Findings

Introduction

Statistics on Pay in the Northern Ireland Civil Service (NICS) have been produced annually by the Human Resources Consultancy Services branch (HRCS), which is part of the Northern Ireland Statistics and Research Agency (NISRA), since 2011/12. The statistics present an overview of NICS pay as of March of each year, with historic data going back to 2006.

In requesting this assessment, the statistics team at the Northern Ireland Statistics and Research Agency (NISRA) is demonstrating its commitment to produce Pay in the NICS statistics that meet the standards required of National Statistics and the Code of Practice for Statistics.

We judge that these statistics meet the highest standards of the Code and we have not identified any requirements for the statistics to achieve National Statistics status. OSR therefore recommends that the UK Statistics Authority designate the statistics as National Statistics.

The statistics are vital for transparency of public sector pay

The Pay in the NICS statistics have a range of users who are predominantly interested in the distribution and equality of pay in the NICS. Most users we spoke to in carrying out this assessment were from departments within the NICS. However, the statistics are also used externally by trade unions and equality monitoring bodies. A primary use of the statistics is to support pay negotiation processes. The other main use of the statistics is to analyse how pay varies for different groups within the NICS, for example by grade, sex and community background.

We heard from users within the NICS that public sector pay has been a growing area of interest following the Covid-19 pandemic, as many public sector workers across the UK experienced a pay freeze. The statistics have been used to brief senior officials and Ministers on how pay has changed over the last year, to inform pay policies, as well in response to queries from Members of the Northern Ireland Assembly and members of the public.

The transparency of data on NICS pay is valuable to the public as it provides insight into government spending on the Civil Service workforce. When the statistics are published each year, they receive wide media coverage. The statistics team keeps track of the coverage it gets each year, including monitoring Twitter impressions and reach. In 2020, there was some misreporting of the headline figures from the Pay in the NICS statistics which incorrectly reported that the increase in median pay was the level of pay award received by all civil servants in Northern Ireland. The statistics team was proactive in working with NICS Human Resources colleagues and the media to correct the statements being made.

Differences between groups and geographically is of high priority to users

As the statistics are used widely for equality monitoring, it is important that they contain the relevant breakdowns to support monitoring under Section 75 of the Northern Ireland Act 1998. This Act requires public authorities to promote equality of opportunity for all groups including by community background, age and disability. Users told us that being able to have statistics on these groups by grade is vital for analysing pay quality and supports briefing of particular issues for engagement with trade unions as part of the pay setting process.

Where users have perceived there to be data gaps, the statistics team has been open and receptive to exploring the feasibility of producing these breakdowns. During the process of this assessment, some users expressed an interest in breakdowns such as working patterns, highest to lowest earner ratios and data on pay bands within Senior Civil Service (SCS), which we shared with the statistics team. Since then, the team has been enthusiastic about reaching out to these users to work with them to understand their reasons for wanting these breakdowns.

We found a common area of interest for users is the desire to make comparisons of pay for civil servants across the UK. These comparisons would support the pay setting process to understand whether pay in the NICS is keeping pace with other parts of the UK. However, there are currently no directly comparable pay statistics for England, Scotland or Wales. The Cabinet Office produces annual Civil Service Statistics, based on the Annual Civil Service Employment Survey, which covers the UK, Scottish and Welsh Governments. The Northern Ireland Civil Service is not included.

The Civil Service Statistics include breakdowns of pay for each grade by different protected groups or by regions, but they do not include a cross-tabulation of protected characteristics and region to allow for comparisons with Northern Ireland. In carrying out this assessment, the statistics team showed enthusiasm and willingness to work with its counterparts across the UK to explore the feasibility of filling this data gap to improve the coherence of pay statistics for the UK. We would encourage the statistics team to take forward this idea to understand whether the user need for UK-wide comparable statistics on Civil Service pay exists outside of Northern Ireland.

Strong user engagement has informed the development of the bulletin

Users are invited to feed into the development of the Pay in the NICS statistics through the annual HRCS Customer Satisfaction Survey, which looks to collect feedback on both the statistics and the service provided by the statistics team. The team also carries out regular ad-hoc engagement with users to ensure the statistics continue to meet their needs. We heard positive feedback from all users about the statistics team’s engagement and in particular, users praised the statistics team for continually looking for ways to improve the statistics and bulletin.

We heard from some users within NICS Human Resources (NICS HR) that they have been given mock-ups of bulletins ahead of publication to be able to feed back on the content and structure. As a result of this and wider user engagement, the statistics team made several improvements to the 2021 Pay in the NICS bulletin. This includes additional information and a short video on what the median is and why it is used, to prevent misreporting of the headline statistics. In response to user feedback about the accessibility of data tables, the team reduced the time period covered in the data tables to 10 years, as earlier data can be found in previous releases. The 2021 statistics also include a more detailed breakdown of industrial staff into specific grades and more departmental comparisons with pay scales in Great Britain.

An external user, however, highlighted that the narrative on pay increases doesn’t necessarily reflect what civil servants in Northern Ireland have seen in their payslips over this time, as pay is generally reported in nominal not real terms (i.e. not adjusted for inflation). We would encourage the statistics team to explore whether there is wider user interest in producing real terms timeseries of NICS pay as part of the statistics.

The quality of the statistics is well understood by the team

We found that users had no concerns about the quality and methods used to produce the Pay in the NICS statistics. The statistics are sourced from administrative data taken from HR systems and cover 100% of staff in the NICS, including both non-industrial and industrial staff. This includes a small number of staff in the Northern Ireland Prison Service (NIPS) who are classified as civil servants as these individuals are mostly staff with teaching rather than custodial duties.

