1. Main points

  • The number of working days lost due to labour disputes in 2016 was 322,000, which was more than in 2015 (170,000) and the eighth lowest annual total since records began in 1891.

  • The increase in working days lost in 2016, compared with 2015, was partly attributable to a dispute involving junior doctors in the National Health Service in England, which accounted for 129,000 working days lost (around 40% of the total working days lost for 2016).

  • The two regions with the highest working days lost per 1,000 employees in 2016 were the North East and London.

  • The private sector saw more stoppages in 2016, but the public sector once again had more working days lost than the private sector.

  • There were 154,000 workers involved in labour disputes in 2016; this figure is higher than the record low figure of 81,000 recorded for 2015 but low compared to historical levels.

Back to table of contents

2. Introduction

This article presents analysis of the three main measures of labour disputes (working days lost, stoppages and workers involved) by industry, region, cause, size and duration. The statistics are put into context by considering estimates of working days lost per 1,000 employees. Data are taken directly from the employer or trade union involved after the Office for National Statistics (ONS) has identified disputes from press reports.

This article gives information on labour disputes in 2016 as well as giving comparisons with earlier years. It presents year total figures on labour disputes in 2016 and provides a more in-depth analysis of figures than that published as part of the monthly Labour Market Statistical Bulletin.

Back to table of contents

3. Annual changes

A comparison of labour disputes in 2015 and 2016 is shown in Table 1. There are three core components to the figures: the number of working days lost through stoppages, the number of workers involved in those stoppages and the number of stoppages themselves. (See section 13 “Technical note” for more details on these definitions.) Information on earlier years is available in Dataset Table 1: labour disputes annual estimates, 1891 to 2016.

As shown in Table 1, the mean number of working days lost per stoppage was higher in 2016 than in 2015. The median number was also higher in 2016 than in 2015. The mean value is generally much higher than the median, since working days lost can be greatly affected by large one-off strikes. For this reason, the median tends to give a more typical measure of the average number of working days lost per stoppage.

This can be seen in Figures 1 and 2, where the median number of working days lost per stoppage in a year is more consistent over time than the mean.

Back to table of contents

4. Review of 1997 to 2016

Figure 3 shows a time series of working days lost between 1891 and 2016. It shows that the amount of industrial action has significantly reduced in the last 30 years. This is a stark contrast to the level of action seen when the miners went on strike in the 1970s and 1980s. The 1910s and 1920s saw even greater levels of industrial action culminating in the general strike of 1926.

Table 2 presents labour disputes figures for the period 1997 to 2016, while Figures 4 and 5 illustrate working days lost and the number of stoppages respectively. Figures 4 and 5 show that there are a number of spikes in the time series in years when a particularly large strike took place, showing the impact individual strikes can have on the statistics. The high number of days lost in 2011, for example, was due to two large public sector strikes, while the 2002 figure was due to one very large stoppage in the transport and storage industry. A longer time series showing stoppages and working days lost can be found within Dataset Table 1.

Figure 5 shows that there has been a significant decline in the number of strikes since 1997. Though volatile, the number of working days lost has remained broadly the same over this period. This shows that although the number of stoppages has fallen, large-scale stoppages have become more common.

The second column of Table 2 shows working days lost per 1,000 employees for each year from 1997 to 2016. This converts working days lost into a strike rate, taking into account the size of the labour force. This also enables comparisons to be made across industries and regions that differ in size, as well as adjusting for employment changes in industries and regions over time. The level of employee jobs has generally risen over time. The strike rate in the last 10 years is generally lower than in previous decades. This rise in employment explains strike rates that differ between years when there are no discernible changes in working days lost. The 322,000 working days lost in 2016 is equivalent to 11 working days lost per 1,000 employees, which is lower than the average over the last 20 years (21 working days).

Back to table of contents

5. Industrial analyses

Historically, certain industries have been more prone to strike action than others, and breaking the labour disputes statistics down into separate industries can reveal some interesting patterns and shifts over time. However, it should be noted that comparisons between industries can also be affected by the methodology that is used for compiling the figures. For example, because very small stoppages are excluded from the figures (see section 13 “Technical note” for more details), it is more likely that industry groups with large firms will have disputes included in the statistics. In addition to this, caution must be exercised while carrying out time series analysis due to changes in industrial classifications over time.

Table 3 shows labour disputes statistics for 2016 broken down into 13 industry groups (classified according to the Standard Industrial Classification (SIC) 2007). Human health and social work was the largest sector in terms of number of working days lost. However, this high figure for working days lost was largely due to a single dispute involving junior doctors in the National Health Service in England. This industrial group only accounted for 4 out of the 101 strikes recorded for 2016.

In terms of the number of strikes, the largest sectors were transport and storage (23 strikes and 49,100 working days lost) and education (21 strikes and 105,400 working days lost). Strikes in the transport and storage sector mainly occurred within public transport.

