Background quality report

Data quality for early conciliation and employment tribunal data for England, Scotland and Wales

Published

Introduction

When an employee wants to make a claim against their employer at an employment tribunal (ET), usually they must notify Acas first. Acas offers early conciliation (EC) to try to reach an agreement between the employee and employer and avoid the employee making a claim to an employment tribunal.
 
In some circumstances, the employer may approach Acas to help resolve a potential dispute. This can also be handled using early conciliation.
 
If the case is not resolved by early conciliation, and the claimant still wants to go ahead, they must submit an ET1 form to make a claim to an employment tribunal.
 
This background quality report relates to our statistical release about the:

  • early conciliation notifications received by Acas
  • their progress through the Acas conciliation system
  • how they're resolved 

We have not published this series since December 2018 because we then introduced a new case management system (CMS) which changed the classification of EC notifications and ET cases. This has made these data not directly comparable to previous Acas EC and ET statistical releases. The way we count cases is also different to HM Courts & Tribunals Service (HMCTS) so our tribunal statistics should not be directly compared to theirs. 
 
These statistics are not classed as official statistics but we've voluntarily applied the UK Statistics Authority Code of Practice for Statistics to help ensure they are:

  • high quality
  • trustworthy
  • provide as much value as possible

Strengths and weaknesses  

Statistics will be published quarterly on the Acas website.  
 
As these are management information (MI) data, they represent full coverage of all cases in our CMS. These data have been presented in both tabular and graphical form to allow for ease of interpretation and separate data files have been provided to enhance use of the data for further manipulation.

Weaknesses in the system include the inherent data quality issues that come with management information. However, work is underway at Acas to improve the quality of the data recorded in our CMS. There is also a trade-off between timely reporting of case numbers and accuracy in the data. Data are extracted 1 month after the end of the reporting period which allows time for case resolution and proper track assignment. However, due to lack of contact with some claimants, there will be some unassigned tracks at this time. We feel that this time period offers the best compromise between these 2 factors.

At present, it is not possible to extract accurate ET multiple case numbers from the CMS. We have made the decision to publish the data without these figures rather than delay publication further because we feel that the data provide value even without these numbers.

Future improvements 

In future publications, we are aiming to:

  • publish geographical breakdowns of case numbers 
  • include multiple ET case data, currently not included due to a known limitation of the CMS

Relevance

These data are the only source of EC statistics and so provide the most complete set of published data on this topic outside of the Acas annual report.  

They are MI data so they offer complete coverage of EC notifications and ET cases in the current system for England, Wales and Scotland.

They have been designed to cover the main freedom of information (FOI) requests that Acas receives.  This reduces burden on both the users and producers of the statistics by having 1 set of data that meets the needs of many users.

Due to the time lag in resolution of cases, data on outcomes are presented 1 quarter behind data on case numbers. This is so we can publish data on case numbers to increase the timeliness of publication.
 

Accuracy and reliability

These data are taken from the Acas CMS and are representative of the full population of EC notifications and employment tribunal cases with a very small number of exceptions.

Process of production of data tables

  • Data is extracted from the Acas case management system 1 month after the end of the latest month reported on.
  • The data is initially sorted based on the receipt dates of the early conciliation notifications to complete Tables 1 and 2 (see section below on case counting).
  • The data is then sorted by the date of outcome for early conciliation for Table 3 and 4. Table 4 covers a slightly different period to capture all the outcomes (see section below on timeframes). Some of the early conciliation notifications are marked on the system as having no track identified. These are manually investigated and the correct track assigned. There is work in progress to avoid the no track notifications requiring manual intervention.
  • The data is then sorted by the receipt date of the employment tribunal claims for Table 5.
  • Finally, for Table 6, the data is sorted by employment tribunal outcome date. Again, any cases with no track assigned are investigated and updated with the correct track.
  • For a minority of cases, there are some single cases that are yet to be identified as part of a group, so it is possible the number of groups is slightly lower than actual with the singles correspondingly inflated.
  • All figures are checked for sense and continuity, and the data is run in an alternative way to check that the 2 sets align (‘dual-run’).

Timeframes

We select 1 month for the data extraction so we can assign a track to most cases for more accurate reporting. For EC, we allocate a 6-week window to try to resolve the case and attempt contact with the parties in the first 2 weeks. So 1 month allows us to assign almost all the cases to their correct track and provides a trade‑off between timely statistics and data accuracy.

For outcomes, most 'jurisdictions' (areas of employment law) have a 3-month time limit from the date of the incident until they progress to tribunal. If claimants take part in EC, this effectively stops the clock for up to 6 weeks  As the outcomes tables lag by a quarter, this gives 4 months to compile the outcomes data. As the majority of cases which are going to progress to ET will have done so at this point, this time frame provides the best trade‑off between timely statistics and accurate reporting.

