Labour Disputes Inquiry QMI

• Accredited official statistic: yes

• Survey name: Labour Disputes Inquiry

• Data collection: survey

• Frequency: monthly

• How compiled: online

• Geographic coverage: UK

• Related publications: Labour disputes in the UK and Labour market overview

This Quality and Methodology Information (QMI) report contains information on the quality characteristics of the statistics (including the European Statistical System's five dimensions of quality) as well as the methods used to create it.

The information in this report will help you to:

• learn about existing uses and users of the statistics

• understand the strengths and limitations of the statistics

• understand the methods used to create the statistics

• help you to decide suitable uses for the statistics

• reduce the risk of misusing statistics

• Labour disputes statistics are estimated using the Labour Disputes Inquiry.

• The Labour Disputes Inquiry aims to record all strike action that has taken place in the UK each month, except for those disputes involving fewer than 10 workers or lasting for less than half a day.

• Disputes are picked up from reports in the media, trade union websites, and from third party organisations, with data then collected directly from the business involved in the dispute, where possible.

• Data are compiled to produce estimates of working days lost, number of workers involved, and the number of stoppages, which are published monthly as part of the Labour market theme day.

• UK data on labour disputes are available back to 1891.

• In April 2020, collection and publication of labour disputes data were temporarily suspended to protect and prioritise our outputs in response to the coronavirus (COVID-19) pandemic; collection of the Labour Disputes Inquiry restarted in June 2022, with collection backdated to January 2022, data for 2020 and 2021 are unavailable.

• In 2023, the average response rate was 62.0%.

Overview

The Labour Disputes Inquiry collects data on working days lost, the number of stoppages, and the number of workers involved in strike action from UK businesses that have employees involved in strike action.

The inquiry tries to record all strike action that has taken place, except for those disputes involving fewer than 10 workers or lasting less than half a day. However, we do collect data if 100 working days are lost because of a single dispute, regardless of the number of workers involved.

Disputes are picked up from reports in the media, trade union websites, and from third party organisations. Where possible, we seek to collect the data directly from the business involved in the dispute, or from intermediary organisations such as employer associations in some sectors.

Uses and users

Data from the Labour Disputes Inquiry are used by a wide range of internal and external customers. Internally, these figures feed into regional national accounts, where they are used to offset against the Compensation of Employees data.

The figures are used by the Department for Business and Trade (DBT) to provide ministers with a brief on labour disputes. Other users in government include the Cabinet Office and the Advisory, Conciliation and Arbitration Service (ACAS).

Outside government, the data attract interest from the media, researchers, local authorities, development corporations, academics, and other users within the private sector. This includes the Trade Union Congress (TUC) and its member unions, plus employer associations in specific sectors.

In accordance with the Resolution concerning statistics of strikes, lockouts and other action due to labour disputes: 1993, the UK provides the International Labour Organization (ILO) with data covering working days lost, number of workers involved and the number of stoppages by industry for each calendar year. This information is then published on the ILO statistical website.

Confidentiality

The information is treated as confidential as directed by the Code of Practice for Statistics. It is only held for as long as it is being used for producing statistics. The statistics produced do not identify any organisations or individuals within an organisation. 

Survey information is also provided to other government departments, approved organisations and approved researchers for statistical purposes only. Details on who can access this information can be found on the approved organisations and government departments page and the approved researchers page. All statistics produced are subject to the Code and the same standards of protection are applied to the data at all times.

Strengths and limitations

Strengths

• Disputes are picked up from reports in the media and trade union websites, with all businesses with employees on strike surveyed; as a result, the Labour Disputes Inquiry has almost 100% coverage of businesses in its sample.

• There is no other official source of labour disputes statistics in the UK, therefore the Labour Disputes Inquiry is the only method of capturing such data.

• The data are published monthly in the timeliest manner possible, approximately six weeks after the end of the reference period, and it would not be possible to publish any sooner because of the time needed for companies to provide the information.

