Chapter 3: Personal well-being and crime

This chapter presents findings from the 2012/13 and 2013/14 Crime Survey for England and Wales (CSEW) on personal well-being. It looks at how victims of crime rate their personal well-being and compares this with non-victims, including looking at different types of crime. As there is likely to be a complex relationship between personal well-being, demographic characteristics, and whether or not a person was victimised, the contributory effect of the victimisation is uncertain.

• Those who were victims of crime in the previous 12 months reported significantly lower personal well-being than non-victims for measures relating to “life satisfaction”, “life worthwhile”, and “happiness yesterday”. There was no significant difference for “anxious yesterday”

• When looking at demographic breakdowns, there was a strong association between young and single victims and low personal well-being. It is possible that this is due to differences in the profile of victimisation, where young people are more likely to be a victim of a violent crime

• Victims of violence with injury gave lower personal well-being ratings than victims of violence without injury. For both measures the differences between victims and non-victims was statistically significant for all personal well-being measures, with the exception of “low anxiety yesterday”

• Of the various theft offences, domestic burglary and theft from the person had the strongest association with a victim’s personal well-being

• Across all four measures, those that believe they are likely to be a victim of crime in the next 12 months had lower personal well-being than those who thought they were unlikely to be a victim

The Office for National Statistics (ONS) has developed a programme aiming to produce accepted and trusted measures of national well-being. This is part of a wider initiative in the UK and internationally to look beyond traditional measures such as Gross Domestic Product (GDP) with regards the state of a country, and consider what really matters to people. ONS regularly monitor 41 different headline measures in areas such as the natural environment, our relationships, health, what we do, where we live, personal finances, the economy, education and skills and governance to measure the progress and well-being of the nation. These measures include four standardised questions which measure personal well-being.

This chapter compares the personal well-being of victims and non-victims of crime, and how this varies by a range of demographic characteristics. It then examines variations by different crime types – for example, how do victims of burglary compare with non-victims? Additionally, the chapter draws on other CSEW measures to illustrate further the emotional impact a crime can have on a victim. It is also possible to look at the personal well-being of respondents who believe they are highly likely to be a victim of a crime over the next year, compared with those who think they are highly unlikely to be so.

ONS began measuring personal well-being in April 2011, as part of the Annual Population Survey (APS). The APS is obtained by combining results from the Labour Force Survey (LFS) and data from the APS ‘boosts’ samples in England, Scotland and Wales¹. These 4 questions have been developed:

1. Overall, how satisfied are you with your life nowadays?

2. Overall, to what extent do you feel the things you do in your life are worthwhile?

3. Overall, how happy did you feel yesterday?

4. Overall, how anxious did you feel yesterday?

People are asked to give their answers on a scale of 0 to 10, where 0 is “not at all” and 10 is “completely”. An aggregate score of 7 or more is described as high (score of 7 or 8) or very high (score of 9 or 10), except for the anxious yesterday question, which is an aggregate score of 3 or less (with a score of 2 or 3 meaning low or 0 or 1 meaning very low)².

The results from the APS are published annually, as part of the Measuring National Well-being programme. The most recent one of these was for the 2013/14 year, and was published in September 2014.

Notes for how personal well-being is measured

1. The ‘Labour Force Survey: User Guide, Volume 1 – LFS Background and Methodology’ has more information.

2. This question is referred to as “Low anxiety yesterday”, because an aggregate score of 3 or less represents those that reported low levels of anxiousness.

The 4 questions relating to personal well-being have been asked in the self-completion section of the CSEW since 2012/13. They have been included to understand the association between personal well-being and experience of crime.

While the standard personal well-being questions from the APS have been used in the CSEW, there are some important differences in the way in which they are administered compared with the APS. Rather than asked as part of the face-to-face or telephone interview (as they are on the APS), the questions come at the beginning of the self-completion section of the survey, following the core questions on perceptions and experience of being a victim of crime, rather than early on in the interview (as they do on the APS). The different mode of interview, survey setting and ordering of questions is likely to have an effect on how respondents answer the well-being questions. Further details on the differences in the personal well-being data from the APS and the CSEW can be found in the ‘Differences between published personal well-being figures and the CSEW’ section.

