Select Committee on Home Affairs Appendices to the Minutes of Evidence


Supplementary memorandum submitted by Barrie Irving, Police Foundation



  The Police Foundation was recently asked by the ACPO to convene a seminar of experts on the "Fear of Crime" concept to:

    —  Examine the state of knowledge in this field

    —  Provide guidance about "what works"

    —  Comment on the mention of "Fear of Crime" in the recent White Paper "Policing a New Century: a blueprint for reform" in the light of the recent statement of Ministerial priorities sent to the ACPO.

  At short notice and mindful of the timetable of the Home Affairs Committee for consideration of the White Paper, the Foundation took soundings from a number of academics and ultimately brought together Geoffrey Pearson (Goldsmiths); Chris Hale (Kent); Martin Innes (on behalf of Nigel Fielding—Surrey) and Tim Hope (Keele). Extensive advice and written material was also made available by Jason Ditton (Sheffield and Glasgow).

  The conclusion of the seminar was that the Police Foundation should, as far as practicable in the time available, synthesise the relevant arguments for the Home Affairs Committee and act as a conduit for appropriate published supporting material.


  The committee will be aware of the way in which "Fear of Crime" features in the "Blueprint for Reform". The summary point which provides the sense of urgency for this evidence is that Police Forces are being asked to reduce "Fear of Crime" across all sectors of society (see attached quotation from Ministerial Priorities communication to ACPO). Indeed, it is in part to enhance their ability to do so that structural working practice, supervisory and sanction reforms are being mooted.

  The logical implication of this new and clearly stated responsibility is that "Fear of Crime", its incidence and prevalence in the population, indeed its epidemiology, can be charted with sufficient rigour to allow police performance in this respect to be assessed along with other more traditional police tasks like reducing and detecting crime.

  It is the considered view of those that the Foundation has consulted that neither the concept of "Fear of Crime" nor the theory behind it, nor its operationalisation within the British Crime Survey and similar surveys, are sufficiently intellectually or practically reputable for Her Majesty's Government to rely on any or all of them in the ways implied by the White Paper and the statement of Ministerial Priorities when taken together.

  The advice given to the ACPO was therefore to argue strongly against any attempt to use British Crime Survey or other similar measures of "Fear of Crime" as a means of evaluating police performance.

  The seminar recognised, however, that there is a political imperative to respond to current levels of public anxiety about crime, and that the police and interested academics have a responsibility not only to expose misguided thinking in this area but also to propose intellectually sound and practicable ways of understanding the causes of that anxiety sufficiently to enable the police and other agencies to control it.


  It takes only a moment of reflection to realise that fear is a complex emotion that is treated in a universally cavalier fashion in common parlance. Fear is a component of excitement in sport and leisure; it serves to alert us to incipient threat (actual or perceived) and is therefore a guardian emotion and both neurophysiologically and experientially it is closely allied to anger and rage.

  While many psychologists have worked with the concept of fear using survey methodology, reputable approaches have usually been cumbersome and could not readily be incorporated into population screening instruments like the British Crime Survey.

  Sociological and criminological investigations of "Fear" are also in danger of straying across disciplinary borders: there are a number of psychiatric conditions that predispose sufferers to express inordinate levels of anxiety in all areas of their lives. It is currently estimated that around 4-6 per cent of the US population suffer from a psychiatric disorder known as "General Anxiety Syndrome."

  The public's experience of the world is heavily affected by exposure to multi-channel media activity that can have, via fiction, fact and "faction" a devastating perceptual impact especially if national media sources set agenda that are then reinforced by local level broadcasting and print.

  These channels of mass communication, largely commercially owned and operated, are driven by audience/readership figures. Stories and presentations that appeal to instinctive emotional drives (sex and self-preservation) tend to win audiences and readers.

  When a member of the public is asked by a pollster about "Fear" and "Fear of Crime", there is a complex mass of experience and processing of that experience brought to play in answering what appears to be a simple question. The questioner, the survey analyst and the user of the results are in danger of remaining blithely unaware of what lies behind the answers.

  To its credit the British Crime Survey has attempted to tie down what can be inferred from the public's responses to questions about "Fear" but as we shall see these efforts have not been successful.

  Jason Ditton and a number of well-known academic colleagues have been reflecting on these issues. The consensus would appear to be that the puzzling characteristics of "Fear of Crime" research results owe more to the poor quality of the concept and its operationalisation in surveys than to any underlying reality.

