User guide to diversity of the judiciary statistics

Question 1

1. Introduction

Accepted Answer

Statistics presenting data on diversity for current judicial office holders, for judicial applications and appointments and within the legal professions which provide the eligible pool of candidates for most judicial roles were first published together in a combined report in September 2020, and will now be published annually. This guide accompanies the second edition of the combined statistical report, published in July 2021.

These statistics are designated as Official Statistics, indicating that they are fit for purpose and are produced in compliance with the Code of Practice for Statistics, in accordance with the Statistics and Registration Service Act 2007.

This designation can be broadly interpreted to mean that the statistics meet identified user needs, are well explained and readily accessible, are produced according to sound methods; and are managed impartially and objectively in the public interest.

This user guide provides a brief background to the judiciary and also includes information on:

users and uses of the statistics
data sources and methodology
the quality of the statistics
changes made to the statistics and plans for future development
links to other related statistics
a glossary of terminology
other explanatory notes

Question 2

2. Background to judicial diversity statistics

Accepted Answer

These statistics bring together information about the diversity for the current judiciary in England and Wales, during selection for judicial roles, and in the legal professions which provide the pool of eligible candidates for most posts requiring legal experience.

While the focus of the statistics is on diversity, they also provide a snapshot count of the number of judicial office holders.

2.1 The judiciary in England and Wales

The statistics provide an overview of the diversity of appointed court judges, tribunal judges, non-legal members of tribunals and magistrates. Figures are published on an annual basis, taking a snapshot of the staffing position as at 1 April of each year. An explanation of judicial roles is available from the judiciary website: www.judiciary.gov.uk/about-the-judiciary/who-are-the-judiciary/judicial-roles/

The Judicial Career Progression chart provides an overview of progression through the judiciary in England and Wales : www.judiciary.gov.uk/about-the-judiciary/judges-career-paths/judicial-career-progression-chart/

2.2 Judicial appointments

The Constitutional Reform Act 2005 (CRA) enshrined in law the independence of the judiciary and changed the way judges are appointed. As a result of the Act, the Judicial Appointments Commission (JAC) was set up in April 2006 to make the appointments process clearer and more accountable. Under the CRA, the JAC’s statutory duties are to:

select applicants solely on merit
select only those with good character
encourage a diverse range of applicants

As part of its diversity strategy, the JAC publishes the diversity profile of applicants at application, shortlisting and recommendation stages.

Shortlisting: Shortlisting is the process used by the JAC to determine who is invited to attend a selection day. The main tools used, either together or separately, are currently:

An online qualifying test, more likely to be used when the volume of applications is large, or
A paper sift, which considers applicants’ self-assessment and other information (for example, independent assessments) and is more likely to be used for those exercises with a smaller number of applicants.

These tools may be used in conjunction with other shortlisting tools, such as a telephone assessment or written scenario test. The same types of selection tool are used for both legal and non-legal exercises. On rare occasions, when applicant numbers are very low, no shortlisting process is undertaken and all eligible applicants are invited to attend a selection day, which will involve an interview and may also involve situational questioning, a leadership presentation or role play.

Recommendations: Before making recommendations to the Appropriate Authority, the Commissioners of the JAC, sitting as the Selection and Character Committee, first assure themselves that applicants are ‘of good character’. The Selection and Character Committee then makes selection decisions based on the panel’s assessment of all the available evidence, and the result of statutory consultation with the judiciary. The Lord Chancellor, Lord Chief Justice or Senior President of Tribunals can reject a recommendation, although do so only on a very exceptional basis.

The JAC makes recommendations under section 87 of the Constitutional Reform Act 2005 (CRA). Recommendations are for a confirmed vacancy. If accepted by the Appropriate Authority, they are guaranteed to be offered appointment.

The JAC may also be asked to identify persons suitable for later selection under section 94 of the CRA (also referred to as ‘recommended to a list’). Those identified by the JAC are regarded as suitable for future appointment to specific roles if, and when, an appropriate vacancy arises. Those applicants are not guaranteed an offer of appointment. Applicants recommended under section 87 and 94 CRA are reported separately in the published statistical tables. In addition, if a vacancy subsequently becomes available for a post for which a selection exercise has recently been carried out, the JAC can make an additional recommendation using the results of that recent exercise. This is the case even if there are no applicants identified following a section 94 exercise for the specific location and/or jurisdiction.

Senior appointments: The JAC is responsible for running selection exercises for posts up to and including the High Court. It also has statutory responsibilities to respond to requests from the Lord Chancellor to convene panels that recommend applicants for appointment to other senior posts. These include the Lord Chief Justice, Heads of Division, and Lord Justices of Appeal. The JAC provides the secretariat for these exercises and, in line with statute, at least 2 JAC Commissioners sit on each 5-member panel.

While senior appointment selection panels are required to determine their own processes, selection exercises may include an application (form or letter), independent assessments, self-assessment, non-statutory consultation (seeking feedback on applicants from the senior judiciary and others), a sift and selection interviews.

From 2015-16, information on senior exercises has been included in these statistics (although figures are shown separately from the overall totals).

Quality assurance in selection: The JAC uses quality assurance checks throughout the selection process to ensure proper procedures are followed, standards are maintained and all stages of selection are free from bias. This includes:

reviewing selection exercise materials, and observing dry-runs of role plays and interviews
monitoring the progression of candidate groups at key stages in the selection process
carrying out equality impact assessments on all significant changes to the selection process and
making reasonable adjustments for applicants who need them

2.3 Legal professions

To become a judge, some degree of legal experience is required. This publication includes data for barristers, solicitors and legal executives. While it is not essential to be a member of one of these professions to apply for judicial roles, in practice most of those who apply will have a background in at least one of these professions.

Solicitors provide general legal advice on a variety of issues, and some may provide specialist advice and represent their clients in court. They can work together with others in private practice or in government departments or commercial businesses. The professional body representing solicitors is the Law Society, and the profession is regulated by the Solicitors Regulation Authority(SRA). Solicitors wishing to carry out reserved legal activities must have in force a practising certificate issued by the SRA (or a relevant exemption). Individuals who have qualified as a solicitor and are on the roll of solicitors do not need to hold a practising certificate to apply for judicial office.
Legal executives (or Chartered Legal Executives) carry out similar work to solicitors, but with a different route to qualification. The Chartered Institute of Legal Executives (CILEX) is the professional body for legal executives.
Barristers are specialists in certain legal fields that solicitors can instruct on behalf of their client to appear in court. The professional body for barristers is the Bar Council, and the regulatory body is the Bar Standards Board. In order to practise as a barrister, an annual practising certificate is required to carry out reserved legal activities. Barristers can be broken down into Queen’s Counsel (QC) barristers and junior barristers. Queen’s Counsel is an office, conferred by the Crown, that is recognised by courts and is awarded for excellence in advocacy (although guidelines for award note it is unlikely that an applicant will have acquired the necessary skills and expertise for appointment without extensive experience in legal practice).

The overall oversight regulator of legal services in England and Wales is the Legal Services Board, which is independent from both the legal professions and government.

Question 3

3. Data sources, coverage and definitions

Accepted Answer

3.1 Data sources

Judicial office holders

Data are a snapshot taken from the judicial HR database, e-HR. This contains details of each judicial office holder (judges, non-legal members and magistrates), including diversity characteristics.

Up to 2019-20, diversity characteristics, including ethnicity, were recorded as self-declared by judicial office holders at time of entry into the judiciary and not changed unless the information was specifically provided to judicial HR teams. However, from 2019/20, judges in courts and tribunals have been able to enter and edit their own diversity information, allowing them to ensure it is up to date and accurate. Additional diversity information is now being collected, though declaration rates are not yet sufficient for publication.

Judges are given the option to decide not to declare any of their diversity information at any point via the option to ‘Prefer not to say’ and thus opt out of providing us their data, ensuring compliance with the General Data Protection Regulation. This impacts on the presentation of diversity information within these statistics^{[footnote 1]}.

Judicial appointments

In January 2020, the JAC replaced its legacy administrative data system JARS (Judicial Appointments Recruitment System) with a new digital platform. Both systems operated in tandem for a period to allow exercises that launched on JARS to draw to a conclusion. For exercises that launched prior to January 2020 and concluded in 2020-21, data is taken from JARS. For all exercises that launched January 2020 onwards, data is taken from the new platform. Both systems store candidate data which is subject to specific legislative provisions set out in the CRA, the Data Protection Act 2018 and Freedom of Information Act 2000. User access is strictly controlled and trail logs are kept for security checks and audit purposes.

Data from the JAC’s diversity monitoring form (which is part of the broader application form) are used to produce reports and to support statistical analysis. Completing the diversity monitoring form is not compulsory and not all applicants make diversity declarations on some or all items within the form. The form includes questions regarding gender, ethnicity, professional legal background, disability, age, socio-economic background, sexual orientation and religious belief.