The primary reason for including the questions in the CSEW was not to provide an estimate of personal well-being as the APS fulfils that duty; it was to explore associations between personal well-being and experience of crime. However, the well-being of respondents is measured at the time of the CSEW interview, and after having any experience of crime. There is not a measure of well-being prior to any such victimisation.

The self-completion section of the CSEW is restricted to 16 to 59 year olds. The personal well-being questions are only asked of a random sub-sample of 1 in 4 respondents. This means the overall number of respondents asked these questions is approximately 5,000 individuals per year.

To enable more detailed analysis, the latest two years data (2012/13 and 2013/14) have been combined to produce a final sample size of just over 10,000.

Due to the relatively low number of respondents who report being a victim of any particular crime in the last 12 months, this analysis has been restricted to the highest volume crime types. Of the 4 overall crime types from the CSEW (violence, robbery, theft, and criminal damage), robbery is not included due to the very small number of victims. As a result, the offences available in this report and in the tables are:

Violence

• Violence with injury

• Violence without injury

• Domestic violence (separate breakdown)¹

Theft offences

• Domestic burglary

• Theft involving vehicles

• Theft from the person

• Other personal theft

• Other household theft

Bicycle theft

• Criminal damage

Looking at the wide range of offences provides an insight into how specific crimes affect people more or less than other types. More information on the crime types can be found in Chapter 5 in the User Guide to Crime Statistics for England and Wales (1.61 Mb Pdf).

Notes for crime types used in analysis

1. Domestic violence offences are included within the offences of either violence with or without injury as appropriate. However violence can also be broken down according to the victim-offender relationship. This measure of domestic violence includes violence committed by partners, ex-partners, other relatives or household members (there is more information in Section 5.1 of the ( User Guide to Crime Statistics for England and Wales (1.59 Mb Pdf) ).

Crime does not affect the population evenly. The CSEW shows certain types of people are more likely to be victims of crime than others. For example, data from 2013/14 showed those aged 16 to 24 and 25 to 34 were more likely to be victims than those aged 55 to 64, 65 to 74, or 75 and over, while those unemployed (as opposed to employed), living in urban areas (as opposed to living in a rural area), or living in the 20% most deprived areas were also more likely to be victims of crime. The Annual Trend and Demographic tables have more information.

Like crime victimisation, those who report having a high sense of personal well-being vary across the population. Data from the 2013/14 Personal Well-being in the UK publication found those aged 40-54 were less likely to have high life satisfaction compared with those in both younger (aged 16-29) and older (aged 65-79) age-groups. There were also differences in other demographic characteristics, such as self-reported health, employment status, and marital status¹. Additionally, the differences vary across the four personal well-being measures. The Personal Well-being Estimates Personal Characteristics tables have more information.

These variations mean that there is likely to be a complex relationship between being a victim of crime and personal well-being. Other factors, such as income, employment or marital status, and health will be inter-related with each other and with both levels of victimisation and personal well-being. For example, the personal well-being of a victim of crime living in a deprived area could be influenced by both their living situation and their victimisation. It should be noted that association does not mean causation, and this initial analysis does not attempt to fully explain this complex pattern. Further multivariate analysis would be required to do this.

Notes for section 1 - Who are victims of crime?

1. The Personal Well-being in the UK publication covers the entire UK; while the CSEW data are only relevant for England and Wales. The Geographic Breakdown tables has country-specific well-being data.

This section looks at whether being a victim of crime is associated with someone’s personal well-being¹. The 4 personal well-being measures are all considered, broken down by those who were victimised and those who were not. It then goes on to look at demographic breakdowns, such as age, marital status, and household income.