Case counting and grouping

When counting EC notifications, the date of receipt we use is the timestamp at the point when the prospective claimant submits the notification on our website.

When counting ET1 receipts, we use the date of receipt at Acas when the record is first input into the CMS.  This is usually later than the submission date with the HMCTS, sometimes by a number of weeks, so the ET1 receipts in a given month at Acas can include some cases submitted in previous months and exclude some submissions from the current month.

For EC notifications, we gather these into group cases if they:

  • are submitted on the same date
  • have the same representation
  • are about the same dispute

We can also create looser relationships based on similar characteristics.

For ET1s, we gather these into batches called ‘multiple' cases which:

  • have a claimant-side representative in common
  • are about the same dispute

The size of the batch or the date of submission does not affect this.  If the ET1s all concern the same dispute but do not share a representative, we do not group the cases in this way. 

Any large-volume claims (such as equal pay) which are too large to be put onto the Acas system are handled administratively and therefore may be excluded from the overall totals. Also, a small number of claims are not open to Acas conciliation and are therefore not counted in Acas totals.

Timeliness and punctuality

There has been a break in publishing these statistics due to a change in the CMS and methodology for registration of cases.  

Each new dataset will be released approximately 2 months after the end of the final month of the quarter that it reports on.  The publication of January to March data may be delayed because we have to publish the statistics in the Acas annual report first.

Accessibility and clarity

To make this data accessible and clear, we have:

  • published full data in tabular form with graphical presentation of the headline figures 
  • provided the full data in a separate open document spreadsheet (ODS) file, so that people can perform further analysis if required 
  • presented tables and graphs in line with the Office for National Statistics (ONS) standards for accessibility, where possible

We are working to publish charts and images in a more accessible way.

Coherence and comparability

Acas carries out the classification of EC cases into track based on indicative jurisdictions identified by Acas officers working on the cases. It is common for the indicative jurisdictions recorded at the EC stage to be different from jurisdictions recorded at the employment tribunal with a tendency for additional jurisdictions to be added on the ET claim. 

As these statistics are neither directly comparable to previous Acas publications before December 2018 on EC and ET cases, nor to HMCTS tribunal statistics, the methodology article for the statistics gives a more thorough explanation of why the figures differ and how they should be used.

The main effects on case numbers presented in this series of publications compared to the previous set of publications on EC and ET case numbers has also been set out in the accompanying methodology article.

Where figures fewer than 5 appear in the tables, these have been suppressed and replaced with a tilde (~) for disclosure reasons. These suppressed figures have been removed from any totals and so the statistics are internally coherent.

Trade-offs between output quality components

There is a trade-off between timely reporting on the statistics and classification of cases into their correct track.

As this issue is ongoing for each quarter and may take considerable time for all cases to be assigned to their correct track, we prioritise timely publication of the statistics over complete assignment of the tracks. This leaves a small number of 'no track assigned' cases at publication date but also enables timely publication of the statistics.

Similarly, as there is a time lag for cases to be resolved, the tables on outcomes of cases will be reported for the quarter behind those for case numbers. This is to allow for timely reporting of cases and the lag will remain throughout all publications in this series. 

Assessment of user needs and perceptions

Data have been presented to meet the requirements of most of the FOI requests that Acas receives about EC notifications and conciliation outcomes.

We will invite users who have made these requests to join a stakeholder group for these statistics to ensure that they remain relevant. 

Performance, cost and respondent burden

Resources taken to collate and publish statistics will be offset by the reduction in ad hoc requests which this publication should largely reduce the need for.

There is no additional burden on the customers for data collection because data for the statistics is automatically obtained during the course of conversations between them and Acas.

We are aiming to use a 'reproducible analytical pipeline' (automated statistical and analytical processes) to produce these statistics in the future which will reduce the time required for table production and manual checking of these data.  

Confidentiality, transparency and security

Confidentiality

All published figures are counts of individuals in particular groupings. Where there are possible disclosure issues in reporting protected characteristics, data are suppressed and represented by a tilde (~) to prevent disclosure of information on individuals. This follows Office of National Statistics (ONS) guidance on disclosure control.

Transparency

The bulletin provides commentary on the key features of the outputs and identifies any issues or caveats to the data. The accompanying methodology article explains how the statistical method has changed from previous releases and how the statistics can and cannot be used, as well as caveats in the data. This quality report provides further information on the method, production process and quality of the output.

Security

All staff involved in the statistical production process follow Civil Service and data protection regulations. The data is stored, accessed and analysed using Acas’s restricted network and IT systems, and access to raw data is restricted. Access to data is only granted to staff who need it to carry out their work.