Within the applications system on JARS, individuals were free to update their diversity data. For these statistics, diversity data are as captured at the point of download from the system (usually close to the date at which the exercise was completed following recommendations for appointment). Diversity data on the new platform cannot be amended by candidates following submission of their application.

Legal professions

Data for the legal professions are taken from administrative data systems used for the purposes of managing membership lists and certification. Figures for inclusion in this publication are provided in aggregate format and published in the form supplied. They represent a snapshot of the position as at 1 April 2021.

Barristers: Data are provided by the Bar Standards Board (BSB), and cover all practising barristers i.e. those that hold a practising certificate.

Solicitors: Data are provided by the Solicitors Regulation Authority (SRA), and cover practising solicitors (those holding practising certificates) excluding Registered European Lawyers (RELs) and Registered Foreign Lawyers (RFLs)^{[footnote 2]}. The practising population is different from the regulated population which includes all solicitors on the roll i.e. solicitors who qualified but do not have a current practising certificate and is therefore a larger group^{[footnote 3]}.

Chartered Legal Executives: Data are provided by CILEX from their membership database. Figures relate to all those who have achieved fellowship of CILEX, excluding students, affiliates, associates and graduates. Data are collected on initial registration and through a private member portal where members can opt to complete their equality and diversity data.

3.2 Coverage and definitions

Judicial office holders

Coverage: These statistics broadly cover the judiciary in England and Wales^{[footnote 4]}. For courts and magistrates, figures cover only England and Wales. Tribunals figures include all tribunals administered by HMCTS and Welsh Tribunals not administered by HMCTS. This includes Employment Tribunal Scotland, in addition to Tribunals in England and Wales. Tribunals that are the responsibility of the devolved Welsh Government are not included.

Count of appointments: The focus of this bulletin is diversity, and accordingly the figures within the bulletin relate to individuals, and not to the posts held.

Where a judge holds more than one appointment^{[footnote 5]}, the statistics are compiled for the appointment considered to be their primary appointment, i.e. the appointment they hold most of the time. Figures are on a headcount basis, and do not reflect the full-time equivalent (FTE) value of part-time salaried judicial post holders (to do so would be to understate representation among part-time individuals). Similarly, for those in fee-paid roles, figures count individuals, not posts held nor appearances in court.
All figures relate to the position as at 1 April of the relevant year.

New appointments and leavers: From 2018-19 onwards, statistics have been published for entrants to, and leavers from, the judiciary.

New appointments include those who started their first appointment, have been promoted from fee-paid to salaried or have had a promotion from a salaried post to a higher salaried post.Judges or members changing appointment, such as extension, change of jurisdiction or returning to sit in retirement are excluded. These are further broken down into:
- New entrants (i.e. those taking up their first judicial role). Judicial office holders may take up a new appointment while in office; new entrants are counted as those starting a new appointment who did not hold another appointment at the start of the financial year. Although judicial office holders can have multiple appointments, they are counted on a head count basis and so will only be counted once by their primary appointment.
- Promotions from one judicial role that happened in the last financial year, either from fee-paid to salaried or from a salaried post to a higher salaried post. In the tables presenting diversity breakdowns for promotions, figures for fee-paid posts are 0.
Leavers count those leaving the judiciary for any reason (including retirement, resignation, death in service and removal by the Lord Chancellor). As judicial office holders can hold more than one appointment, they are only counted when they leave their primary appointment and hold no other appointments.

Contract type: The bulletin provides breakdowns of fee-paid, salaried and salaried part-time judges and non-legal members of tribunals^{[footnote 6]}. For both courts and tribunals, fee-paid positions are paid according to the number of sittings or days worked. The number of sitting days varies depending on the type of appointment, and will generally be at least 15 days a year. All figures exclude those who are sitting in retirement as a fee-paid judge - improvements have been made from 2019 onwards to ensure exclusion of all those sitting in retirement.

Judicial appointments

Exercises included: The statistics cover all selection exercises run by the Judicial Appointments Commission, although any which are run for the Welsh Government (under the Government of Wales Act) are excluded, and figures for senior selections (for Court of Appeal and above) are shown separately.

The JAC makes recommendations for appointment to one of 3 Appropriate Authorities (the Lord Chancellor, Lord Chief Justice or Senior President of Tribunals). For the purpose of presenting information in the Official Statistics bulletin, the date of the report to the Appropriate Authority marks the point at which the JAC’s involvement in the selection exercise is considered to have ended. Those exercises for which recommendations have been considered by the Appropriate Authority within the financial year are included within the annual statistics^{[footnote 7]}.

In the event that a recommended candidate approved by the Appropriate Authority subsequently turns down the offer of a post, further recommendations may be made to fill the vacancy request. Where additional recommendations come from the same pool of candidates who applied initially and are appointed within the financial year, these will also be reported in addition to those made previously. Should further recommendations be made after the end of the financial year, these will not be included within the annual statistics. However, in the event of 10 or more additional recommendations being made in any single selection exercise after the end of the financial year, the additional recommendations may be published in the following year’s statistics bulletin.

The bulletin presents information on the outcome of selection exercises by the date of the report to the Appropriate Authority. This has implications for revisions (see section on revisions).

Applications: In selection exercises prior to December 2012, applicants were screened to ensure they met the eligibility criteria when they first applied. Ineligible applicants did not continue through to the next stage of the selection process. For exercises that completed from October 2013, information regarding applicants relates to all those who applied for a particular post, regardless of eligibility. The number of applicants excluded because of eligibility concerns is generally low, largely confined to entry-level roles and should, in most cases, make little substantive difference. However some caution should be taken when comparing the profile of applicants in exercises carried out at different times for this reason.

Where there are vacancies for 2 or more posts which are run as a single selection exercise^{[footnote 8]} figures presented refer to individual applicants on a headcount basis, as opposed to the number of applications. Candidates may apply for both posts but would only participate in the exercise once. However, where the same person applies for selection through different selection exercises, each application is counted.

Recommendations: On rare occasions where a recommendation made by the Appropriate Authority is rejected, the JAC would make a further recommendation to the Appropriate Authority in line with legislation. If this occurred prior to the publication of the statistics they would be included in the published figures, unless timescales make inclusion impractical. If following publication, then any amendment to the published statistics would be considered a revision (see section on revisions).

Where a vacancy subsequently becomes available for a post for which a selection exercise has recently been carried out, any additional recommendations made prior to the end of the reporting year would be included within the published statistics (and, otherwise, would be included in the subsequent bulletin).

Eligible Pool: The eligible pool provides context for the diversity statistics of different selection exercises. It presents the gender, ethnicity and professional background of everyone who meets the formal eligibility criteria and certain additional selection criteria for a post. It should be noted that just because a candidate is included in the eligible pool, this does not mean they have a desire to apply for a given role, nor that they have the relevant talent and experience needed.

The data relating to the gender, ethnicity and professional background of the eligible pool is collated from data provided by the legal professions and the judiciary on the basis of the selection exercise eligibility criteria. With the exception of specialist posts, selection exercise eligibility criteria fall into 4 main categories:

statutory requirement of 5 years or more post qualification experience
statutory requirement of 7 years or more post qualification experience
statutory requirements of 5 or 7 or more years post qualification experience and subject to additional selection criteria. For salaried posts, additional criteria often include that the Lord Chancellor expects that individuals must normally have served as a judicial office holder for at least 2 years or have completed 30 sitting days in a fee-paid capacity
no statutory eligibility criteria (for non-legal posts)

For the first 2 categories (which are typically applied to fee-paid legal posts), data are supplied by the SRA, the BSB and CILEX as outlined below. Note that, based on advice from these professional bodies, the definitions used since 2019-20 differ from those used previously though the difference has been assessed as small and unlikely to materially impact on the resulting conclusions^{[footnote 9]}.

For the third category (which is typically applied to salaried legal posts), the data represent the information available on the composition of the pool of judicial office holders in England and Wales, taken from published statistics for the previous year.

Eligible pool figures are not calculated for non-legal posts, because there are no common statutory eligibility criteria, and are only calculated for characteristics where suitable data is available – currently gender, ethnicity and legal role.

The eligible pool figures presented in the statistics should be considered as best estimates of the pool, which while a good guide to the diversity of those theoretically eligible are unlikely to be precisely accurate. For example, there are some legal professionals who are not captured in the pool as defined here but who may be eligible to apply for certain judicial appointments.

Additionally, in order to estimate an overall eligible pool for legal exercises, individual exercises are weighted according to their share of recommendations rather than aggregating the eligible pools (as these vary greatly from around 1,000 to over 100,000, this would give disproportionate weight to exercises with large pools but few recommendations).