Notes for section 2 - Victims of crime and personal well-being

1. Though, as has already been touched upon, association does not mean causation.

This comparison is of an overall victim/non-victim breakdown¹. While the severity and cost of a crime can vary markedly on the crime type suffered, previous international studies have found correlations between any experiences of being a victim of crime and both lower happiness and general life satisfaction (Denkers and Winkel, 1998 and Michalos and Zumbo, 2000). Additionally, Staubli et, al, (2013) found evidence of time dependence; for property crimes (for example burglary, car theft, or criminal damage offences) the impact on life satisfaction was only short-term, whereas for personal crimes (such as violence or theft from the person) a negative influence was sustained if the incident took place at any point in the previous 2 years.

Figure 3.1 shows the percentage of those giving positive ratings across the 4 personal well-being measures, broken down by whether or not the respondent was a victim of any crime in the previous 12 months. It shows that for 3 of the 4 measures, non-victims were more likely to give a higher rating than victims (the 2 percentage point difference between victims and non-victims in relation to “low anxiety yesterday” was not statistically significant).

The gap was most noticeable for the “life satisfaction” and “happy yesterday” measures. 75% of non-victims for both measures gave positive ratings, while for victims it was 67% for “life satisfaction” and 68% for “happy yesterday”. The difference for “life worthwhile” was 5 percentage points (79% for non-victims and 74% for victims).

Notes for overall victims of crime

1. The victim measure is based on victimisation of any of the main CSEW crime types (including both personal and household). There is more information in the User Guide to Crime Statistics for England and Wales (1.59 Mb Pdf).

Table 3.1 shows that generally across all age groups, the personal well-being of victims was lower than those of non-victims. However, this difference was less evident among the older age groups. For “life satisfaction” for example, in the 16 to 24 age group the difference in those giving a positive personal well-being score was 13 percentage points higher for non-victims than victims (78% and 66% respectively). In contrast, the difference for those aged 55 to 59 was just 4 percentage points (71% for victims and 75% for non-victims). This pattern of a lower association between well-being and experience of crime in the older age groups was apparent in the other personal well-being measures, although less marked.

The stronger association between personal well-being and experience of crime for young people may reflect differences in the nature of such victimisation. The recently released Focus On: Violence and Sexual Offences 2013/14 showed that, in the year to March 2014, adults aged 16 to 24 were more likely to be a victim of violent crime compared with any other age groups (5.0% of those aged 16 to 24 were victims, compared with 2.4% for 25 to 34 year olds, the second highest age group and 0.9% of those aged 55 to 64). The ‘Victims of violent crime’ section highlights the strong association between violent crime and low personal well-being.

Table 3.2 shows that the difference in personal well-being scores between victims and non-victims were more pronounced for single and separated people than for those who were married/civil partnered or cohabiting. For example, for “life satisfaction” the difference between victim and non-victim for those married/civil partnered (5 percentage points) and cohabiting (3 percentage points) was substantially smaller than those who were single (12 percentage points) and separated (17 percentage points).

A similar relationship to that which exists between marital status and experience of crime is also evident in that of age and victimisation. Younger people (who are most likely to be a victim of a violent crime) are also more likely to be single, while older age groups are more likely to be in long-term relationships. This helps explain the variation in personal well-being scores across the different marital status categories.

Appendix tables 3.05 and 3.06 (733.5 Kb Excel sheet) provides a fuller set of breakdowns for all 4 personal well-being measures (for both personal and household characteristics), with additional groupings for victims of any crime, as well as non-victims.

As well as overall crime, it is possible to look at different crime types and the relationship they have with personal well-being scores. In the analyses that follow, we define “victims” as being victims of the specific crime category presented and “non victims” as not having experienced that particular crime. This means that the “non victim” group will include some people who were victims of another crime type. So, for example a respondent who has experienced burglary but not violent crime will be defined as a victim in the burglary tables and a no victim in the violence tables.

Violent crimes cover offence types from minor assaults, such as pushing and shoving that result in no physical harm, to much more serious violence where significant injuries are suffered. Violent crime can be broken down by whether or not the victim was injured, or by the relationship of the victim to the perpetrator. A common subset of violent crime is domestic violence, where the victim-offender relationship involves partners, ex-partners, other relatives or household members¹.