Figure 6 shows working days lost per 1,000 employees for the public administration and defence, education, and transport and storage sectors over a 10-year period. It shows how the strike rates for this sector compare with the total strike rate for the economy as a whole.

Back to table of contents

6. Regional analysis

Dataset Table 2 shows regional strike rates between 2007 and 2016, with a further breakdown of the figures for 2016 by industrial grouping. When interpreting these figures, it is important to bear in mind that the industrial composition of employment in a region is a major influencing factor on the scale of labour disputes it experiences. The regions with the highest strike rate in 2016 were the North East and London (16 working days lost per 1,000 employees). All of the regions except Northern Ireland and Wales showed an increase compared with 2015. Looking at the figures over the 10-year period from 2007 to 2016, the North East has shown the highest strike rate and the East of England has shown the lowest.

Figure 7 compares working days lost per 1,000 employees between 2015 and 2016. Most regions showed an increase in strike action over this period. The North East and London had the largest strike rate in 2016 (16 working days lost per 1,000 employees). Northern Ireland had the lowest strike rate in 2016 (4 working days lost per 1,000 employees), compared with 2015 when it had the highest (21 working days lost per 1,000 employees).

Back to table of contents

7. Cause of disputes

Dataset Table 3 shows stoppages in 2016 by principal cause and industry grouping. Disputes over pay also include stoppages over feared or alleged reductions in earnings, as well as disputes over the size of pay increases. Disputes over pension provisions are also classified as disputes over pay.

In 2016 the main cause of industrial action, in terms of working days lost, was duration and patterns of hours worked, which accounted for 134,800 working days lost (41.8% of all working days lost in 2016). Most of the 134,800 working days lost due to duration and pattern of hours worked were attributable to a dispute involving junior doctors in the National Health Service in England.

However, in terms of the number of workers involved, the main cause of industrial action in 2016 was redundancies, which accounted for 86,100 workers (55.9% of all workers involved in industrial action in 2016). Figure 8a and Dataset Table 4 give information on working days lost by cause of dispute in each year. The figures are often dominated by one or two very large strikes, which can make comparisons over time difficult.

Looking at the figures from 2007 to 2015, pay was the main cause of disputes in each year except for 2009 and 2010 when, following the economic downturn, redundancies were the main cause. However, the figures for 2016 show a different picture with most disputes being accounted for by causes other than pay or redundancy. Figure 8b breaks down this “Other” category into more detail and shows that the main cause of disputes within this category was duration and patterns of hours worked. This was mainly due to a dispute involving junior doctors in the National Health Service in England.

Back to table of contents

8. Disputes by duration

Labour disputes statistics cover the number of days that strike action took place, not the number of days the parties involved in the dispute were actually in disagreement.

Table 4 and Figure 9 show the duration of the 101 stoppages in progress in 2016. These show that around one-third of the number of stoppages (34 out of 101) lasted for only 1 day. These accounted for 66.1% of workers involved and 81,000 working days lost (25% of all working days lost).

There were 13 stoppages that lasted for more than 10 days, accounting for 44,400 working days lost (13.8% of all working days lost). However, these stoppages involved only 3,500 workers (3.5% of all workers involved in 2016).

Back to table of contents

9. Disputes by size

Table 5 shows disputes in 2016 by size and Figure 10 shows how the 101 stoppages recorded for 2016 break down by duration. Out of the 101 stoppages in 2016 over half (63) had less than 500 days lost; however these stoppages lasting less than 500 days accounted for only 3.4% of all working days lost in 2016. This shows the impact that large strikes can have on the figures.

Back to table of contents

10. Disputes by public and private sector

Figures 11a and 11b illustrate the breakdown of working days lost and the number of stoppages between the public and private sectors. The figures are also shown in Table 6. Over the 10-year period shown in Figure 11a there have been far more working days lost in the public sector than in the private sector, even though the private sector is much larger. The private sector has had fewer working days lost than the public sector in every year since 1999.

Looking at the figures for 2016, there were more working days lost in the public sector (243,000) than the private sector (79,000) but there were more strikes in the private sector (60) than in the public sector (41). In the public sector there were 45 working days lost per 1,000 employees, while in the private sector there were only 3 working days lost per 1,000 employees.

Back to table of contents

11. Trade union ballots

Annual trade union ballot data from 2002 to 2016 are presented in Tables 7 and 8. The number of ballots has fallen to 488 this year, the lowest since records from Electoral Reform Services began in 2002.

The number of ballots resulting in a vote for strike action in 2016 shows a decrease on the previous year, from 503 in 2015 to 436 in 2016. The time series for trade union ballots is illustrated in Figure 12. This chart shows that a high percentage of ballots calling for strike action result in “yes” votes, with 93.6% in 2016.