Legal professions

Coverage: As noted above, figures for barristers and solicitors are based on the practising population i.e. those holding practising certificates. Those who are not practising, and for solicitors, those who are not eligible for judicial appointment, are not included. Figures for Chartered Legal Executives are based on all those who have fellowship of CILEX, as only fellows of CILEX are eligible to apply for judicial appointments.

Post-qualification Experience: For barristers, years of post-qualification experience (PQE) are based on years after the completion of pupillage (and therefore becoming fully qualified) – this will include any years where a practising certificate was not held. This differs from the number of years from admission (or Call) to the Bar (‘years of Call’) which is an alternative measure and includes any time between Call and completion of pupillage.

For solicitors, years of PQE are based on the number of annual practising certificates held^{[footnote 10]} which is different to the number of years on roll as a practising certificate may not be held every year. Compared to the number of years since qualification, this approach better takes account of those who have had a career break. Note however that having a practising certificate does not necessarily mean that a solicitor was working as such during that year.

For Chartered Legal Executives, years of PQE relate to years since fellowship.

Seniority: Definitions of seniority for each profession are included in this publication, though it is important to note that these are not equivalent (so that comparisons between professions should be avoided):

Solicitors: solicitors at the lower level and partner at the higher level. The senior level of partner includes owners and managers of law firms. There is no equivalent way of identifying seniority for in-house solicitors who are all included in the lower level of solicitor.
Barristers: junior barristers (lower level) and Queens’s Counsel (QC) at the higher level.
Chartered Legal Executives: Chartered Legal Executives that are only fellows represent the lower level, while those who are also partners represent the higher level of seniority. There is no equivalent way of identifying seniority for in-house Chartered Legal Executives so they are all included in the lower seniority level.

3.3 Diversity characteristics

These statistics cover diversity characteristics where data is available and considered to be sufficiently robust (deemed to be when the declaration rate is at least 60%).

Gender

This report uses the term gender when describing the differences between men and women in the data. For the avoidance of doubt, this is based on self-declared data provided about sex (males and females) not data relating to gender identity^{[footnote 11]}.

For the judiciary and JAC data, unknowns represent those who preferred not to declare their information. Within the judiciary data, for a small number of records where the gender field was not populated, sex was imputed from other information (e.g. title or name).

Ethnicity

For the judiciary, judicial appointments and the legal professions, ethnicity is recorded by self-declaration on administrative systems on a non-mandatory basis, with the individual selecting the most appropriate category based on their own self-perception from the 2021 Census definitions (18 categories), or stating they choose not to declare.

For publication:

where possible, ethnicity is presented in data tables in aggregated form, using the 5+1 2011 Census definitions of White, Asian or Asian British, Black or Black British, Mixed, and Other Ethnic background (the ‘+1’ representing unknown, where individuals have either not responded or have stated their preference not to give a declaration)
in the report, or in data tables where the above approach could pose serious disclosure risks^{[footnote 12]}, those from the Black, Asian, Mixed and Other categories are grouped together (referred to as the Black, Asian and minority ethnic group) and compared to those in the White group. We acknowledge that grouping ethnicity in this way does not capture the different lived-experience of people within these groups. However, it is necessary, from a statistical perspective, to consider groups with sufficient numbers to make meaningful comparisons. The comparisons presented maintain consistency with previous iterations of this bulletin and is consistent with the approach used in other statistical publications.

Age

For the judiciary and judicial appointments data, age is calculated from date of birth. Data is grouped for publication, based on the distribution of ages among the judiciary. These are under 40, 40-49, 50-59 and 60 and over.

For the combined publication, the same age groupings have been used for judicial appointments and legal professions, which differ from the groupings used in previous JAC publications. Age is calculated as at the date of close of applications^{[footnote 13]}.

Professional background / legal role

For the judiciary, this refers to the legal profession which individuals had predominantly been employed within prior to taking up judicial office. This information is collected by self-declaration on a non-mandatory basis, reflecting the perception of the individuals themselves. Options include, but are not limited to, ‘solicitor’, ‘barrister’, ‘legal executive’ (Fellows of the Chartered Institute of Legal Executives) and ‘other’. Some ambiguity may also exist where individuals have had multiple prior roles (for example an individual that had been both a solicitor and a barrister would need to choose just one of these to enter, which is likely to be the most recent profession at the time of taking up judicial office and figures will not capture the prior professional experience not recorded in such cases).

For applicants for judicial appointments, legal role is presented using information from a question on the application form regarding the professional background of applicants. Options include, but are not limited to, ‘solicitor’, ‘barrister’, ‘salaried judicial office-holder’, ‘CILEx’ (Fellows of the Chartered Institute of Legal Executives) and ‘other’. Changes to the questions asked on professional background have enabled more comprehensive data on the full professional background of applicants, in addition to their current legal role. From 2019 onwards, legal role has been reported using 2 methodologies: applicants who have a current legal role of solicitor and applicants who have declared ever holding the role of solicitor (which is compared to ever barrister i.e. those currently holding a legal role of barrister and those who have declared holding the role of barrister in the past).

Further information on the development of the ‘ever’ legal role methodology is included in the 2018-19 JAC statistics publication.

Disability

Disability data is currently published only for JAC data. It is recorded as a binary characteristic of whether individuals have declared that they have, or do not have, a disability. Disability comes in many forms, and the impacts, needs and adjustments that may be required vary from individual to individual. In order to make statistical comparisons, reasonable numbers are required, and while simple binary categories do not reflect these differences, increasing granularity would substantially reduce analytical capability.

For the judiciary, disability information is collected on a non-mandatory basis by self-declaration, representing the perception of the individuals themselves, but it is not currently possible to differentiate between those with no disability, and those who have chosen not to provide the information. Disability status may change over time; an individual’s diversity information is as provided at point of entry unless they contact the relevant HR staff to update their disability information should their status change.

Sexual orientation

This is currently available only for JAC data, and recorded by asking applicants to declare whether they identify as a gay male, a gay female/lesbian, bisexual, or heterosexual. This is collated for statistical purposes into a binary category, grouping gay, lesbian and bisexual individuals together, in comparison to heterosexual individuals. We acknowledge that this simplifies the diversity of sexual orientation, and does not capture all identities. Consistent with the Equality Act 2010, this protected characteristic is distinct from and independent of gender identity. Accordingly, the familiar acronym LGBT is not appropriate for use when looking solely at sexual orientation as a protected characteristic. Sexual orientation figures are currently only presented aggregated across all selection exercises conducted within the financial year, due to the small numbers for individual exercises.

Religion or belief

This is also only available for JAC data. It is recorded with a range of options, including Buddhist, Christian, Hindu, Jewish, Muslim, Sikh, other religions and no religion. Religion is presented grouped across all the exercises reporting in a year. While declaration of religion continues to be lower than for other characteristics, it is above the threshold at which we would have concerns about representativeness and bias. It would not be statistically meaningful to present the full granularity of declared religions by selection exercise, given the very low numbers involved for many religions.

This is not a protected characteristic under the Equality Act 2010, however it is an important aspect of diversity, though currently only available for JAC data. The JAC added questions on social mobility in October 2015, in line with the Social Mobility and Child Poverty Commission’s recommendation that government and employers should collect data on the social background of new and existing post-holders. This information has been published for exercises from 2017-18 onwards. Information captures the type of school attended from ages 11-18, identifying whether applicants attended an independent/fee-paying school, went to a state school, or attended a school abroad. It also captures whether applicants attended university, whether either one or both parents went to university, neither went to university, or that the candidate did not attend university.

3.4 Summary of declaration rates for all diversity characteristics

The following table compares the current availability of data^{[footnote 14]}, and where appropriate, declaration rates, for the different diversity characteristics. Where declaration rates are below 60%, data is not included within the publication. It is anticipated that where these rates are currently above 40%, they may improve to be suitable for inclusion over a 3-4 timescale.

