Creating a strikes database for the UK

Description

This project is funded by The Amiel-Melburn Trust, and The Lipman-Miliband Trust.

Every year the Office for National Statistics (ONS) has traditionally provided annual data on the number of strikes, the numbers of workers involved in these strikes and the number of days not worked in these strikes (roughly the number of workers involved x number of days the strikes lasted). The ONS also provided data by year on the number of ballots undertaken for strikes and industrial action short of a strike (IASOS) as well as their ballot outcomes.

However, the ONS has never provided data on the frequency of industrial action short of a strike (IASOS) such as work-to-rules and overtimes bans which are often referred to as ‘cut-price actions’.

In May 2019, the ONS released its last to-date annual report on strikes, covering the year 2018. Data covered by year the number of strikes, the numbers of workers involved in these strikes and the number of days not worked in these strikes.

For the first time in over a decade, in 2017 the ONS did not also provide annual data on the number of ballots undertaken for strikes and IASOS (and their ballot outcomes). This was because it believed using the data from the Electoral Reform Services (ERS) – now Civica Election Services - was no longer providing a reliably comprehensive picture of the number of ballots undertaken, especially given that other balloting agencies existed and even though ERS was then the main balloting agency.

When the ONS restarted strike data collection in the summer of 2022, it undertook to provide data for the period of the pandemic for which data collection was suspended. So far this has not happened so that there is a gap in ONS strike data covering 1 February 2020 to 31 December 2021.

Just as importantly the ONS now publishes the data on a monthly basis and not also on an annual basis so that there is also an additional absence of annual data for the years 2022, 2023 and 2024 (in addition to the years 2019 to 2021). Although the number of days not worked annually can be calculated from the monthly number of days not worked (2022: 2.514m; 2023: 2.660m; 2024: 0.733m), this is not true of the workers involved nor the number of strikes.

Consequently, there is a need to provide strike data on the missing period and on a more comprehensive annual basis as well as on balloting incidences (frequency, outcomes). This gap cannot be filled by StrikeMap because it is reliant upon union activists and members to voluntarily submit instances to it and because it counts a strike in each individual workplace as one strike rather than part of a single strike which may involve many different workplaces of the same employer. It is not, therefore, a suitable source to fill gaps in data from ONS reporting.

Fortunately, something of a ‘silver lining’ from the ‘black cloud’ of the Trade Union Act 2016 – in the context that unions seeking to have legal standing are required to gain a certificate of independence from the Certification Officer - is that unions registered with the Certification Officer must provide in their annual returns data on all ballots and industrial action undertaken. This also includes data on IASOS so that for the first time ever there is longitudinal data on IASOS, that is, annual time series data on IASOS.

Lastly, the data from the Certification Officer is by union so that the overall data can be disaggregated by union unlike the ONS data.

Research overview

This research covers the six years between 2018 and 2023, providing previously unavailable data by which a more accurate picture can be gained of the state of industrial collective conflict in Britain. This is especially so on the issues of balloting and industrial short of a strike.

Read the briefing note for key findings.