Health and Safety
Executive / Commission
Statistics
These statistics cover accidents that are covered by the provisions of RIDDOR (Reporting of Injuries Diseases and Dangerous Occurrences Regulations, 1995). They therefore cover accidents that are reportable to either HSE, the relevant Local Authority or the Office of the Rail Regulator. For the vast majority of such reportable accidents the Health and Safety at Work etc Act, 1974 is the main legislation applicable. During the course of a year there will also be many other fatal accidents that, although they take place during working hours, are primarily covered by other legislation. A key exclusion from the HSE figures is road traffic accidents to workers who travel as part of their work (e.g. lorry drivers and travelling sales representatives). For these the police have primacy. Similarly the CAA's Air Accident Investigation Branch have primacy for air accidents and the Marine Accident Investigation Branch of the MCA have primacy for incidents involving seafarers.
The figures for 2008/09 are, at this stage, provisional. They will be finalised in June 2010 following any necessary adjustments. This arrangement allows for the fact that the investigations which follow fatalities are often complex and take considerable time. In the course of these investigations new facts can emerge to affect judgements on issues such as whether the accident was work-related or whether the worker was based at the site of the accident. This means that initial views regarding the reportability of the accident or the industrial sector to which should be assigned can prove to be incorrect. The delay of a year in finalising the figures allows for such matters to be fully resolved in the light of formal interviews with all relevant witnesses, forensic investigation and coroners rulings. There is also the ‘year and a day rule’ under Regulation 4 of RIDDOR to cover situations where someone dies after some delay following an accident.
The following table shows provisional information on the numbers of fatal injuries for the twelve months 1/04/2008 to 31/03/2009.
| Main industry: | Agriculture | Extractive & utility supply | Manufacturing | Construction | Services1 | All Industries |
|---|---|---|---|---|---|---|
| Employees | 10 | 6 | 32 | 33 | 48 | 129 |
| Self employed | 16 | - | - | 20 | 15 | 51 |
| Workers* | 26 | 6 | 32 | 53 | 63 | 180 |
| Members of the public | 5 | 1 | 2 | 4 | 403 | 415 |
| Total fatalities | 31 | 7 | 34 | 57 | 466 | 595 |
The following table shows a regional breakdown of the fatal injuries to workers.
| Government Office Region (GOR) | All Industries |
|---|---|
| South West | 22 |
| South East | 15 |
| London | 20 |
| East | 12 |
| East Midlands | 11 |
| West Midlands | 12 |
| Yorkshire | 24 |
| North West | 22 |
| North East | 9 |
| Wales | 5 |
| Scotland | 26 |
| Not known | 2 |
| Total | 180 |
Changes in the size of the workforce will impact on the number of fatalities in any one year. Therefore, when making any comparisons either on a year-to-year basis or between one sector and another it is important to look at rates of fatal injury per 100 000 in the workforce. The following table shows the provisional rates of fatal injury to workers in the main industrial sectors. These are expressed as the number of fatal injuries per 100 000 employees, self employed or workers, as appropriate.
| Main industry: | Agriculture | Extractive & utility supply | Manufacturing | Construction | Services | All Industries |
|---|---|---|---|---|---|---|
| Employees | 4.0 | 3.4 | 1.2 | 2.6 | 0.2 | 0.5 |
| Self employed | 7.7 | - | - | 2.3 | 0.6 | 1.3 |
| Workers* | 5.7 | 3.1 | 1.1 | 2.4 | 0.3 | 0.6 |
For the purposes of comparison with Table 1 the following table shows the average numbers of injuries per year for the previous 5 years.
| Main industry: | Agriculture | Extractive & utility supply | Manufacturing | Construction | Services1 | All Industries |
|---|---|---|---|---|---|---|
| Employees | 14 (22) | 7 (9) | 35 (32) | 51 (53) | 67 (62) | 175 (178) |
| Self employed | 27 (24) | 1 (-) | 2 (1) | 19 (19) | 9 (11) | 57 (55) |
| Workers* | 40 (46) | 7 (9) | 37 (33) | 70 (72) | 76 (73) | 231 (233) |
| Members of the public | 5 (2) | 2 (2) | 3 (2) | 5 (3) | 372 (362) | 386 (371) |
| Total fatalities | 46 (48) | 9 (11) | 39 (35) | 75 (75) | 448 (435) | 617 (604) |
1 The figures for services include railway incidents reported to the Office of Rail Regulation (ORR). For member of the public figures, the yearly average of 386 becomes 86 if railway-related incidents are excluded, and the 2007/08 figure of 371 becomes 70.
