Prepared by Office for National Statistics for the …

1 health and Safety Executive Changes in shift work patterns over the last ten years (1999 to 2009) Prepared by Office for National Statistics for the health and Safety Executive 2011 RR887 Research Report Crown copyright 2011 First published 2011 You may reuse this information (not including logos) free of charge in any format or medium, under the terms of the Open Government Licence. To view the licence visit , write to the Information Policy Team, The National Archives, Kew, London TW9 4DU, or email Some images and illustrations may not be owned by the Crown so cannot be reproduced without permission of the copyright owner. Enquiries should be sent to ii health and Safety Executive Changes in shift work patterns over the last ten years (1999 to 2009) Matthew Steel Office for National Statistics Government Buildings Cardiff Road Newport NP10 8XG The health and Safety Executive (HSE) have approached the Office for National Statistics (ONS) to carry out an updated analysis of shift work in the UK based on the last ten years worth of data (1999 to 2009) from the Labour Force Survey (LFS).

2 This paper examines changes in shift work patterns across the UK workforce over the last decade. The report investigates differences in the reporting of shift work based on gender, age, region, occupation and industry and also compares engagement in the different types of shift patterns. This report and the work it describes were funded by the health and Safety Executive (HSE). Its contents, including any opinions and/or conclusions expressed, are those of the author alone and do not necessarily reflect HSE policy. HSE Books Background The health and Safety Executive (HSE) have approached the Office for National Statistics (ONS) to carry out an updated analysis of shift work in the UK based on the last ten years worth of data (1999 to 2009) from the Labour Force Survey (LFS).

3 This paper examines changes in shift work patterns across the UK workforce over the last decade. The report investigates differences in the reporting of shift work based on gender, age, region, occupation and industry and also compares engagement in the different types of shift patterns. Summary of Key Findings The overall proportion engaged in shift work has remained consistent across the past decade. Where gender differences were identified, it continues to be males rather than females that are more likely to be engaged in any form of shift work pattern (irrespective of whether the frequency reported is most of the time or some of the time ). Where we compare the proportional data between 1999 and 2009, a downward trend is evident among males in the number that have reported working shift work most of the time across the time series. Where we consider age, it continues to be individuals grouped within the youngest age band (16 to 24 years of age) that are most likely to report engaging in shift work patterns.

4 The number of males within the youngest age band engaging in shift work most of the time has increased across the decade in question where we compare the proportions between 1999 and 2009. Although an increase was also seen in reported shift work most of the time among females in the youngest age group, this increase was not statistically significant. Across the time series it was individuals working within the countries Wales and Scotland that were most likely to report engaging in any pattern of shift work. It was consistently the country Northern Ireland that displayed the lowest reported levels of any frequency of shift work. Where we consider government Office regions, the highest levels of reported engagement in shift work were in the North East of England and Scotland. The lowest levels of shift work participation were reported in Northern Ireland and London. Those working within the Personal protection occupations remain the most likely to engage in any form of shift work pattern.

5 Individuals employed within the transport and communication industry classification were the most likely to be working shift patterns. The most prevalent form of shift work pattern reported by those engaging in shift work was the two shift system. This pattern often comprises of two shifts of eight hours in duration which are usually alternated weekly or over longer intervals. 2 Methodological considerations In attempting a time-series analysis of LFS data spanning a decade, a number of issues required careful consideration.

6 One particular issue concerned the re-weighting exercises which have been performed on the LFS datasets in the years of 2007 and 2009. For current analytical purposes, datasets between 1999 and 2006 have been weighted using the 2007 person level weights. Datasets between 2007 and 2009 have been weighted using the 2009 person level weights. Further consideration was necessary following revisions to the structure and frequency of the LFS. In 2006 the survey switched from a seasonal-quarter to a calendar-quarter basis, in order to fall in line with EU Regulation. As a consequence of the shift work questions appearing only in a single LFS quarter, the realignment from seasonal to calendar quarters altered the months where the question would be asked. Where the variable was originally asked within the seasonal quarter of spring (March May), from 2006 it would be administered in the calendar quarter of April-June. A major change such as this clearly presents difficulties for analysis of trend data.

7 The aforementioned re-weighting exercises were performed only on a limited series of LFS back data on a calendar-quarter basis. As a direct result, March data on shift work were omitted from datasets prior to 2006. Where only two of the three months worth of shift work data were present, the decision was made to apportion the observed ratio in the two month s for which data was present to the month where data was missing. The current paper therefore makes the assumption that the reporting of shift work patterns would remain consistent across the three months and that similar reporting would occur in June as it would in April and May. In 1999 LFS response options for the frequency of shift work at the variable SHIFTWK altered from occasionally and sometimes to usually and most of the time . The variable name altered to reflect this change from SHFTWK to SHFTWK99. Although this should not impact on current analysis, any attempt to compare current analysis with the previous report (McOrmond 2004) may prove difficult due to the potential discontinuities (Please refer to Appendix i for further information about the shift work questions which are administered on the LFS).

8 Background shift work Shift work can be thought of as a work pattern under which different groups of workers succeed each other in the same job at the same site or location. At the beginning of the shift work, the work of the previous shift is taken over and at the end of the shift, the work is handed over to the next shift. Shift work can often entail working during unsocial hours in the early morning, at night or during the weekend and the weekly rest days do not always coincide with normal rest days ( weekends). 3 Shift work can be prevalent in firms or organisations which remain open and working or providing services outside of the expected working hours.

9 We often tend to treat expected working hours as those conforming to the parameters of 8am and 6pm on weekdays. Individuals whose work pattern may vary to suit day-to-day needs but, who are not part of any regular schedule do not fulfil the LFS criteria for shift work. Some respondents, especially on covering duties in the health sector, may not recognise that they are working as part of a larger shift pattern. LFS interviewers are trained to be aware of this potential under-reporting and encouraged to probe to ensure complete coverage. Respondents who reported engaging in shift work were able to categorise this frequency under the response options most of the time or occasionally . Where respondents were unable to distinguish between these two options, advice was provided by interviewers to treat more than fifty per cent of their employment hours as most of the time . Frequency of shift work As shown in Table 1, the overall proportion of respondents who reported engaging in shift work has remained largely consistent across the previous decade.

10 Having said this, there does appear to be a slight reduction in the overall number of respondents reporting working shift patterns most of the time . This downward trend, where we compare the proportions engaging in shift work most of the time between 1999 and 2009, is only statistically significant among men. Table 1: Proportions of people in employment working shift patterns by sex; United Kingdom; 1999 to 2009 Per Cent Males Females All 1,2 3 1,2 3 1,2 31999 2009 1999 2009 1999 2009 Most of the time Occasionally Never Unweighted base 22,465 22,448 19,939 20,978 42,404 43,426 1 Weighted base (000s) 9,559 12,942 8,071 11,331 17,630 24,273 1 Weighted to 2007 population totals 2 Following realignment from seasonal to calendar quarters, data for June was assumed to follow the same pattern as that of April and May. 3 Weighted to 2009 population totals In previous analysis by McOrmond (2004), the proportion of individuals never working shift patterns were compared in an attempt to identify changes in shift work.