 40(2) FOIA: 126.1. Mean pay for men. 126.2. Mean pay for women. 126.3. The percentage difference between median pay for women and median pay for men. 126.4. The percentage difference between mean pay for women and mean pay for men.

[127]There was no breach of section 16. Signed Sophie Buckley Date: 18 March 2024 Judge of the First-tier Tribunal