• Many businesses and employer associations are collecting detailed workforce information relating to strikes, which they are publishing or sharing, allowing for an improved coverage of responses.

Limitations

• Small disputes may be overlooked because of lack of press coverage.

• The Labour Disputes Inquiry is a voluntary survey, involving a sensitive topic, which can sometimes result in high refusal or low response rates for some months.

• Stoppages are one aspect of labour disputes that has not been comparable over time, with the definition changing and evolving because of the complex nature of disputes (including disputes coordinated between unions or across sectors).

Response rates

Since the reintroduction from January 2022, response rates for the Labour Disputes Inquiry have been varied. We consider a dispute to have a true response if the business has filled out the survey or if a third party has given us a direct response for a business. The latter includes detailed workforce statistics published publicly by some public sector organisations. Any response rolled forward, or imputed another way, is not considered as a true response and will not be included in the response rate for any given month.

The average response rate for 2015 was 84%. In 2022, the average response rate was 56.4% increasing to 62.0% in 2023.

The increase in 2023 can be contributed to improved stakeholder engagement, including relationship building with employer associations and other government departments.

Recent improvements

Temporary suspension and reintroduction

The effects of the coronavirus (COVID-19) pandemic on the Office for National Statistics (ONS) capacity and capability meant that we had to review the existing labour market releases and suspend some labour market publications. On 3 April 2020, we outlined our plans to temporarily suspend the data collection and publication of labour disputes statistics in a statement on the impact of COVID-19 on labour market outputs.

By June 2022, there was a growing interest and need for these statistics from users, therefore we restarted the collection of the Labour Disputes Inquiry. In September 2022, we published the article Labour disputes, UK: July 2022 update and future work to update users on our progress-to-date, to give a snapshot of the data, and outline plans to reintroduce the statistics as part of the UK labour market release from the November 2022 publication.

Apportionment

As the information is usually given by the businesses involved in the action, the data are considered to be accurate. However, it is not always possible to obtain complete information from the business. For example, a business may not be able to give accurate estimates on the full- or part-time breakdown of employees involved in a dispute, or geographical breakdowns of where the disputes are taking place.

In instances where incomplete data are given, these can be apportioned by calculating averages of the business's total employment, as defined on the Inter-Departmental Business Register (IDBR). These averages, aggregated by employment type and region, are then applied to returned data to enable us to produce regional breakdowns and to calculate the number of working days lost. This is not an entirely new methodology but has been automated and refined as part of the reintroduction of these statistics.

Sector assignment

Historically, businesses are asked to describe their main business activity and to define whether they operate in the public or private sector. A recent analysis of responses highlighted inconsistencies in responses from businesses for these questions. In particular, businesses that the ONS defines as public corporations and therefore public sector were often classifying themselves as private sector, thus being misclassified in the data.

To reduce these inconsistencies and improve quality of the data, a new methodology was introduced. During the linking to IDBR for our improved apportionment methodology, we link on variables for legal status and Standard Industrial Classification 2007 (SIC 2007) for each business. Introducing this methodology ensures businesses are classified consistently for labour disputes and in-line with other ONS business surveys.

This methodology is not entirely new; analysts would often validate responses against the IDBR to confirm metrics such as industry and sector. However, this has now been automated in the system and applied to all responses in the sample and not on an ad-hoc basis.

Publication tables

Prior to temporary suspension in April 2020, labour disputes statistics were published monthly across three datasets: LABD01, LABD02 and LABD03, and annually in the Labour disputes in the UK article. As part of the reintroduction, we reviewed the outputs, discussed with stakeholders and amalgamated the monthly datasets into the single Labour disputes in the UK dataset, which meets accessibility requirements.

This section provides a range of information that describes the quality and characteristics of the data and identifies issues that should be noted when using the output.

Relevance

(The degree to which the statistical outputs meet users' needs.)