	Judiciary (Judicial HR data)	Judicial Appointments (JAC data)	Barristers (BSB data)	Solicitors (SRA data)	Chartered Legal Executives (CILEX data)
Gender	See published data tables	See published data tables	See published data tables	See published data tables	See published data tables
Age	See published data tables	See published data tables	See published data tables	See published data tables	See published data tables
Ethnicity	Around 90%	Around 90%	Around 90%	Around 80% overall, but under 50% for those with under 6 years PQE^{[footnote 15]}	Around 80%
Professional background	Over 90%	Above 90% for current legal role	Not applicable	Not applicable	Not applicable
Disability	Not currently possible to identify accurately	Around 90%	Around 50%	Around 80% (but not included in this publication)	Around 50%
Social mobility	Not recorded before 2019	Around 90%	50-55%	Recently started recording	Recently started recording
Religion or belief	Not recorded before 2019	Around 80%	50-55%	Under 40%	Recently started recording
Sexual Orientation	Not recorded before 2019	Around 90%	50-55%	Under 40%	Recently started recording
Post-qualification experience	Not applicable	Not currently published	Virtually 100%	Virtually 100%	Virtually 100%
Seniority	Not applicable	Not applicable	Virtually 100%	Virtually 100%	Virtually 100%
Practice area	Not applicable	Not applicable	Not included in this publication but available^{[footnote 16]}	Not considered reliable	Not included in this publication but available^{[footnote 17]}

3.5 Confidentiality and disclosure

Judiciary: There is no suppression of small numbers within the published data tables. Currently figures are presented for each characteristic, so that the risk of using some information to deduce other characteristics for an individual is considered minimal. Where figures are presented on the intersection of characteristics in Table 3.3, similar appointments have been grouped together in order to further minimise the risk of disclosing sensitive information. All individuals have the right to withhold their diversity information (via choosing ‘prefer not to say’).

At the senior levels, lists and biographies of judges are published on the Judicial Office website. While there are low numbers for some appointments in these statistics, which may allow senior judges to be identified, it is considered that this does not disclose information which is not already in the public domain.

Publication of these figures is in accordance with the privacy notice maintained by the judicial HR team.

Judicial appointments data: Exercises with fewer than 10 recommendations are aggregated so that applicants cannot be personally identified and are presented in the following groups:

Small court exercises (High Court and below)
Small tribunal exercises
Small exercises (the two groups above combined)
Senior judicial (above High Court) exercises

In larger exercises, there may be cases where certain breakdowns presented result in low numbers within that breakdown. It is considered that this is an acceptable risk to confidentiality; the applicants’ anonymity is still protected because the process of application itself is confidential and applicants can come from a wide range of areas within the legal profession and judiciary. Therefore, even if there is only one candidate with a particular characteristic it should not be possible to identify that person. By contrast, smaller exercises for more specialised posts sometimes accept applicants from a very narrow pool of eligibility, increasing the risk of a particular person being identified in the statistical results. This risk is mitigated by aggregating such exercises together.

In the accompanying statistical tables, percentages have been suppressed and replaced with an asterisk if they are based off a category containing fewer than 10 individuals. This is because percentages can be volatile for small groupings and are not considered reliable from a statistical perspective. RRIs have also been suppressed if either of their constituent recommendation rates contain numbers too small to be displayed.

Legal professions: Data included in this publication are aggregated and given the large numbers involved, the risk of identifying individuals from the published figures is considered to be very low.

Question 4

4. Methodology and calculations

Accepted Answer

4.1 Representation percentages and declaration rates

Representation percentages - the representation of particular groups within a diversity characteristic - are calculated excluding unknowns. This is the standard approach used across the Ministry of Justice, and more widely across government. They are presented for applicants and those recommended for appointment in judicial selection exercises, as well as judicial office holders and the legal professions.

Representation percentages allow comparison of the distribution of each diversity characteristic at different stages of the selection process, and for the current judiciary. This is particularly useful for judicial applications, giving a clear picture of the diversity of the pool of applicants, and how closely they represent the general population, and, where applicable, the eligible pool. It is also useful at the recommendation stage to illustrate the end result from a diversity perspective. However, representation among those recommended for appointment is the combined result of the representation among applicants and rates of success for each group in being recommended for appointment. Consideration of whether there is any significant difference in outcomes for a particular selection exercise can be viewed independently of initial level of applications by considering recommendation rates (below).

It is appropriate to consider this alongside the declaration rate – the proportion of the total number of individuals who gave a declaration for the characteristic. Only where the declaration rate is sufficiently high to mean that coverage of the characteristic is good (a widely adopted standard is a minimum threshold of 60%) will the representation percentage be presented. Where declaration rates fail to meet the minimum threshold of 60%, representation percentages are withheld as the level of uncertainty is too great for representation percentages to be meaningful. The higher the declaration rate, the better the coverage and the greater the certainty over the representation figures.

4.2 Recommendation rates (judicial appointments)

The recommendation rate is a simple measure of the proportion of applicants in one group that were recommended for appointment, derived from the total number of applicants as the denominator, and the number of those applicants that were recommended for appointment as the numerator. Direct comparison can be made between the recommendation rate for one group (such as women) compared to the recommendation rate for the other group (such as men) to determine, of those from each group that applied, whether there were equal outcomes for both groups (similar rates of recommendation for both groups), or whether there was a difference in outcomes, with one group being recommended at a statistically significant different rate to the other group.

While the recommendation rates for each group allow direct comparison within a characteristic, these rates are entirely dependent on both the number of applicants to an exercise and the number of vacancies being recruited for in the exercise. As such, while comparisons can be made within a single exercise, it would not necessarily be meaningful or valid to make simple comparisons across different exercises or across time, where the scale of applicants relative to the number of vacancies would differ. When considering recommendation rates, it is important to consider these alongside the representation percentages of applicants in the eligible pool, where available.

To make more meaningful comparisons across time or across different exercises requires a measure of difference in outcomes on a standard scale. This standardised measure of difference in outcomes is described as the Relative Rate Index (RRI).

4.3 Relative Rate Index (judicial appointments)

The Relative Rate Index, or RRI, gives a standardised measure of difference between groups, independent of variation in the overall rates of recommendation. However, when considering the RRI, it is important to consider, where available, the representation percentages of applicants relative to the eligible pool (or, if not available, the representation in the relevant working age population).

The use of RRIs for judicial appointments statistics was reviewed by MoJ statisticians, with recommendations published in a report which contains further background to the approach^{[footnote 18]}.

The RRI is the rate of recommendation for one group divided by the rate for another group within a diversity characteristic, thus creating a single standardised ratio measure of relative difference in outcomes between those 2 groups. This is most suited to binary comparisons (for example: women and men, Black, Asian and minority ethnic and White, disabled and non-disabled).

RRIs are also used to compare outcomes for solicitors relative to barristers, the particular comparison of interest for professional background, while noting this does not account for outcomes of those from other professional backgrounds. As interpretation of the RRI is to see this value as the comparison of outcomes of a group of interest (the group as the numerator in the calculation) to a baseline group (the group as the denominator), it is logical that the baseline should be the historically over-represented group.

An RRI value of 1 indicates no difference (that is, the recommendation rate of one group is precisely the same as the rate of the other group, so when dividing one by the other, a value of 1 is obtained). An RRI greater than 1 means the group of interest (women, Black, Asian and minority ethnic individuals, solicitors, people with disabilities) had a greater likelihood of being recommended for appointment than the baseline group, while an RRI of less than 1 indicates that the group of interest was less likely than the baseline to be recommended for appointment. For example, a gender RRI of 1.5 would be interpreted as women being 1.5 times as likely (50% more likely) to be recommended than men. Similarly, a gender RRI of 0.5 would be interpreted as women being half as likely (50% less likely) to be recommended than men.

RRIs can be calculated for different stages of the application process, and following the recommendations of the review, the data tables include the following:

Eligible pool to recommendation
Application to recommendation
Eligible pool to application
Application to shortlisting
Shortlisting to recommendation

This enables relative differences between groups to be compared for different stages of the process. However, within the publication, depending on the characteristic being discussed, only application to recommendation RRIs or eligible pool to recommendation RRIs are given – these present the best overall indication of relative success for different groups, which takes into account their prevalence among those eligible. All the exercise stage RRIs (Eligible Pool to Application, Application to Shortlisting, Shortlisting to Recommendation, Application to Recommendation and Eligible Pool to Recommendation) are presented in the accompanying tables.

Statistical significance: Where RRIs are calculated, their statistical significance is assessed – this provides a measure of the likelihood of the RRI being that large (or small) due to chance i.e. where the number of candidates is low, RRIs can fluctuate which may not indicate an underlying disparity. Statistical significance is estimated by calculating confidence intervals using a ‘bootstrapping’ approach which enables better estimation for smaller exercises^{[footnote 19]}. A 95% confidence interval is calculation – broadly, the range of values which we can be 95% confident the true RRI lies between – and where this interval contains 1, an RRI is not statistically significant.

Practical significance: To further aid interpretation using the “4/5th rule of thumb for adverse impact”^{[footnote 20]},^{[footnote 21]}, RRI values that fall within a range of 0.8 to 1.25 (the zone of tolerance) are not likely to indicate a difference in outcomes resulting in a disparity having practical significance. This does not imply that an RRI falling outside of this range is indicative of the presence of a meaningful disparity. The nature of selection exercises inevitably results in low numbers. In some cases, the numbers are too low to calculate the RRI. However, even where an RRI can be calculated, numbers within some selection exercises are low for making meaningful attributions of a potential difference. As such, caution should be taken when considering whether an apparent difference in rates, as measured by an RRI falling outside the range of 0.8 to 1.25, could represent a meaningful difference.