In 2006, HRConnect was awarded a contract to bring all departments in the NICS onto the same HR system. This allowed for the creation of the Pay in the NICS statistics as the data were being collected in a consistent way across departments. The one current exception is the NIPS staff in scope for these statistics, whose HR details are recorded on the NIPS’ separate standalone HR system Compass. The majority of administrative staff in NIPS have already transferred to HRConnect and the remaining Prison Grade staff may transfer to HRConnect in the future. The statistics team has worked with HRConnect and Compass teams since the creation of these statistics, to develop the necessary extract routines to get the fields and breakdowns it needs.

We heard positive feedback from the data suppliers and managers about the strong relationships they have built with the statistics team. In particular, they commented on the ease with which errors are highlighted, investigated and amended in all versions of the data. This positive relationship has made it easier to investigate anomalies and respond to data queries where the data are not included in the bulletin.

The statistics team has demonstrated its understanding about the strength and limitations of the data by publishing a Quality Assurance of Administrative Data (QAAD) report. The QAAD report covers all the NICS statistics produced by HRCS which includes sickness absence, employment, equal opportunities, recruitment and pay. The QAAD report is thorough in that it not only details the quality elements of the statistics (such as data collection, quality assurance and potential bias), but it provides information on the relevance of the statistics. This demonstrates a good understanding of the suitability of data and professional capability of the team. The QAAD report is routinely updated, along with the quality methodology in the bulletin. We heard positive feedback from users about the clarity and usefulness of the quality information.

Improvements to timeliness will better address user need

Some users commented that the timeliness of the statistics adversely impacts the public value of the statistics. The Pay in NICS statistics are usually produced in December of each year and refer to data as at the end of March of the same year. There is therefore a gap of around nine months between the reference date and the date of publication.

The main reason for this is the delay in implementing the annual pay award for NICS staff, which is beyond the team’s control. For several years, the annual pay award has been implemented in July, meaning up-to-date pay data are only available after that date. The data must then be processed and quality assured. An additional delay has been caused by the requirement to wait for data from the Annual Survey of Hours and Earnings (ASHE), which HRCS uses to provide a comparison between the pay in the NICS and pay in the wider public and private sectors in NI, which are normally not available until later in the year.

However, in 2021, the statistics team has made efficiency improvements to the production process compared to previous years to improve the timeliness of their output and are investigating further potential improvements for subsequent editions. The 2021 statistics were published on 18 November 2021, an improvement to timeliness of three weeks compared with 2020. This improvement has been achieved by a number of process changes including improved planning and process documentation, taking a more proactive approach to obtaining answers to queries from data suppliers, and liaising and agreeing with customers to defer other pieces of work until after publication.

The statistics team is examining the potential for making additional changes to streamline the data quality and validation process for production of the dataset and bulletin, with a view to further improving timeliness. For example, the team has updated its work plan GANNT chart to allow it to clearly identify, and deal with, logjams or inefficiencies.

The statistics team is also submitting a proposal to the NISRA Tech Lab, an initiative designed to harness new technologies through a targeted investment bringing together skilled personnel to provide a centralised dedicated technical resource to the Agency. The team’s proposal would look to explore the possibility of automating the production of the bulletin, with the aim of further improving timeliness and accuracy, and reducing the need for manual checking and updating. This is a welcome step to build in innovation to the way the statistics are produced to improve processes and better meet user need.

The team adopts best practice on data governance

The statistics team adopts and exhibits best practice on data governance. HRCS acts as a data processor for NICS HR and has published a Code of Practice Data Governance and Confidentiality Statement, which sets out the arrangements that HRCS has put in place to manage the data securely. Data control and sharing arrangements are stipulated within the Service Level Agreement (SLA) and Memorandum of Understanding (MoU) between NICS HR and HRCS. A separate agreement is in place between NIPS and HRCS outlining the arrangements in place for the secure transfer of data from Compass.

All HRCS staff must complete mandatory on-line data protection and information security training on joining the branch and complete further annual refresher training. HRCS staff apply statistical disclosure control to all data sourced from HRConnect and Compass to ensure that no one nor any sensitive information relating to them are identifiable from the data. A Data Protection Health Check carried out across HRCS in 2021 indicated that successful data compliance appears highly likely as there were no major outstanding issues from the review to show significant deviation from policy, standards or guidance.

We heard from one of the regulator bodies in Northern Ireland that it was so impressed by HRCS’ approach to data governance and confidentiality, that it has taken this best practice and adopted it in its own organisation. This is a great example of sharing best practice and demonstrating trustworthiness in statistics production.

A demonstrably capable and helpful team

The statistics bulletin and tables are presented impartially. Users told us that it is valuable to be able to refer to independently produced pay statistics, when talking to senior leaders or trade unions about pay, as they are free from political pressure. The statistics team is seen as professional, capable and helpful. User feedback we collated supported this view.

The roles are clearly defined and the team’s workplan sets out the timeline and responsibilities for publishing the statistics each year, which even includes user engagement activity. The team has produced a manual for producing the statistics which includes comprehensive annotation and desk notes for running the code and producing the tables. In carrying out this assessment, the statistics team provided Continuous Professional Development (CPD) logs which showed participation in Government Statistical Service (GSS) learning and NISRA-wide events on non-technical professional skills. The responsible statisticians told us that they are encouraged to participate in training and development as part of their role, demonstrating a commitment to working with credibility.