Labour dispute statistics are used by a wide range of internal and external customers, see Uses and Users in Section 4: Quality summary.

The widespread use and importance of these statistics was particularly evident in 2022, during their temporary suspension. In June 2022, given widespread strikes across the country, there was a growing interest and need for these statistics from users, and so we restarted the collection of the Labour Disputes Inquiry.

Since their reintroduction, these data have been widely reported in the media. They have also been used extensively by other government departments in briefings to ministers and to supplement their own analysis and forecasting in policy making. The additional breakdowns by region and industry included since the reintroduction have aided decision making at a local and sector level.

We plan to regularly review outputs with stakeholders to ensure they remain relevant.

Accuracy and reliability

(The degree of closeness between an estimate and the true value.)

Disputes are picked up from reports in the media and trade union websites, with all businesses with employees on strike surveyed. As a result, the Labour Disputes Inquiry has almost 100% coverage of businesses in its issued sample. However, small disputes may be overlooked because of lack of press coverage. Overall, the survey has good coverage and our sampling method should not impact the accuracy or reliability of estimates.

The Labour Disputes Inquiry is a voluntary survey, involving a sensitive topic, which can sometimes result in high refusal or low response rates for some months.

Information on disputes is widely reported in the media and on trade union websites, including detailed balloting information for some disputes. This published information can be used to validate the data received from businesses and to get reliable estimates for imputation for non-responders.

Analysis has been done on the reliability of using published information to impute for non-responders. In most cases, the expected number of workers on strike published by trade unions is higher than values returned by businesses. We believe this is because of unions publishing figures based on their membership in each business, however, not all members will choose to take strike action or be scheduled to work on the strike day. Applying this method is likely overestimating the number of workers and thus working days lost, but we are confident this method is sufficiently robust and more reliable than excluding these businesses from our data.

The labour disputes estimates are published monthly, six weeks after the reference period. These estimates are considered as provisional and are updated as a revision the subsequent month to take on late returns. Once a year, methods are reviewed, implemented and data revised for a longer period. All revisions are conducted in line with the National Statistics protocol on revisions.

Coherence and comparability

(Coherence is the degree to which data that are derived from different sources or methods, but refer to the same topic, are similar. Comparability is the degree to which data can be compared over time and domain, for example, geographic level.)

The International Labour Organization (ILO) Resolution concerning statistics of strikes, lockouts and other action due to labour disputes: 1993 has provided the framework for the methodology used by the UK. The UK does not satisfy the resolution in all respects, largely because of the practical difficulties in collecting some of the detail (for example, the basis on which the dispute was settled). Nevertheless, the main recommendations of the resolution are followed.

There is no other official source of labour disputes statistics in the UK, therefore this is the only method of capturing such data.

UK data on labour disputes are available back to 1891. Data are available by industry groups, by standard statistical region from 1958 to 1995, by region from 1996 and by cause (from 1959). We also publish estimates on a public and private split, with data dating back to 1996.

Because of the temporary suspension of labour disputes data in response to the coronavirus (COVID-19) pandemic, no data are available for the period of 2020 and 2021.

Stoppages are one aspect of labour disputes that have not been comparable over time. Labour disputes are becoming more complex over time, so we consider there to be a break in series for number of stoppages, from January 2022 data onwards. Therefore the number of working days lost is thought to be a better indicator of the impact of labour disputes than the number of recorded stoppages or the number of workers involved.

The labour disputes statistics cover all UK industry using the Standard Industrial Classification (SIC). Data for 1995 to 2002 have been classified according to SIC 1992, data from 2003 to 2008 have been classified according to SIC 2003, and from 2009 data have been classified according to SIC 2007. The broad industrial classifications used by the Labour Disputes Inquiry, however, have meant that this change in classification had minimal impact on the outputs.