  These puzzling characteristics include:

    —  Reported prevalence does not reflect differing local circumstances in any logical way

    —  Coherent attempts to change levels of "Fear" have either no effect or a paradoxical effect ("fear" goes up)

    —  "Fear" is negatively related to risk

  The Police Foundation seminar and Dr Ditton agreed that the most parsimonious way of explaining these odd results is to look at what a respondent being surveyed is being asked to do. First, respondents are asked to guess the likelihood that something will happen to them and, having done that, they are asked to imagine the damage they would sustain if the threat became a reality. On the basis of that complex process, respondents then have to infer what effect this hypothetical experience would have on their ambient emotional state.

  Valid and reliable information is unlikely to emerge from this decisional maze. People are poor estimators of probability especially where crime is concerned and imagining damage is a hugely idiosyncratic process. Finally, ambient emotional states (anxious, euphoric, happy, sad etc) are mostly physiologically determined—reactions to individual events feed the overall state but in highly complex ways. Psychiatrists do not know why some people succumb to post traumatic stress disorder while others who suffered the same trauma do not.

  Professor Anthony Giddens has hypothesised that science and modern technology have removed old certainties and produced an anxious generation. This anxiety can be counter-productive but it can also be nothing more than a fashion—post modern angst. Whatever it is, it probably feeds the tendency to say yes to fear of crime researchers.

  Geoffrey Pearson reminded the group of the way in which "respectable fears" have been generated in every generation. But in the last twenty years there appears to have been an increasing tendency for something now called "Fear of Crime" to establish itself as a persistent and ubiquitous component of UK and Western angst.

  A few possible explanations of this phenomenon need to be considered:

    —  Fear of crime is politically popular: it appears to provide governments with a new moral target and a well-established arsenal to attack it.

    —  Fear of crime can be found almost anywhere at any time—it is a research-friendly concept. (A review in 1995 found 200 published articles: Ditton's more recent review quotes 600.)

    —  The quirky nature of the research findings allows an enormous range of popular interpretations.

    —  The findings can easily be used to support police lobbying for additional investment.

  "Fear of Crime" as a distinct field of criminological research can be traced back to Lyndon Johnson's 1967 Crime Surveys. Originally the level of public concern about crime was interpreted as an indicator of the importance politician's should attach to crime rate. High levels of concern were taken to imply the need to reduce crime levels. They were not read as a diagnosis of a public malaise to be treated in its own right.

  Since then there has been an extraordinary and largely unnoticed reinterpretation both here and in the U.S. The public are now treated as if they suffer from a condition that is, in some important respects, independent of crime and detection rates. It is out of this transmogrification that the police `reassurance' strategy and the recent statement of Ministerial Priorities has been born.


  Regardless of local variations and changing rates of crime a decade of "Fear of Crime" research results can be summarised thus:

Table 1

1.The number of respondents who claim that they are a bit worried or very worried about burglary is 60±5 per cent.
2.The number of respondents who claim that they feel a bit or very unsafe when walking alone in their area at night is 35 per cent±6 per cent.
3.The number of respondents who claim that they feel a bit or very unsafe when at home alone at night is 10 per cent±1 per cent.

  Source—Ditton J, 2002

  Table 1 is derived from BCS reports, 1994 and 1996, and Scottish Crime Survey reports for 1984, 1992 and 1996. DITTON, Jason, FARRALL, Stephen, BANNISTER, Jon and GILCHRIST, Elizabeth, "Crime surveys and the measurement problem: Fear of crime". Ch. 8, pp. 142-156 in V. Jupp, P Davies and P. Francis (eds) Doing criminological Research, 2000, Safe, London.

  The impact of investment on reducing fear of crime is paradoxical.

Table 2

1 Low investment
2 Medium investment
3 High investment
4 No investment
Before-After per cent change ActualBurglary
Before-After per cent change Worry aboutBurglary
Before-After per cent change in Burglaryworry by those Aware of anti-Burglary action

  Table 2 is taken from Ekblom, P., Law, H, and Sutton, M (1996) Safe Cities and Domestic burglary, HO Research Study No 164, London, HO RSD; derived from pages xiv, 13 and 70.

  While target hardening investment against burglary works—it tends to be associated with increasing worry about burglary. The only logical result is obtained from those who are aware of the investment and then only if investment is medium or high. To the cynical eye this looks like a reaction to a campaign not the underlying reality.