Of the two, statistical significance is the more important measure. If an RRI is not statistically significant, we cannot be confident that the estimated RRI reflects a real difference between the groups being compared. If an estimated RRI falls outside of the “zone of tolerance” - and might therefore be referred to as “practically significant” - but is not statistically significant, its result should not be taken too seriously as it may be down to chance. If an estimated RRI is within the “zone of tolerance” - and not “practically significant” - but is statistically significant, we can be confident that there is a real difference between the two groups, even though it is fairly small. It is then up to the discretion of policy makers to decide whether the small difference is worth investigating further and acting upon.

4.4 Professional background - calculation of “ever” legal role (judicial appointments)

The 2018-19 JAC publication was the first in which “ever” legal role was reported in addition to current legal role. The “ever” legal role measure compares “ever” solicitors (those who have declared ever holding a role as a solicitor) to “ever” barristers (those who have declared ever holding a role as a barrister). Chartered Legal Executives data is not yet included because the totals of this group are too small within the judicial appointments data to make meaningful comparisons to other groups.

In the accompanying statistical tables, the values for “Solicitor in the past” and “Barrister in the past” are a count of those who have declared previously holding a role as a solicitor or barrister. “Adjusted solicitor numbers” and “Adjusted barrister numbers” consider those who have been both a solicitor and barrister in the past. If someone has held both roles they have been assigned a value of 0.5 for both solicitor and barrister to avoid double counting. Numbers are rounded up to the nearest whole number and therefore totals may not match.

A worked example using the methodology has been presented below: Solicitor in the past = 340 Barrister in the past = 216 Both a solicitor and barrister in the past = 41 Adjusted solicitor numbers = 340 + (0.5 x 41) = 360.5 (rounded to 320) Adjusted barrister numbers = 216 + (0.5 x 41) = 236.5 (rounded to 196)

There are around 10% more applicants identified as solicitors using the wider definition of ever legal role. The ‘ever legal’ role typically results in increased RRIs when comparing solicitors relative to barristers, though broad patterns are similar; however, a smaller difference in success rates between solicitors and barristers was generally observed for senior roles using the ‘ever legal’ role. This is because an applicant’s current legal role is likely to be a salaried Judicial Office holder and so expanding the definition of legal role is likely to highlight previous experience as a solicitor and/or barrister. Further details were included in the background notes for the 2018-19 statistics.

It is important to note that the “ever legal” approach does not account for the relative periods of time an individual has spent in each profession (as this information is not captured). This may mean that some of those classified as ever solicitors may have spent the majority of their career as a barrister and should be kept in mind when interpreting the results.

Question 5

5. Users and uses of these statistics

Accepted Answer

5.1 Known users

Among the individuals, groups and types of organisations with an interest in these statistics are:


Internal customers: Ministry of Justice, Judicial Office and JAC	This group includes ministers and officials within MoJ, Judicial Office, Her Majesty’s Court and Tribunal Service and within the JAC in policy, operational or analytical roles. We have an ongoing dialogue with these users and receive most feedback from within this group.
Legal professional bodies (e.g. the Bar Council, The Law Society, Chartered Institute of Legal Executives)	These statistics include data supplied by professional bodies and in developing the new combined report we have engaged closely with them to incorporate feedback. More generally, we receive feedback from the Judicial Diversity Forum which brings together representatives from the legal professions.
Other groups with an interest in judicial diversity issues (e.g. JUSTICE, the Black Solicitors Network)	Published figures were recently analysed in detail for JUSTICE’s report on diversity in the judiciary and we are happy to provide support and advice where this is useful.
Existing judges and candidates for judicial appointment	We receive occasional ad-hoc requests via the judicial statistics mailbox, or through officials
Parliament	Statistics are used to answer parliamentary questions from both MPs and Lords
Journalists/media	These statistics are sometimes reported on or cited, most often by specialist press.
Academics, researchers and members of the public	We receive occasional ad-hoc requests via the judicial statistics mailbox or via Freedom of Information requests
Other public bodies	Judicial diversity statistics are included in the ‘ethnicity facts and figures’ publication produced by the Cabinet Office Race Disparity Unit.

5.2 What the statistics are used for

These statistics have a variety of uses, some of which include:

To inform the development and monitoring of policy and actions relating to judicial diversity. This includes the work of the Judicial Diversity Forum, which brings together officials from MoJ, JAC and the judiciary with the legal professions and draws on the patterns shown in these statistics to identify and monitor actions to improve judicial diversity
Use as evidence for equality impact assessments relating to judicial issues (including the recent consultation on raising the retirement age for judges), and to inform other aspects of judicial policy including annual evidence to the Senior Salaries Review Body
Use by charities, campaigning groups and others to hold government to account, and they inform reviews of judicial diversity including as recently published by JUSTICE and the Lammy Review.
Feeding in to cross-cutting publications to provide comparisons between diversity in the judiciary and elsewhere, including MoJ’s ‘Race and the Criminal Justice System’ statistics, and the Ethnicity Facts and Figures
Responding to occasional Freedom of Information (FoI) or other ad-hoc requests^{[footnote 22]} received directly or via the judicial statistics mailbox (judicial.statistics@justice.gov.uk)

5.3 User engagement

We always welcome feedback on the content, format and timing of these statistics.

In developing the new combined publication, a working group was formed with representatives from policy and diversity teams in MoJ, Judicial Office, the JAC and the legal professional bodies. This group met monthly during development of the report and advised on content and presentation. In addition, the way in which the outcomes of judicial selection exercises are presented in the statistics (specifically the use of the relative rate index) was independently reviewed by MoJ statisticians in consultation with a range of users, with the results of the review published.

Following the publication of the combined judicial diversity statistics, we will seek to engage with users to invite feedback and assess use of the new statistics including:

Monitoring online use of the statistics, including use of Google Analytics
Inviting feedback from users via a short user feedback questionnaire
Continued engagement with professional bodies and others interested, in particular to seek to address the planned areas for development identified below.

5.4 User feedback and changes made

The table below sets out selected details of the history of the combined judicial diversity statistics and the publications it superseded. These changes are made based on user feedback, or following consultation with users.

As noted above, we will seek and respond to feedback on the new combined report following publication, and log this as part of this guide.


2010	Judicial appointments: the first Official Statistics bulletin was published in February 2010. Prior to that, the diversity results of selection exercises were published online. Publishing these data as Official Statistics aimed to improve users’ confidence in the information
2015	Judiciary data: age breakdown included for the first time
2018	Judicial appointments: Inclusion of information related to social mobility for the first time as part of the 2017-18 statistics
2019	Judiciary data: statistics on entrants and leavers added as experimental statistics, in response to answering requests for evidence from the Senior Salaries Review Body, FoI requests and other general enquiries. The new tables increased value for users, presenting flows in and out of the judiciary to allow an assessment of how diversity of the judiciary is changing. Following review of the approach and in the absence of user concerns, these have been incorporated into the statistics from 2019-20 onwards. Judicial appointments: inclusion of reporting on ‘ever’ legal role for the first time, providing comparison of solicitor and barrister applicants on a basis which takes account of their full professional background.
2020	Development and publication of the first combined judicial diversity statistics. Judicial appointments: review of the use of RRIs published, making a number of recommendations for improvements (annex A lists these together with the resulting actions)

5.5 Latest additions to the publication

Three-year analysis (judicial appointments)

To provide more granularity with minimal risk of disclosure on Gender, Ethnicity and Profession intersections, we have combined data from the most recent three years of exercises (2018-2019, 2019-2020, 2020-2021) in the judicial appointments process. This meant that numbers of applicants and recommendations for each group were large enough for us to make meaningful comparisons.

RRIs were calculated using the same method used in RRIs with single-year data. That is by dividing the recommendation rate of the target group by the recommendation rate of the comparison group. These calculations then compare each ethnicity group to the White candidates which are the dominant group as well as different intersection groups to each other. For example, we calculated RRIs comparing White women to White men, Black, Asian and minority ethnic solicitors to Black, Asian and minority ethnic barristers and White female solicitors to White female barristers.

Due to constraints on the data available from the legal professions, the eligible pool for the three-year data is not an average of numbers across those years - which might have been more accurate. Instead, this is drawn solely from the most recent year of data, like the eligible pool for the single-year data. This was deemed by the working group to still be a good enough approximation to be a useful comparison to include in the statistics.

Though both the single-year and the three-year-combined eligible pools are derived from the most recent year of data for the legal professions and the judiciary, the overall eligible pool for their “all legal exercises” counts are different. This is due to the different weighted average applied for each type of eligible pool, whether with 5 years’ PQE, 7 years’ PQE or previous judicial experience. Weights are applied by how many recommendations there were for each type. The single-year and three-year data differ in the amount of the different eligibility types of exercises that were run in the time they cover, so their weightings are different.

