The Scottish Government

« Previous | Contents | Next »

Listen

SCOTTISH CRIME SURVEY 2003

CHAPTER ONE: INTRODUCTION

This report presents the main findings from the 2003 Scottish Crime Survey (SCS), a large scale household survey of public experiences and perceptions of crime, involving interviews with a sample of 5,041 adults (aged 16 or over) throughout Scotland. As with all previous sweeps of the SCS, the 2003 SCS relates to crimes that took place during the previous calendar year; that is, between January and December 2002.

The main aims of the SCS are to:

• provide a complementary measure of crime to police recorded crime statistics
• examine trends in the extent and nature of crime over time
• examine the risk of falling victim to crime
• collect information on a number of crime-related issues (e.g. concern about crime; attitudes to the police; drug misuse; domestic violence).

Throughout this paper, the term 'crime' is used to refer to any incident of victimisation recorded by the survey. No technical distinction is made between 'crimes' and 'offences', as in the police recorded crime statistics.

Background

This is the sixth in a series of crime surveys in Scotland, and the last one in its current form. In 1982 and 1988, Scotland participated in sweeps of the British Crime Survey (BCS), co-ordinated by the Home Office. However, data collection was restricted to southern and central Scotland, thereby excluding the experiences of those living in the Highlands and Islands. In 1993, the Scottish Office commissioned the first independent Scottish Crime Survey which extended coverage to the whole of mainland Scotland together with the larger islands. Sweeps of the survey in this format were then repeated in 1996, 2000 and 2003. Interviews with around 5,000 adults have been achieved at each sweep of the survey and were administered by interviewers, face-to-face with respondents, using paper questionnaires.

In February 2003, the Scottish Executive commissioned a fundamental review of the design, content and management of the SCS. The review was prompted by the need to overcome a number of limitations to the current survey; notably, the relatively small sample size which has resulted in an inability to produce reliable trends for some forms of victimisation (margins of error are particularly large around 'lower incidence' crimes such as those involving violence) and the inability of the SCS to report findings at anything below a national level. The intermittent nature of the survey and delays in publishing the results also limited the usefulness of the SCS in policy terms.

Following this review process, a new Scottish Crime and Victimisation Survey (SCVS) was launched in June 2004. This new survey represents a major shift in design, methodology and sample size from previous surveys. The SCVS involves continuous data collection, with a rolling monthly sample of adults across Scotland. Adults aged 16 or over are interviewed over the telephone, rather than face-to-face, and the total annual sample size for the survey is 27,500, which will make possible estimates of victimisation at police force area level and allow more robust estimates of lower incidence crimes. Unfortunately, the new methodology has meant that the self completion modules (for drug use and domestic abuse) have had to be dropped as these are inappropriate questions to be asked in a telephone survey. First results from this new survey reporting on interviews conducted in 2004/05 are due to be reported in autumn 2005.

Purpose of crime surveys

The principal value of crime surveys is that they provide a complementary index of crime to police recorded crime statistics. For a variety of reasons, not all incidents of victimisation are reported to, or recorded by, the police. In addition, police recorded crime statistics are constantly affected by changes in policing practice and police recording practice (for example a 'zero tolerance' campaign will result in more crime being recorded). This will particularly be the case with police recorded crime statistics for 2004/05 because of the introduction of a new Scottish Crime Recording Standard by the police in 2004. The new crime recording standard is expected to increase the number of minor crimes recorded by the police, such as minor crimes of vandalism and minor thefts, and thus affect trends in overall recorded crime. Crime surveys can overcome some of these limitations and provide a more complete picture of victimisation levels by including incidents that are not recorded by the police and by repeating the survey using the same methodology, resulting in trend information that is comparable year on year.

However, there are some limitations to victimisation surveys, and the SCS is no exception. The SCS does not provide a complete picture of crime in Scotland. It is a survey of adults living in private households and does not, therefore, provide information on crimes against people not living in private households (for example those in hospital, prison or homeless); those under the age of 16; and crimes against corporate bodies (for example shoplifting).