Table 3.3 shows the personal well-being measures broken down by overall violence, while Table 3.4 has breakdowns for violence with injury, violence without injury, and domestic violence. There were statistically significant differences for 3 of the personal well-being measures across all offence groupings (“life satisfaction”, “life worthwhile” and “happy yesterday”), while “low anxiety yesterday” was also significant for the domestic violence breakdown.

Of all offence types, victims of domestic violence gave the lowest positive ratings for all 4 personal well-being measures. Only 40% of those who were victims of domestic violence gave a positive rating for “low anxiety yesterday” compared with 53% for victims of the total violence category. For “life satisfaction” 41% of victims of domestic violence gave a positive rating (compared with 54% for victims of all violence), for “happy yesterday” it was 46% (compared with 57%), and for “life worthwhile” it was 49% (compared with 58%).

As might be expected, victims of violence with injury gave lower personal well-being ratings than victims of violence without injury. For both measures there were statistically significant differences between victims and non-victims for all personal well-being measures, with the exception of “low anxiety yesterday”.

International studies have seen similar results – both Powdthavee (2005) and Davies and Hinks (2010) found lower well-being in cases where a household had suffered a violent crime². Being a victim of violent crime has been found to have a range of impacts on someone, from emotional reactions such as emptiness or grief, fear or anxiety, and feelings of helplessness or panic, to physical reactions such as aches and pains, nightmares, and changes in appetites³. More specifically, domestic violence has been widely acknowledged to be associated with an increased risk of negative outcomes, such as poor health, mental illness, increased anxiety and social dysfunction (Coker et al. 2002 and Ratner, 1993).

The Nature of Crime tables, which are published alongside the other Focus On publications⁴, can provide further insight on the impact of a certain crime on the victim. These tables provide further information on the circumstances of crimes measured by the CSEW, such as location, offender, injury sustained, and emotional impact. This report will touch briefly on the statistics looking at the emotional impact of the crime, as there may be clear links between this and the victim’s personal well-being.

The emotional impacts of crimes are analysed in Appendix table 3.07 (733.5 Kb Excel sheet). As might be expected, the percentage of victims who were emotionally affected by a violent crime was high compared to the rest of the offence types. The percentage of victims who stated they were emotionally affected by the crime was highest for domestic violence and robbery (93% and 91%). The serious impact of violent crime is even more obvious when looking at those who were affected ‘very much’ by the crime; the top 5 categories overall were (in order of most serious) domestic violence (43%)⁵, robbery (27%), violence with injury (24%), burglary (the one theft offence – 21%), and violence without injury (20%). This evidence falls in line with the personal well-being figures in that of all CSEW crime types, violence has the biggest impact on victims.

Notes for victims of violent offences

1. These domestic violence statistics are collected in face-to-face CSEW interviews. As a result, they should be treated with caution. Prevalence rates for domestic violence derived from the self-completion module are around 5 times higher for adults than those obtained from face-to-face interviews (Chapter 7 of Walby and Allen, 2004).

2. Powdthavee’s study also included the impact of burglary on well-being.

3. For example, ‘The Impact of Violent Crime on You and Your Family’, from Victim Assist Queensland, Australia.

4. The 2 most recent publications are Focus On: Property Crime 2013/14, published 27 November 2014, and Focus On: Violent Crime and Sexual Offences 2013/14, published 12 February 2015.

5. Domestic violence is a separate breakdown and is a subset of crimes from both violence with injury and violence without injury – they are not mutually exclusive.

The overall theft offence category covers any personal or household crime where an item has been stolen – domestic burglary, vehicle-related theft, theft from the person, other theft of personal property, other household theft, and bicycle theft.