For example, the 20-21 data does not have any exercises that had a 5 years’ PQE eligibility requirement, so this does not contribute to the weighting for the single-year overall eligible pool. Because they differ in the make-up of exercise eligibility in the “all legal exercises” group, the weighted average calculated for the overall eligible pool is different, and as a result, the overall eligible pool counts are different.

5.6 Future developments

Although the combined diversity statistics publication was developed working closely with key users, following publication we continuously seek wider feedback and develop the statistics as required in response to this. In addition, there are a number of areas where we have already identified that further work is required. We will seek to address this prior to the publication of the 2021-22 statistics.

Area	Planned Developments
Coverage of characteristics / improved declaration rates	As noted above, there is currently more limited coverage of diversity characteristics for the judiciary and the legal professions than for the judicial appointments process. We hope that reporting rates will improve to allow more complete coverage to be captured in this publication in future years, though this is unlikely to happen immediately. In addition, there have been some minor changes to the collection and recording of diversity information for applications for judicial selection, which may affect the presentation of these statistics in future (for example, the new form captures gender in a non-binary way).
Time series / historic data	Currently the publication presents limited data on trends over time, and does not make it easy for users to obtain data for different years; we will explore ways to bring the most recent and previously published data together better.
Definition of ‘BAME’	Currently, the ethnicity breakdowns throughout the publication use the Black, Asian and minority ethnic (BAME) grouping which does not include non-British white groups (e.g. white Irish) which differs e.g. from the use by the Cabinet Office Race Disparity Unit. We will consider whether this presentation should be changed and seek views from users of the statistics. In considering a potential change, we must ensure that statistics remain somewhat comparable to previous years, as this is vital for tracking progress.
Non-legal roles – eligible pools	Currently, eligible pools are presented only for legal exercises due to the availability of data. However, in many cases, there may be an appropriate pool which could be calculated e.g. based on membership of a relevant professional body. We will explore whether this data can be obtained in a reliable way.
Legal roles – better understanding of experience	Following on from analysis of average PQE of applicants presented in the 2020 and 2021 publications, we will be conducting further research to understand how experience in the legal professions relates to chances of success in judicial selection exercises

For example, there is ongoing work to improve declaration rates for the characteristic groups. This includes:

For the judiciary, Judicial Office is currently rolling out a new, improved data declaration system, including a campaign to encourage declaration by judicial office holders. Their declaration rates for disability, social mobility, religion or belief and sexual orientation should improve annually as a result, with data being publication-ready in the next few years.
For the legal professions, all the professional bodies have either just installed or have plans to establish new IT systems to enable diversity characteristic declaration from their members:
The BSB introduced a new data declaration system in 2018 and subsequently saw a significant increase in their declaration rates. Assuming a continued annual increase, it is hoped that the declaration rate for characteristics including disability, sexual orientation, social mobility and religion will be suitable for publication in approximately 3 to 4 years. Additionally, it is anticipated that information collected during training (where response rates are higher) will be used to populate the records for newly-qualifying barristers.
The SRA has updated the diversity questions on its systems and is actively engaging with solicitors to increase declaration rates.
CILEx is running a campaign to encourage members to improve their declaration rates following the introduction of a new data declaration system. Their declaration rates should improve annually as a result, with data being publication-ready in the next few years.

While it is likely that these new systems will prompt significant improvements in declaration, it is difficult to identify a reliable estimate of how quickly declaration rates will increase at this point.

Question 6

6. Revisions

Accepted Answer

6.1 Revisions policy

The procedure for handling planned or unplanned revisions to these statistics will be in line with the published revisions policy for MoJ statistics, available from www.gov.uk/government/statistics/ministry-of-justice-statistics-policy-and-procedures.

In particular, where errors are found in the published figures, an assessment will be made as to whether these are materially significant. If so, a revision will be announced and published at the earliest opportunity. If not (i.e. if any errors are considered to be minor) then the figures will be revised with the next publication.

The following outlines how the specific data within the publication is treated.

6.2 Judiciary

Data for the judiciary are an annual snapshot from a live administrative system, taken roughly two months after the period to which they relate. As the database is live, it is likely that were data for the same period extracted at a later date, the precise figures could be different, although we would expect any differences to be minimal. However, except where clear errors are identified, figures published for earlier years are not revised in any way.

6.3 Judicial appointments

The published statistics, though quality assured, are liable to revision. This could either be because of a late amendment to the database or because of recommendations made by the JAC after the initial report to the Appropriate Authority (see section on recommendations).

The standard process for revising the published statistics to account for these late amendments is to publish them in the next annual edition if the revision accounts for an additional 10 or more recommendations being made. However, revisions that consist of less than 10 recommendations will not be published. This is because a comparison of the original presentation of the exercise and the revised presentation of the exercise could identify those applicants recommended since the publication of the bulletin. In accordance with the disclosure policy for these data, releasing information on exercises of less than 10 recommendations may constitute a threat to applicants’ privacy (see section on confidentiality and disclosure).

6.4 Legal professions

The figures published are as provided by the professional bodies. It is not anticipated that these will be revised except when future year’s publications are produced; at this time, we will seek advice from the data suppliers as to whether any revisions are required to previously published figures.

6.5 Revisions made

The 2021 report is the second such combined publication, and no revisions have been made to date.

Question 7

7. Related statistics

Accepted Answer

7.1 Other jurisdictions

As noted, these figures relate to the judiciary of England and Wales. Diversity statistics for other jurisdictions are available:

Northern Ireland: www.nijac.gov.uk/publication/equality-monitoring-report-2020

7.2 Other diversity statistics

Diversity statistics are published by the Ministry of Justice for other elements of the justice system

The Cabinet Office Race Disparity Unit presents a range of data on ethnicity, including the judiciary and other professions, in their Ethnicity Facts and Figures

7.3 Legal professions publications and data

Each of the legal professions considered within the combined publications publishes information on diversity within the profession. This provides further detail than the relatively high level summary captured in these statistics

Question 8

8. Contacts for further information

Accepted Answer

Enquiries or comments about the content of this user guide or the publication of judicial statistics in general should be directed to:

Julian Sandler or Rita Kumi-Ampofo
Judicial Statistics team
judicial.statistics@justice.gov.uk

Press enquiries should be directed to the Ministry of Justice press office:
Tel: 020 3334 3536
newsdesk@justice.gov.uk

General information about the official statistics system of the United Kingdom is available from: uksa.statisticsauthority.gov.uk/about-the-authority/uk-statistical-system/

General enquiries about the statistical work of the MoJ can be sent to statistics.enquiries@justice.gov.uk.

Question 9

9. Annex A: Response to recommendations relating to review of the relative rate index

Accepted Answer

In January 2020, a review of the use of the Relative Rate Index (RRI) in judicial appointments statistics was published. The following summarises how the recommendations of this review have been considered and are being taken forward.