Another limitation of the SCS is that it is not a perfectly reliable measure of crime. The survey is only representative of the experience of the people who agree to take part; the fewer people who participate, the less likely it is that the survey reflects the experiences of the people of Scotland as a whole. There might also be errors in the recall of participants. It is also possible that public perceptions of crime and victimisation may change over time, and result in changes in how people report crime from survey to survey.

The SCS results, therefore, like the results from other sample-based surveys, are subject to error. To compensate for this, they key results presented in this report are given with their calculated confidence intervals. These are bands within which we can be fairly confident the 'true' value lies. The figures presented in this survey provide the best available indicator of the levels and trends of victimisation in Scotland.

The survey

The SCS interviewed 5,041 Scottish adults aged 16 or over in their own homes between March and June 2003. Interviews were conducted jointly by the survey companies TNS Social Research and MORI Scotland. This represents a change to the running of the SCS, which was co-ordinated and run by the company MVA for the 1993, 1996 and 2000 surveys. In 2003 the methodology of the earlier sweeps of the survey was replicated as consistently as possible to minimise any 'contractor' effect. However, it is possible that there have been some differences in the training and briefing of interviewers and other survey staff and the execution of the survey which might affect the findings; to what degree this might be the case is unknown.

The response rate for the survey was 68 per cent, slightly lower than the 2000 survey (71%) and significantly lower than the response rate for the 1996 survey (77%), mirroring a general decline in participation in surveys across the UK and other western countries (Martin and Matheson, 1999). The general structure of the interview is detailed in Appendix B. The survey was almost identical to that run in 1993, 1996 and 2000. A summary of the methodology used is also given in Appendix B and definitions of the terms used throughout this report are given in the glossary detailed in Appendix F. Full details of the methodology can be obtained from the Technical Report produced by the survey companies to accompany the dataset, a copy of which can be obtained from the Scottish Executive.

It is important to note that all figures presented in this report are the result of fresh analysis of the data, including the re-analysis of previous sweeps of the survey. This has resulted in some figures presented here being slightly different from previously published findings. Where this has occurred it has been clearly marked in the tables presented in Appendix A. The re-analysis has primarily affected the figures for housebreaking, where the definition has been changed in order for the SCS's housebreaking measure to directly correspond to police recorded domestic housebreaking, including housebreaking to domestic buildings other than dwellings. This has resulted in an increase in rates and estimates for housebreaking and a parallel decrease in the category, 'other household theft'. Full details of the differences between these definitions can be found in Appendix B.

Unlike previous reports on the SCS, this report presents data for the whole of Scotland, and not separately for southern and central Scotland. Previous reports have presented these separately in order to examine trends from the 1982 and 1988 British Crime Surveys which covered central and southern Scotland. Measuring trends over this length of time was not considered necessary in this report, as trend information is available for the ten year period from the first independent SCS in 1993, up to the 2003 SCS. There is therefore no longer any need to differentiate between all Scotland and central and southern Scotland alone.

The report

This report presents the main findings from the 2003 SCS which relate to crime which occurred in the calendar year 2002. More detailed findings on drug use and on domestic abuse from the 2003 SCS will be published in Spring 2005 alongside the findings from the 2004 SCS drugs and domestic violence modules.

The report will first detail estimates of the extent of crime and victimisation in Scotland for 2002 in Chapter 2. It will then provide information on the trends in self-reported victimisation in Chapter 3 by comparing the results from the 2003 SCS to those found in previous year's sweeps. Chapter 4 provides a comparison of the findings with official police recorded crime statistics. It is well understood that individual risks of becoming a victim are not equal. Chapter 5 will explore such unequal risks of victimisation. Chapter 6 examines the public's perception of crime and their worry about crime, and the final chapter will look at the prevalence of drug use in Scotland.

It should be noted that most previous publications on the SCS have conducted comparisons with crime levels in England and Wales through comparing SCS findings with those from the British Crime Survey (BCS). This has not been possible with the 2003 SCS because of changes in methodology in the BCS. The BCS is now a continuous survey and since 2001 all published data has referred to crime which occurred in the year prior to interview with interviews taking place continuously over a financial year. The BCS data are therefore not comparable with the findings from the SCS because of the different reference periods used. However, the new SCVS will be using the same reporting periods, and so findings from 2004/05 onwards will again include comparisons with England and Wales.

« Previous | Contents | Next »