The breadth of theft offences mean the potential impact on a victim may vary greatly. Table 3.5 shows the ratings for the 4 personal well-being measures across all theft offences. As well as the overall theft offences category, both domestic burglary and theft from the person had statistically significant differences between victims and non-victims for 2 of the personal well-being measures. For the overall category, the differences were for “life satisfaction” and “happy yesterday” (both had 69% for victims compared with 74% for non-victims).

Table 3.6 provides the personal well-being for each of the different theft sub-categories. The table shows that there is a similar pattern across most of the theft offences, in that victims have lower personal well-being when compared with non-victims. However, due to the small number of victims for some of the theft offences (for example theft from the person and other household theft), only a few of them are statistically significant. Only vehicle-related theft, and a couple of personal well-being measures for other theft of personal property and bicycle theft, goes against this general trend.

Victims of domestic burglary were less likely to have positive ratings for “happy yesterday” (63% for victims compared to 74% for non-victims) and “low anxiety yesterday” (51% compared to 59%). These results fall in line with other research, which has found that burglary victims often experience emotions similar to those of victims of violent crime, including suffering mental health issues, with people worrying about protecting their homes and families in the future (there is research by Safe Essentials and Victim Support). Additionally, both Cohen (2008) and Kuroki (2012) found significant negative correlations between past burglary and happiness.

The differences in the personal well-being of victims and non-victims of theft from the person were largest for “life satisfaction” (58% for victims compared with 73% for non-victims), and “low anxiety yesterday” (44% for victims compared with 59% for non-victims). These are thefts where something has been taken directly from a victim, but without the threat or use of physical force. Examples include snatch theft, stealth theft (including pick-pocketing), and attempted thefts. Items stolen are often of high personal value to the victim (mobile phones or wallets and cash)¹, the loss of which can lead to the victim reporting a lower personal well-being score. In addition to these differences, the “life satisfaction” of victims of bicycle theft was statistically significantly less than non-victims (63% compared to 73%).

As mentioned in the ‘Victims of violent offences’ section, Appendix Table 3.07 (733.5 Kb Excel sheet) looks at the emotional impact a crime had on a victim, as first reported in the Nature of Crime tables. Given the wide range of theft offences, it is not surprising that there is a range in those reporting they were emotionally affected. The range extends from theft from the person (85%) to other household theft (73%). With regards to those that were very much affected by the crime, burglary was the highest ranking theft offence, with 21% reporting they felt this way.

Notes for victims of theft offences

1. Table 7.3 of the 2013/14 Nature of Crime tables (published 27 November 2014), personal and other theft.

In the CSEW, criminal damage is defined as the intentional and malicious damage to the home, other property or vehicles.

Figure 3.2 shows that for all measures other than “low anxiety yesterday”, there were differences between victims and non-victims across the personal well-being measures. The difference was largest for “life satisfaction”, with a difference of over 7 percentage points (66% compared with 74%), followed by “life worthwhile” and “happy yesterday”, with a difference of 7 percentage points (72% compared to 79% for life worthwhile, and 67% compared to 74% for happy yesterday).

Previous analyses have identified links between “signs” of crime (like vandalism and litter) and increased anxiety about crime and the fear of such victimisation (Kershaw et al, 2000). This ties in with a wider theory, originally coined by Wilson and Kelling (1982), termed the ‘broken windows’ theory, where smaller crime types like broken windows, graffiti, or large amounts of littering in turn lead to more serious crimes occurring¹.

Notes for victims of criminal damage offences

1. It should be noted that this theory has been debated heavily in the years since it was published; for example Thacher (2004) outlines studies which found only a modest relationship between disorder and serious crime (and even then, any relationship was likely due to wider social forces).

International studies have found fear or worry about crime happening to someone can have a real impact on personal well-being (for example, in Germany Hanslmaier, 2013 or South Africa, Møller, 2004). Due to the structure of the CSEW, respondents are asked either the personal well-being questions or many of the fear/worry of crime measures covered in Chapter One. This means it is not possible to look at the effect these measures might have on personal well-being. There is more information on the structure of the CSEW in the User Guide to Crime Statistics for England and Wales.