Recommendation	Response
Coverage
We recommend that applications, shortlisting and recommendations should be presented within the context of the eligible pool (of potential applicants who meet the minimum eligibility criteria). This is to ensure we cover the full scope of the JAC’s work and for the statistics to be as relevant as possible.	Complete. Eligible pool figures are presented for all legal exercises, and used as the basis for most RRIs presented within the bulletin, apart from those involving professional background due to the vast difference in size of solicitor and barrister eligible pools
Where eligible pool figures are not available: a. clear and appropriate warnings should be included within the report that care should be taken when drawing conclusions based on partial coverage of the statistics. b. further work should be undertaken to enable the availability of eligible pool data.	In progress. Eligible pool figures are not currently available for non-legal exercises and we will explore what may be possible to include in next year’s publication.
Consider further analysis initially within future ‘deep dive’ work but with a view to inclusion within the combined JAC statistical publication at a later point: a. on whether the eligible pool includes people who meet minimum eligibility requirements but have little or no chance of applying successfully, and if so the impact(s) of this. […] . Further analysis could determine this and potentially produce statistics after adjusting for years in the field. b. of more detailed or robust data from the professional legal bodies which could provide greater insight into why certain eligible individuals may not be applying as well as any potential ‘blockers’ further on in the recruitment process.	In progress. For the 20-21 publication, we have analysed judicial applications data and used that to provide a guide to average level of experience for applicants. This then informs the comparisons with diversity in the professions. Before next year’s publication, we will explore with the legal professions whether their data offers scope for further insight, and consider whether a more credible pool of potential applicants could be presented for comparison.
Additional intersectionality analyses (e.g. looking at the combination of ethnicity and gender) should be considered along with providing indications of emerging trends where there are no statistically significant disparities in any single year, and where such disparities would become statistically significant if a few years data were rolled together. Such analyses initially fit within ‘deep dive’ work but should be with a view to inclusion within the new combined JAC statistical publication at a later point.	Complete. For the 20-21 publication we have included breakdowns of gender, ethnicity and professional background intersection groups, doing so by combining data from the previous three years of judicial selection exercises.
Key statistical indicators
The relative difference between the success rates (or RRI) is recommended as the key summary indicator to flag up potential disparities between those with different diversity characteristics.	Complete. The RRI has been used as the headline indicator throughout the JAC sections of the report
It is necessary for applications, shortlisting and recommendations to be presented within the context of the eligible pool. It is therefore recommended that the key RRI findings in the new combined JAC statistical publication must at minimum include coverage of the RRI from the eligible pool (denominator) to recommendation (numerator) and where disparities are found […] provide further information about where in the interim stages of the process disparities exist. Consultation with stakeholders also revealed a desire for the key RRI findings to by default include the RRI from application to recommendation. This could be included perhaps along with the complementary RRI from eligible pool to application.	Complete. For legal exercises, most RRIs presented in the report compare recommendation rates from the eligible pool, with recommendation rates and representation percentages at different stages used to highlight where disparities exist. There are some instances where recommendation rates from application are presented, for reasons described above. Other RRIs are presented in the data tables, but not within the report (for reasons of brevity).
The new combined JAC statistical tables should include all RRIs mentioned in 2. above; from eligible pool to recommendation, application to recommendation, and for each interim stage in the selection process (eligible pool to application, application to shortlisting, and shortlisting to recommendation).	Complete. We have included these different RRIs within the statistical tables
To aid interpretation of the relative difference between the success rates (or RRI) as the key summary indicator, some underlying statistical context could be provided by the success rates on which the RRI is based; and if user consultation deems necessary the representation percentages at each stage.	Complete. Following consultation with users, particularly at the JAC, we have decided to use representation percentages as the main way of exploring patterns in overall RRIs. We will review this in light of any feedback on the new publication.
While statistical properties of the odds ratio enable it to be an easy to use metric for some deep dive analyses […] Any usage should therefore be accompanied by an illustration about what a particular odds ratio value means in practice.	Out of scope. Odds ratios are not used within these statistics, and we will consider this recommendation where relevant
Accompanying metrics / rules
The 4/5ths rule to indicate whether an RRI value is sufficiently far from one (or parity) to reflect a disparity of outcomes is considered very useful from a policy perspective. […] Given this review did not uncover any evidence that suggested the bounds should either be increased or decreased, it is recommended that the tolerance zone remains unchanged for the new combined JAC statistical publication [..’]. Consideration should be given to reviewing this at some point in the future.	Complete. The tolerance zone has been retained for this publication
It is important for the JAC statistical publication to flag up any disparities that can be considered statistically significant. Therefore, for a disparity to be regarded as a high priority for further action, we recommend it is both statistically significant and outside the tolerance zone (presently 0.8 to 1.25). A suggested guide to interpretation, which could be finessed with user testing, is: a. Inside the tolerance zone (presently 0.8 to 1.25) – this signifies no practical disparity b. Outside the tolerance zone (presently 0.8 to 1.25) but not statistically significant – this implies that the disparity could well be due to chance c. Outside the tolerance zone (presently 0.8 to 1.25) and statistically significant – this implies that we can have confidence that a disparity exists and that our best estimate suggests that it is of a size that is important practically	Complete. We have broadly followed the proposed approach, and have clearly identified where disparities are or are not statistically significant
Some indication of the potential impact of statistical bias due to non-reporting of diversity characteristics should be provided in the JAC statistical publication […]. The total number, declaration rate and the related number of unknowns should continue to be published. Where possible these should be accompanied by a breakdown of the number of unknowns to understand the extent to which people are choosing any ‘prefer not to say’ and ‘prefer not to share data outside the JAC’ options and so are engaging with the monitoring process rather than simply not responding. Where such options exist but the related data are unavailable, further work should be undertaken to try and obtain these.	Partly complete. We have retained information on declaration rates, though not yet added the more detailed breakdowns indicated.
Where exercises are grouped together in the JAC statistical publication it would be helpful for the total numbers, declaration rates and numbers of unknowns (where possible with the additional breakdowns described in 3. above) to be provided separately at the application stage (instead of only at aggregated grouped level) for each exercise involving ten or more applications.	Not yet considered. We will consider the most appropriate way to present this information in future.
The potential for statistical bias is best dealt with at source. While declaration rates are consistently high overall, they vary by characteristic. Only one declaration rate in the 2019 JAC statistics publication tables was less than 70% and this was also less than 60%. The declaration rate threshold could therefore be raised from 60% to 70% or even 80% to encourage even better reporting which will in turn lead to more accurate statistics. The level of this threshold should be reviewed every few years.	Considered but not completed. In publishing the first and second combined statistics, we have kept the existing threshold for now, as for judicial and legal professions data declaration rates are lower than for judicial application. We will continue to review this annually as more data becomes available.
Exercises with fewer than ten recommendations should continue to be aggregated into groupings for confidentiality reasons so applicants cannot be personally identified.	Complete. We will continue with the existing method of aggregating smaller exercises in the 2020-21 publication
Where exercises are aggregated and the eligible pools are known, an overall group eligible pool could be estimated by summing the eligible pools for each exercise.[…]. Statistical anomalies could be avoided by checking the trend in the aggregated exercises is consistent with trends seen in the individual exercises.	Complete. While we have developed a method to enable estimation of RRIs for all legal exercises, this is based on weighting by recommendation, rather than summing eligible pools
Unless there are overriding confidentiality, presentational, or statistical concerns, there is no need for figures to be suppressed […] the provision of confidence intervals enables users to gain a good understanding of the uncertainty around key statistical indicators. While the calculation of standard confidence intervals needs careful modification for small groups, an alternative approach is recommended that renders such modifications unnecessary. Any presentational concerns arising from small groups leading to particularly high RRIs should be monitored and dealt with on a case by case basis until such a time that a useful rule can be put in place. If the current practice of suppressing figures where denominators are less than 10 is continued, it should be applied after first excluding any unknown values.	Complete. The recommended approach to calculation of confidence intervals for small groups has been adopted.
Presentational features
For the summary section to be user friendly we advise it has a similar style to other Ministry of Justice publications which use a bullet point structure with visual aids such as arrows and signs. The chosen approach should enable the publication to be easily readable with clear key messages. The summary of main points should ideally be one page and should contain one main point per section	Complete. We have reduced the number of main points (when compared to the previous publication) and followed an approach consistent with other MoJ publications
To enable the publication to be as navigable as possible we recommend including a table of contents. Efforts should be made to keep the report as concise as possible without any unnecessary graphs. Contextual information should be provided in an accompanying statistics definitions and measurements document and additional information provided in the appendices.	Complete. A table of contents has been included, and the overall length of the publication is reduced compared to the previous separate ones in combination. The publication includes key context (‘things you need to know’) and this document provides more details of definitions and measurement.
The key messages should be sufficiently well explained so that lay users don’t have to refer to another part of the publication. This includes key points from plots and summary statistics relating to the RRI e.g. ‘the following disparities were significant both practically (being outside the tolerance zone of no disparity) and statistically: females in the eligible pool were 43% less likely to be recommended than males in the eligible pool with them having a success rate of 20% compared to 34% for males’.	Complete. We have attempted to ensure that the headline messages relating to the RRI are clear, including testing with users. We will review this following publication when it is possible to gather a wider range of feedback.
All key messages should also include clear and well positioned caveats to avoid the potential for misleading conclusions to be drawn. For instance: a. In the summary that disparities may be explained by differences in other factors. b. where no eligible pool figures are available that care should be taken when drawing conclusions based on partial coverage of the statistics. c. where annual comparisons need to be approached with caution due to changes in exercises over the years.	Complete. We have attempted to ensure this where relevant. In particular, a conscious decision has been taken to avoid annual comparisons within the sections related to judicial appointments.
The key headline data should be easy for publication users to find and use. This could be achieved by including tables showing the key RRI statistics for each diversity strand.	Complete. While we have not included a table as suggested, the new publication has been developed with a range of users to ensure headline messages are clear. We will review this further following publication.
Presentation of the figures/plots should be reviewed. For instance, forest plots with accessible colouring could be used to show the RRIs accompanied by 95% confidence intervals and the zone of tolerance. Any (horizontal) bar charts could include the actual percentage figures along with accessibly coloured bars.	Complete. We have introduced forest plots as suggested, and improved bar charts by including percentage figures.
Explanations of the RRI and other technical terms should be reviewed to ensure they are understood by lay readers and that the terminology is clear and accurate. In particular the 4/5ths rule (or tolerance zone) should be described as a measure of practical rather than statistical significance while the declaration rate is about declarations rather than unknowns. Consideration should be given to moving these sections to an appendix with short concise wording developed for the main part of the report	Complete. We have attempted to do this, including review of the draft publication by key users within the JAC who provided feedback to further refine the draft. We will also review this following publication and seek to address any issues for the following year’s publication.