The labour disputes statistics include estimates by public and private sector. Businesses are defined as public or private sector at the time of collection. Unlike estimates of Public sector employment, there are no estimates published in reflection to include or exclude any major reclassifications. As mentioned previously, since January 2022 estimates, businesses are now classified based on their classification on the Inter-Departmental Business Register (IDBR), which may result in a break in series for sector.

Accessibility and clarity

(Accessibility is the ease with which users are able to access the data, also reflecting the format in which the data are available and the availability of supporting information. Clarity refers to the quality and sufficiency of the release details, illustrations and accompanying advice.)

Data are published in the Labour disputes in the UK dataset, which meets the Government Analysis Function guidelines on releasing statistics in spreadsheets.

The main series from the dataset are also released in time series format, alongside other major labour market indicators. This allows those unfamiliar with spreadsheets to pick up the data in an additional format and as a graphic.

Timeliness and punctuality

(Timeliness refers to the lapse of time between publication and the period to which the estimates refer. Punctuality refers to the gap between planned and actual publication dates.)

Estimates from the Labour Disputes Inquiry are published monthly. These data are released in the timeliest manner possible, around six weeks after the reference period; it would not be possible to publish any sooner because of the time needed for businesses to compile the information.

The timeliness of these statistics are consistent with other labour market indicators, which enables us to produce a comprehensive snapshot of the labour market for each release.

For more details on releases, see the release calendar on the ONS website. If there are any changes to the pre-announced release schedule, public attention will be drawn to the change and the reasons for the change will be explained fully at the same time, as set out in the Code of Practice for Statistics.

Concepts and definitions (including list of changes to definitions)

(Concepts and definitions describe the legislation governing the output and a description of the classifications used in the output.)

Working days lost

Working days lost is our main measure of labour disputes. It is defined as the number of days not worked by people involved in a dispute at their place of work. In measuring the number of working days lost, account is taken only of the time lost in the "basic working week". Overtime work is excluded, as is weekend working where it is not regular practice.

Working days lost are calculated by considering how many employees are involved in a dispute and how many days the dispute was ongoing in a given month. For full-time workers, we consider each day of strike as one working day lost. For part-time workers, we consider each day of strike as half a working day lost. This method will not capture the working pattern of all part-time workers but allows us to be consistent over time.

Stoppages

Another indicator of labour disputes is stoppages. The statistics cover stoppages of work in progress in the UK caused by labour disputes between employers and workers, or between workers and other workers, connected with terms and conditions of employment. These include "lock-outs" by employers, where organisations close their sites and will not allow workers to participate in their normal working day and "unlawful" or "unofficial" strikes. The statistics exclude disputes that do not result in a stoppage of work, for example, work-to-rules and go-slows.

A stoppage is defined as a dispute between a single trade union and a single employer. Where a union co-ordinates disputes with several employers, or several unions co-ordinate action with a single employer, this would be counted as multiple stoppages.

In general, we consider a stoppage as a dispute about a specific reason between one union and one company. We would consider the following examples as multiple stoppages.

• One union co-ordinates strikes across several companies (for example, a trade union in the transport sector co-ordinates action across several operators).

• Several unions covering different subsets of employees strike within the same company (for example, the trade union for teaching staff co-ordinate action with the trade union for non-teaching staff within the education sector).

• The union gets a new mandate to strike for a different reason (for example, workers take action related to pay and in the next month take action because of changes in terms or conditions).

• A combination of the above.

As a result of difficulties in defining a stoppage, the number of working days lost is considered to be a better indicator of the impact of labour disputes than the number of recorded stoppages or the number of workers involved. We consider this method of defining a stoppage to be consistent since the reintroduction in 2022, however, there will be some inconsistencies when comparing with data prior to 2022.

Workers involved

The figures for workers involved are for workers directly involved in strike action at the establishment where the dispute occurred. Workers involved in more than one stoppage during the year are counted in the statistics for each stoppage in which they take part. Part-time workers are counted as whole units.