The full CSEW sample is asked if they think they will be a victim of crime in the next year. This means it is possible to look at the personal well-being of people who think it is either likely or unlikely that they will be victimised. Figure 3.3 shows that there were statistically significant differences for all 4 personal well-being measures. The largest differences between those who thought they were very or fairly likely to be a victim compared with those who were very or fairly unlikely was for “life satisfaction” (67% and 76%) and “low anxiety yesterday” (52% and 61%).

The differences here are similar to those found in Figure 3.1, which looked at the personal well-being of victims of any crime compared to non-victims. As with Figure 3.1, there are statistically significant differences for “life satisfaction”, “life worthwhile”, and “happy yesterday”. Unlike victims of crime however, for those who thought it likely they would be a victim of crime in the next 12 months there was a statistically significant difference for “low anxiety yesterday” (the 2 percentage point difference between victim of any crime and non-victim was not statistically significant). See Chapter 2 for more information on the perceived likelihood of becoming a victim of crime.

This section compares 2 different sources – the CSEW personal well-being measures and the Annual Population Survey (APS) measures used in the Measuring National Well-being programme. As the largest survey of households in the UK, the APS includes responses from around 165,000 respondents (across all age bands). Even when constrained to the same parameters as the CSEW (16 to 59 year olds, England and Wales only), the sample size for 2013/14 is almost 90,000.

The demographic breakdowns from all CSEW respondents show similarities to those found in the personal well-being measures derived from the APS (‘Who are victims of crime?’ has more information). For example, both figures show an apparent dip in personal well-being scores for the 45 to 54 age-group, especially for “life satisfaction”. Unemployed and single or divorced people also reported lower personal well-being scores in both series. The demographic breakdowns for the APS can be found as part of the Personal Well-being in the UK, 2013/14 publication.

However, when directly comparing the 2 sources, as shown in Table 3.7, there is a statistically significant difference between all 4 personal well-being measures. For 3 of them, the CSEW total is lower than the official published measures; the exception is the “happy yesterday” measure, where the CSEW total is 3 percentage points higher (74% compared with 72%). For the other measures, the published figure is either 3 (“life worthwhile” and “low anxiety yesterday) or 4 (for “life satisfaction”) percentage points larger than the CSEW figures.

There are several possible reasons why these differences may exist. Research into survey methodology has shown that the context and order in which questions appear in a survey can lead respondents to answer the same questions differently (for example Dillman et. al, 2009). These effects are particularly influential on questions about attitudes (like the well-being questions) than for questions about facts such as age or employment status.

The personal well-being questions in the CSEW are asked in the self-completion section of the survey. This comes after questions on a range of topics including perceptions of crime, experiences of the Criminal Justice System, and, if they were a victim in the previous 12 months, a range of questions on the nature of their victimisation.

The official well-being questions collected in the APS are found early on in the questionnaire, after the basic questions on household and individual demographics. Tinkler and Hicks (2011) (240.8 Kb Pdf) state this is to allow time to build rapport between the interviewer and respondent without allowing later questions to influence response to the subjective well-being questions.

Given the placement of the questions in the respective surveys, it is therefore unsurprising that for 3 of the well-being questions, the totals recorded in the CSEW are lower than the official measures. Schwarz et al. (1987) found that responses to evaluative questions can be determined in part by the respondent’s current mood and by the immediate context. After speaking (often at length) about experiences and thoughts on crime, it is not surprising that CSEW personal well-being measures are, in general, lower than the APS figures.

An additional point is that the Drug Use and Drinking module is carried out via self-completion. Pudney (2010) found that, on average, lower scores to personal well-being questions are received if the interview is carried out via self-completion rather than administered by an interviewer¹. Again, these different collection methods appear to affect how people respond.

Notes for Section 5 - Differences between published personal well-being figures and the CSEW

1. There is more information on how the APS in ‘APS design and its implications for the personal well-being statistics’ as part of the ‘2013/14 Personal Well-being in the UK’ publication.