HSE banner
Picture of broken reading glasses situated on a desk

Violence at work

Definitions and additional sources

Violence to employees can cause pain, distress and even disability or death. Physical attacks are an obvious danger. However, serious or persistent verbal abuse can be a significant problem too, as it can cause damage to employees’ health through anxiety and stress. See also HSE’s pages on work-related stress.

HSE defines violence as “any incident in which the person is abused, threatened or assaulted in circumstances relating to their work”.

However, with a wide range of available data sources on the extent of violence at work and its consequences, it is important to understand that definitions of violence will vary from source to source, with no source covering the entire range of violent incidents included within HSE’s definition.

The three main sources of data on the overall British working population are:

These sources have been used to produce the majority of the detailed statistics on these pages. However, a number of other bodies have collected relevant information specific to their sectors, which provides a useful insight into the nature or scale of the problem within some of the most at-risk occupations or industries. Some examples follow.

The British Retail Consortium includes measures of workplace violence in its annual Retail Crime Survey4. The 2007 report (Retail crime survey 2007) showed that in 2006, the survey estimates that there were 6 incidents of physical violence per 1000 employees in the retail sector, 16 threats per 1000 employees and 15 incidents of verbal abuse per 1000 employees.

In 2008, the British Medical Association (BMA) produced a report into violence in the workplace, gathering the experience of doctors in Great Britain. The report can be downloaded from the BMA website5. Headline findings included the fact that violence was a problem for almost half of all respondents, whether working as a GP or in a hospital setting. One in five doctors reported an increase in violence over the past year, while the majority reported no change.

The NHS Counter Fraud and Security Management Service Division (CFSMS) collects annual statistics from NHS trusts giving total number of assaults and number of assaults per 1000 staff. The figures can be found on the CFSMS website6. A staff survey was conducted by the same organisation in 2006 (NHS Violence Survey 2006) and showed that over half of all respondents (54%) had been verbally abused in the past 12 months, with 17% threatened with physical violence and 6% actually physically assaulted. It should be noted that these figures are not representative of the NHS, since the sample was selected partly to seek the views of those with face to face contact with the public in open-access public environments. These staff groups were thought to be particularly at risk of abuse by the public.

The NHS violence survey had been conducted previously in 2003 and 2004 and the year on year comparison suggests violence is increasing. However there is also an insight into reporting trends, which may help to explain the levels of rises in violence from sources relying on reporting mechanisms. In 2003, more than half (54%) of those saying they had suffered verbal abuse said they had not reported it, while in 2006 this was down to just under a third (31%). Similarly, 66% of those saying in the 2003 survey that they had been physically assaulted had not reported the incident while in 2006 the figure had fallen as far as 13%.

Call centre staff are subject to the risk of verbal abuse. Research by the Call Centre Association (CCA)7 in 2005 (Click for CCA Survey 2005) investigated the extent to which members had experienced abuse and how the issue was being dealt with.

The Teacher Support Network8 is an independent charity supporting workers in the education sector. In 2007 they produced a report on the impact of violence and disruption in schools on teacher wellbeing, (Teacher Support Network Survey 2007) which includes some figures on violence against teachers. The survey was self-selecting and so the figures cannot be seen as representative of the sector, however they are still worryingly high.