Geography (including list of changes to boundaries)

The headline estimates of labour disputes statistics are UK-level, with regional estimates published in the main datasets.

In most instances, the region in which a dispute takes place is defined by the responding company, where they are asked to name all regions in which the dispute took place. In instances where this is not defined by the company, we assume the dispute took place at all locations where the company operates, according to the company information on IDBR.

Labour disputes statistics are not revised to take account of boundary changes.

How we collect the data, main data sources and accuracy

Labour disputes statistics are estimated using the Labour Disputes Inquiry. The Labour Disputes Inquiry tries to record all strike action that has taken place in the UK each month, except for those disputes involving fewer than 10 workers or lasting for less than half a day.

Disputes are picked up from reports in the mainstream media and trade union websites, with data then collected directly from the company involved in the dispute, where possible. A questionnaire is dispatched to all businesses with employees on strike via email, with contact details sourced from the Office for National Statistics (ONS) Survey Data Collection platform, enabling us to reach businesses via known responders from other surveys.

Prior to the 2022 reintroduction, companies would be contacted via email with a follow-up via telephone. Because of the large number of strikes in 2022 and 2023, there was insufficient resource to follow-up with every company by telephone. We hope to reintroduce this in the future but have improved imputation methodology to account for this in the interim.

Since the 2022 reintroduction, the ONS has built relationships with a number of employer associations, membership organisations and other third-party organisations in main sectors. These relationships allow us to collect data for several companies within a sector, for certain co-ordinated strikes. This improves data collection and response rates for specific industries and reduces the burden on individual businesses.

How we process the data

Returned forms are entered into the system and assigned a unique identifier, allowing us to match the data to other ONS systems at a later stage.

The next stage is to impute for any non-responding companies. There are several methods of imputation used, depending on different factors. If a company has responded to the survey in a recent period, we roll their worker numbers forward and use alongside known strike dates from trade union or media outlets. If a company has not responded before, but is part of a large, coordinated strike, we use known information from similar companies in the dispute to implement a nearest neighbour approach. For all other companies, we use information on the strike or ballot published by the trade unions to make a best possible estimate of the number of people on strike.

The next stage is our apportionment and sector assignment stages. More information on these can be found in Recent improvements in Section 4: Quality summary.

How we analyse and interpret the data

The team then analyse the latest data in the context of the previous month, the previous year and historically. Changes in the level are highlighted, especially any regions or industries with a large proportion of the total working days lost. Trends are presented at the monthly curiosity meeting and developed into lines to take for Labour market overview and for media queries.

How we quality assure and validate the data

There are two main quality assurance stages to labour disputes: assurance of the microdata and assurance of the outputs.

On a microdata level, we ensure that data input into our system is quality assured by another analyst. As the data are sourced directly from companies, they are considered to be accurate. However, information on disputes is widely reported in the media and on trade union websites, including detailed balloting information for some disputes. This published information can be used to validate the data received from businesses and to get reliable estimates for non-responders. We can also compare responses on main business activity and location of strikes to known business information held on the Inter-Departmental Business Register (IDBR), which adds another layer of quality assurance.

At an outputs level, although there are no official forecasts for labour disputes, it is usually possible to predict the direction of movement month-on-month and the highest impacted industry from the extensive research that has been conducted.

How we disseminate the data

Data are published monthly on the Labour market theme day in the Labour disputes in the UK dataset.

How we review and maintain the data processes

As part of the Labour disputes reintroduction in 2022, a full review of the labour disputes business process model was undertaken. This allowed the team to better understand and document the process end-to-end, with a list of essential and desirable improvements logged. The essential improvements were made alongside the introduction (see previous section), while the desirable improvements were logged.

The main desirable actions we would like to implement in the future are: responder research to find out the main issues with data capture, move collection fully online to align with other ONS surveys, and transform the processing system to a more strategically aligned platform.