Back to table of contents

12. Quality and methodology

The Labour Disputes Quality and Methodology Information document contains important information on:

  • the strengths and limitations of the data and how it compares with related data

  • uses and users

  • how the output was created

  • the quality of the output including the accuracy of the data

Back to table of contents

13. Technical note

Coverage

Information regarding labour disputes within the UK is collected by the Office for National Statistics (ONS) from a variety of sources. The information is collected directly from the employer or trade union involved after we have identified disputes from press reports. We publish figures on labour disputes each month. They appear in the Labour Market Statistical Bulletin.

Definition of stoppages

The statistics cover stoppages of work in progress in the UK during a year caused by labour disputes between employers and workers, or between workers and other workers, connected with terms and conditions of employment. A distinction can be drawn between stoppages that started in the current year and those that started in earlier years.

A stoppage in progress is defined as a dispute that has continued from a previously recorded dispute by the same organisation and for the same cause. Prior to 2015, a dispute was counted as a new stoppage if there was a gap of more than 1 month between instances of industrial action. From 2015, disputes with a gap of more than 1 month between instances of industrial action are counted as a single stoppage.

The statistics exclude disputes that do not result in a stoppage of work, for example, work-to-rules and go-slows; this is because their effects are not quantifiable to any degree of certainty. Stoppages involving fewer than 10 workers or lasting less than 1 day are also excluded unless the total number of working days lost in the dispute is 100 or more.

Stoppages over issues not directly linked to terms and conditions between workers and employers are omitted, although in most years this is not significant. For example, in 1986 one stoppage was considered to be political (a protest in the coal industry against the visit of an MP) and it was excluded from the figures. The total working days lost amounted to less than 1,000. The next known dispute to be excluded was in 1991. This involved a boycott by self-employed market traders who, after increased rent and changes to the market rules, kept their stalls closed for about 20 weeks.

The statistics include “lock-outs”, that is, where an employer prevents their employees from working by refusing entry to the place of work, and “unlawful”, that is, unlawfully organised strikes. However, no distinction is made between a “strike” and a “lock-out” or between “lawful” and “unlawful” stoppages. This is principally because of the practical difficulty in deciding in which category a particular stoppage falls. It was for similar reasons that a distinction between “official” and “unofficial” disputes was no longer made after 1981.

Working days lost

Working days lost are defined as the number of days not worked by people as a result of their involvement 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 a regular practice.

Where an establishment is open every day, and runs two or more shifts, the statistics will record the number of working days lost for each shift. In recording the number of days lost, allowance is made for public and known annual holidays, such as factory fortnights, occurring within the strike's duration. No allowance is made for absence from work for such reasons as sickness and unauthorised leave.

Where strikes last less than the basic working day, the hours lost are converted to full-day equivalents. Similarly, days lost by part-time workers are converted to full-day equivalents. The number of working days lost in a stoppage reflects the actual number of workers involved at each point in the stoppage. This is generally less than the total derived by multiplying the duration of the stoppage by the total number of workers involved at any time during the stoppage, because some workers would not have been involved throughout.

In disputes where employers dismiss their employees and subsequently reinstate them, the working days lost figure includes those days lost by workers during the period of dismissal.

For disputes where employers dismiss their employees and replace them with another workforce, the statistics cannot assume that working days lost by the sacked workers continue indefinitely. In such cases, the statistics measure the number of days lost in terms of the size of the replacement workforce. For example, where an employer initially recruits 100 workers and wishes to build up to 300, the number of working days lost on the first day will be 200 and will then progressively reduce on subsequent days, eventually to zero when the new workforce reaches the target of 300.

Figures given for working days lost per 1,000 employees use employee jobs for each year taken from our most recent estimate of Workforce Jobs.

Number of stoppages

There are difficulties in ensuring complete recording of stoppages, in particular for short disputes lasting only a day or so, or involving only a few workers. Because of this recording difficulty and the cut-off applied, 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.

Workers involved

We aim to record the number of workers that are involved at any time in the stoppage. For example, consider a 3-day strike where there were 200 workers involved on the first day; 300 on the second day, of whom 100 were involved for the first time; and 200 on the third day, of whom 50 were involved for the first time. The total number of workers involved in the dispute is 350 – the sum of all those involved on the first day, and those joining for the first time on subsequent days. However, the number of workers taking strike action for the first time during a dispute cannot always be easily ascertained. In such cases, the statistics record the highest number involved at any one time (300 in the above example). Take another example, where there are 200 workers involved in a stoppage on each of days one, two and three. It may be necessary to assume that there were a total of 200 workers involved, although it is possible, but unlikely, that as many as 600 workers could have been involved. For this reason, the statistics may under-estimate the number of workers involved in a dispute. However, the estimate of the number of working days lost is unaffected by this consideration.

Ballot data

Although the Electoral Reform Service (ERS) conduct the majority of industrial action ballots for trade unions in the UK, other organisations also conduct ballots. In addition to this, ballots with less than 50 union members do not have to appoint a scrutineer.

Back to table of contents

Contact details for this Article

Richard Clegg
richard.clegg@ons.gsi.gov.uk
Telephone: +44 (0)1633 455400