The finalised total for worker deaths in 2007/08 represents a net increase of 4 compared to the provisional figure of 229 released last year. This is due to late reporting and the time taken to conclude some fatal investigations.
For the purposes of comparison with Table 2 the following table shows the average numbers of fatal injuries per year for the previous 5 years.
| Government Office Region (GOR) | All Industries |
|---|---|
| South West | 20 (28) |
| South East | 24 (22) |
| London | 23 (26) |
| East | 18 (19) |
| East Midlands | 15 (18) |
| West Midlands | 23 (22) |
| Yorkshire | 16 (16) |
| North West | 33 (22) |
| North East | 9 (7) |
| Wales | 16 (18) |
| Scotland | 30 (31) |
| Not known | 6 (4) |
| Total | 231 (233) |
For the purposes of comparison with Table 3 the following table shows the average rates of fatal injury per year for the previous 5 years.
| Main industry: | Agriculture | Extractive & utility supply | Manufacturing | Construction | Services | All Industries |
|---|---|---|---|---|---|---|
| Employees | 6.1 (9.1) | 4.3 (5.2) | 1.2 (1.1) | 4.2 (4.2) | 0.3 (0.3) | 0.7 (0.7) |
| Self employed | 14.7 (12.5) | 10.5 (-) | 0.8 (0.4) | 2.3 (2.2) | 0.3 (0.4) | 1.5 (1.4) |
| Workers* | 9.9 (10.6) | 4.2 (4.8) | 1.1 (1.1) | 3.4 (3.4) | 0.3 (0.3) | 0.8 (0.8) |
Fatal injuries at work are thankfully rare events. As a consequence, basic statistical principles dictate that the annual count is highly subject to chance variation. Moreover, the effect of this chance variation can be estimated to give an indication of the amount the figure would fluctuate if the inherent dangerousness of work conditions were to stay unchanged from one year to the next. For example, it can be estimated that this year’s count of 180 could have been anywhere between 153 and 207 based on chance alone. This theoretical point is borne out at a practical level when the causal factors behind individual fatalities are examined. It is often found that an unfortunate set of chance events have occurred together with shortcomings in safety precautions.
Given the above, it is necessary to make some allowance for the play of chance when seeking to use data on fatalities to make inferences about trends in the inherent dangerousness of work conditions. The best way of doing this is to analyse data over a number of years. In this way fluctuations due to chance will be evened out.
Chart 1 shows the trend in the rate of fatal injury over the last seventeen years. It can be seen that the rate for 2008/09 provides evidence that the underlying downward trend that was apparent in the late 1990s and the early part of this decade is still continuing. Prior to the availability of the rate for 2008/09 the figures for the preceding five years were starting to suggest that the downward trend was levelling off. Therefore, this year’s data has changed the picture in an important way. Moreover, the rate for 2008/09 represents a statistically significant decrease compared to the average rate for the previous five years.
Given that the figure for 2008/09 is out of line with the data for the previous five years there is reason to be cautious about any conclusions. However, the most likely explanation for the striking drop in the figures this year is that the play of chance has fallen in a highly favourable way. This has come after two years when based on the underlying trend it would appear that the figures were very much in the upper range of what might be expected. It is also worth mentioning that the recession could be having an effect. There is indeed research evidence to show that previous recessions have led to a decline in non-fatal injury rates (see research report 386 and HSE strategy launch Background paper: the economic evidence). However, at this early stage, any such linkage can only be tentative supposition.
Statistics of fatal injuries given in these pages are compiled primarily from notifications made under RIDDOR (see Background section above).