Question 10

10. Glossary

Accepted Answer

Statistical Term	Definition	Example/Usage in Publication
Confidence Intervals	This is the range of values within which the true value will fall 95% of the time	In the publication we provide the 95% confidence interval for RRIs - the range the true rate difference is 95% likely to sit within. Results are described as statistically significant if this range does not overlap the value when there would be no difference between groups, i.e. 1 in RRIs.
Statistically significant	This describes the confidence we have that the derived values are caused by a real difference between the two groups and did not result due to random chance.	Through the publication a disparity is described as statistically significant if there is a less than 5% probability that the estimated disparity could have happened by chance; this means we can feel confident that the result is due to a real difference between the groups
Practically significant / practical significance	This examines whether an RRI really shows a noticeable difference between the recommendation rates of two groups in the actual exercise that was run. If the difference is not of practical significance, it is probably not something that makes a noticeable difference to the outcomes of candidate pools in the real world and is likely to be much less impactful than other factors.	For example, in a candidate pool of 100 male and 100 female candidates, if the RRI for female to male = 0.9 and within the tolerance zone. This means that 10% of male and 9% of female candidates were recommended. In this case, the difference is only one candidate, and so probably not noticeable to the applicant pool in practice.
Recommendation rate	The recommendation rate for any group is the number of applicants in that group who were recommended for appointment, as a percentage of those who originally applied, or as a percentage of the total number of that group in the eligible pool.	If 100 women applied for a role and 10 women were recommended for appointment, the recommendation rate for women in that exercise is 10%.
Relative Rate Index (RRI)	Relative Rate Index or RRI calculations are used to describe the difference in recommendation rate between two groups of candidates. RRIs are calculated by dividing one recommendation rate by the other. It is known as a relative rate index because it explains the difference between the two groups without giving the recommendation rate.	If 25% of female candidates are successful, and 50% of male candidates are successful, the RRI is therefore 25%/50% = 0.5. This RRI means that the success rate for female candidates is half of that for male candidates; or equally that the success rate of male candidates is double that of female candidates. An RRI of 1 means there is no difference between the recommendation rate of the two groups. A value of less than 1 means that the target group has a lower recommendation rate than the comparison group, while an RRI of more than 1 means that the target group has a higher recommendation rate.
Zone of tolerance	This is the range around the statistics representing a disparity which is not large enough to be considered important. In this publication we use the ‘4/5th rule of thumb for adverse impact’. (see footnote 21 in guide)	As an example, RRI values that fall within a range of 0.8 to 1.25 (the zone of tolerance) are not likely to indicate a difference in outcomes resulting in a disparity having practical significance.
Weighting	Used to bring the sample being analysed more in line with the population being analysed. That is, instead of a candidate counting as 1, they might count as 1.25 or .75 respondents depending on how low or high their representation reflects the whole exercise.	In this publication, in order to produce an overall eligible pool combined across all legal exercises with differently sized eligible pools, a weighting was made by the number of recommendations for each exercise
Disparity in/of outcomes	This represents the difference in results of an indicator. Results falling within the zone of tolerance are considered to represent no practical disparity.	In this publication, a practical disparity is when one group’s pass rate is less than four-fifths (80%) of another group’s pass rate.

Most notably, prior to 2019-20 all judges were assigned a gender with no unknowns. From 2019-20 onwards, those judges who opt to ‘prefer not to say’ are classified as unknown gender in the statistics. ↩
RELs and RFLs are not eligible for judicial appointment and therefore do not form part of the ‘eligible pool’. Consequently they are excluded for consistency. ↩
Some solicitors without practising certificates may still be eligible to apply for judicial roles; these are not captured in the eligible pool used for these statistics. ↩
Scotland and Northern Ireland are different jurisdictions, and where separate statistics are published. ↩
The majority of judicial office holders (over 90%) hold only one appointment, although more tribunal judges than court judges have two or more roles. ↩
In courts, just over half of judges are in fee-paid positions, with just under half being salaried. Among tribunal judges, around three-quarters are fee-paid, with around a quarter salaried. The large majority of salaried judges in both courts and tribunals work on a full-time basis. ↩
In addition, on rare occasions and for operational reasons, it is planned that recommendations will be made to the Appropriate Authority through more than one report sent on different dates, such as where requirements for different jurisdictions are separately considered. Under those circumstances, the exercise will be considered to have been completed when the last report has been sent for that exercise. This means that the result of the exercise can be provided in a single, comprehensive presentation, rather than in stages, to support easier understanding for the users of the bulletin. ↩
For example, individuals may apply simultaneously to both the Fee-paid Medical Members of the First-tier Tribunal (Mental Health) and the Mental Health Review Tribunal (Wales). ↩
Prior to 2019-20, years of post-qualification experiences were based on years since admission to the roll (solicitors) and years from being called to the Bar and having completed pupillage (barristers). ↩
For solicitors, years as a Registered European Lawyers (REL) or Registered Foreign Lawyer (RFL) are also used to calculate post qualification experience. So a solicitor who has held a PC for 4 years and worked as a REL for 2 years before that, will be counted as having 6 years PQE. ↩
This represents the current availability of data. From 2020-21, JAC will in theory capture non-binary gender and should numbers be sufficient to include this as a separate category in future without risk of breaching confidentiality this will be done for future publications. ↩
This is most notable for individual judicial selection exercises ↩
Accordingly, it is possible that age group distributions at the shortlist and recommendation stages may deviate slightly from the age groups presented. Such differences, if any, would be very small and non-material ↩
Declaration rates can vary for subsets of the data (e,g. specific appointments or selection exercises), so that the figures here should be treated as a broad guide based on the overall dataset ↩
The declaration rates for recently qualified solicitors have declined since the SRA moved its authorisation process on line and measures are being taken to address this. ↩
Information on practice area is available for barristers, but not currently for solicitors (information is collected at a firm level, but that captured for individuals is not considered reliable and may not have been recently updated). For barristers, at authorisation to practice the percentage of income that comes from different listed areas of practice. ↩
Information on practice area for Chartered Legal Executives is available but not included within the publication. CILEx members (from associate grade upwards) are required to notify their practice area/area of specialism to us when submitting their CPD and prior conduct forms annually. ↩
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/869028/Methodology_Review_Report.pdf ↩
This approach is used for all confidence intervals with the exception of those for the “ever” legal role variable, where non-integers mean that the approach used results in errors and as a result confidence intervals are calculated by formula instead. ↩
Dan Biddle. Adverse Impact and Test Validation: A Practitioner’s Guide to Valid and Defensible Employment Testing. Aldershot, Hants, England: Gower Technical Press. pp. 2–5. ISBN 0-566-08778-2. ↩
Toward a Coherent Test for Disparate Impact Discrimination: Peresie, J.L. 2009 ↩
Typically, these requests relate to a more detailed breakdown than is available within the published statistics. In such cases, a copy of the requested information will also be published. A summary of all FoI releases received by MoJ can be found at www.gov.uk/government/collections/freedom-of-information-disclosure-log ↩

Cookies on GOV.UK

1. Introduction

2. Background to judicial diversity statistics

2.1 The judiciary in England and Wales

2.2 Judicial appointments

2.3 Legal professions

3. Data sources, coverage and definitions

3.1 Data sources

Judicial office holders

Judicial appointments

Legal professions

3.2 Coverage and definitions

Judicial office holders

Judicial appointments

Legal professions

3.3 Diversity characteristics

Gender

Ethnicity

Age

Professional background / legal role

Disability

Sexual orientation

Religion or belief

Social mobility

3.4 Summary of declaration rates for all diversity characteristics

3.5 Confidentiality and disclosure

4. Methodology and calculations

4.1 Representation percentages and declaration rates

4.2 Recommendation rates (judicial appointments)

4.3 Relative Rate Index (judicial appointments)

4.4 Professional background - calculation of “ever” legal role (judicial appointments)

5. Users and uses of these statistics

5.1 Known users

5.2 What the statistics are used for

5.3 User engagement

5.4 User feedback and changes made

5.5 Latest additions to the publication

5.6 Future developments

6. Revisions

6.1 Revisions policy

6.2 Judiciary

6.3 Judicial appointments

6.4 Legal professions

6.5 Revisions made

7. Related statistics

7.1 Other jurisdictions

7.2 Other diversity statistics

7.3 Legal professions publications and data

8. Contacts for further information

9. Annex A: Response to recommendations relating to review of the relative rate index

10. Glossary

Is this page useful?

Help us improve GOV.UK

Help us improve GOV.UK