Since 1 April 2006, enforcement of safety on railways has been the responsibility of ORR, and they have provided HSE with the relevant figures since that date. Prior to this date, enforcement was the responsibility of HSE’s Railways Inspectorate. More information on these arrangements can be found at http://www.hse.gov.uk/railway/index.htm. ORR publish detailed railways-specific safety statistics, which can be found at http://www.rail-reg.gov.uk/server/show/nav.1210.
The term ‘workers’ describes both employees and self-employed combined. Those on a training scheme, or on work experience, are classified here as employees. References to a ‘member of the public’ are those persons killed as a result of an accident, which has arisen out of or in connection with work activity, although they are not ‘at work’ themselves. For example, some of the following member of the public fatalities may be notifiable to the enforcing authority: customers on retail premises; residents in residential care homes; passengers on trains; trespassers on railways.
Fatal diseases are outside the scope of these figures, irrespective of whether the disease is notified under RIDDOR.
The following are clarifications of the fatal injuries that are generally included or excluded in these statistics. It must be stressed that this is a generalised view and that each record of a fatal injury is considered on an individual basis. Although a particular fatal incident may fall outside the scope of these figures, the relevant health and safety enforcing authority (HSE, LA, or ORR) may still have an interest.
Fatal injuries included in the statistics are primarily those deemed to be reportable under RIDDOR. These include workers involved in road maintenance or roadside refuse collection and workers or members of the public killed on railways (including those killed as a result of suicide or trespass); plus the following:
Whilst deaths as a consequence of physical violence done to a person at work are included, there is anecdotal evidence to suggest that compliance with the reporting requirement within RIDDOR may be patchy. Because such deaths are primarily covered by other legislation, and the police have primacy in the investigation, the need to report under RIDDOR may sometimes be overlooked.
Fatal injuries excluded from the statistics are mainly those deemed non-reportable under RIDDOR. The most important exclusions are as follows:
The Standard Industrial Classification has been revised several times since first introduced in 1948. The currently used version, SIC 2003, made minor revisions to SIC 1992. Standard Occupational Classification 2000 (‘SOC 2000’), also the responsibility of ONS, attempts to classify jobs in terms of their skill level and skill content, and is used by many other government organisations.
Rates of fatal injury for employees produced by HSE are based on employee job estimates produced by the ONS from the Short Term Employment Survey. Such estimates are subject to revision based on information from the Annual Business Inquiry. When HSE finalises the provisional statistics, rates are revised using the employment data available at that particular time. Rates of fatal injury are not normally revised to incorporate subsequent revisions to employment estimates by the ONS.
Rates of fatal injury for the self-employed are based on self-employed job estimates from the Labour Force Survey (LFS), produced by ONS. Second jobs are included in the estimates used.
The smooth trend line shown in Chart 1 is calculated by choosing an "expected" fatal injury count for each year such that:
The smoothness of the rate series is measured (inversely) by the sum of its squared second differences. Maximising smoothness of the rate series expresses a view that the underlying state of the world (in respect of the inherent dangerousness of the nation's workplaces) changes gradually from year to year.
The theoretically predicted year to year variation is based on an assumption that the annual count of fatal accidents follows a Poisson distribution, and condition 2 is implemented as a constraint on the overall chi-squared value to be less than or equal to its expected value. The potential flexibility of the fitting process (separate fitted value assigned to each observation, but subject to maximal smoothing) means that the number of degrees of freedom lost in the fitting process is not precisely determined. It follows that the expected value of the total chi-squared (i.e. the number of observations minus the degrees of freedom due to the fitted model) is also somewhat indeterminate. Since the resulting fit is close to linear we have assumed it accounts for two degrees of freedom.
Names and details of fatalities that occurred in 2008/09 in HSE-enforced sectors are available at the following link: http://www.hse.gov.uk/foi/fatalities/2008-9.htm
The fatal injuries figures representing the full-year are National Statistics.
National Statistics are produced to high professional standards set out in the National Statistics Code of Practice. They undergo regular quality assurance reviews to ensure that they meet customer needs. They are produced free from any political interference.