  What is not in this table, but exemplifies the state of the "Fear of Crime" research industry, is the worry reported by those who live in areas where anti-burglary investment has been high but who are unaware of it—they actually became more worried. If this had been a proper double-blind experiment only those unaware of what was being done to them would feature in the results!

  Ditton has also exposed the danger of assuming that the public know what makes them feel less anxious (safer?). He reports the result of an extensive investigation of street-lighting introduction using a before and after test of estimates of effectiveness and experience of reality.

Table 3

Will (before)
Did (after)
Will/did people's fear of crime decline?
69 per cent
25 per cent
Will/did the amount of crime decline?
59 per cent
17 per cent

  Table 3 is taken from DITTON, Jason, FARRALL, Stephen, BANNISTER, Jon and GILCHRIST, Elizabeth, "Crime surveys and the measurement problem: Fear of crime". Ch. 8, pp. 142-156 in V. Jupp, P Davies and P. Francis (eds) Doing criminological Research, 2000, Safe, London.

  Experts in the field agree that hypothetical questioning is a dangerous activity, especially when questions come armed with official sanction. The British Crime Survey tends to favour using "going out at night or not" as a test situation. Yet methodologists have not really grappled with the problem that most people do not, in fact, go out in the streets at night and therefore cannot easily report on this experience.

  Again the drive to extract data is indicated by the device of asking those who don't go out what they think they would feel like if they did. The way the questions are being asked is tending to drive the results.

  There are, however, sound investigations of what worries people. Top worries by age group have been reliably charted as follows:

Table 4

12-17 years
18-25 years
26-40 years
41-65 years
65+ years
Exams 60 per cent
Money 69 per cent
Money 64 per cent
People 63 per cent
Health 64 per cent
Money 31 per cent
Exams 41 per cent
People 49 per cent
Health 47 per cent
People 37 per cent
People 20 per cent
Job 25 per cent
Job 38 per cent
Money 38 per cent
Ageing 37 per cent
Future 19 per cent
People 22 per cent
Health 28 per cent
Job 17 per cent
Lonely 23 per cent
R/ships 18 per cent
Future 19 per cent
R/ships 13 per cent
World 13 per cent
Money 22 per cent

  Table 4 is taken from Aind de Roiste (1996) "Sources of Worry and Happiness in Ireland", Irish Journal of Psychology, 17 (3), pp. 193-212.

  In a straight race, crime in the UK does not feature and this is also true for middle class Americans. However, it is interesting to note that if researchers throw away their disreputable questioning techniques and ask people to talk about their reactions to crime—the overriding emotion associated with Burglary, Car crime, Assault and Vandalism is anger, with being made "Afraid" coming a poor second.

  The drive to extract "Fear of Crime" value from survey data has also infected some reporting behaviour. In particular, "don't knows" and "no answers" are increasingly excluded from analysis and categories of partial agreement/disagreement are combined in self-serving ways. Examples have been catalogued by Ditton and others but this practice was noted by all participants in the Police Foundation seminar.

  The context of asking about "Fear of Crime" also affects results and in an incisive critique Ditton and colleagues used this phenomenon to check consistency of results, when typical "Fear of Crime" questions were asked of the same people at different times.

Table 5

First Question
Second Question
  Not at all+ not much Quite a bit+ a lot  
Not at all + hardly ever46 1359
Some of the time + all of the time18 2341
  6436 100

  Table 5 is discussed in FARRALL, Stephen and DITTON, Jason (1999) "Improving the measurement of attitudinal response: An example from a crime survey", International Journal of Social Research Methodology, Vol. 2, No. 1, pp. 55-68.

  The researchers then expanded this technique and awarded worry scores to subjects depending on consistency over a number of questions and answers. This produced a more or less normal distribution of worry with very few being worried all the time or never. Using the technique to gauge the proportion of the population who worry consistently about three major crime types, less than 1 per cent fell into the extreme category. Expanding their critique to the issue of "staying in" and "going out", Ditton and colleagues carefully disaggregated the usual BCS questions to discover the real shape of the underlying distribution—while this research still suffers from the methodological problems detailed above, it starts to put the "Fear of Crime" into perspective.

Table 6

per cent
1Feel safe in, stay in, feel safe out
2Feel safe in, go out, feel safe out
3Feel unsafe in, go out, feel safe out
4Feel safe in, stay in, feel unsafe out
5Feel safe in, go out, feel unsafe out
6Feel unsafe in, stay in, feel safe out
7Feel unsafe in, stay in, feel unsafe out
8Feel unsafe in, go out, feel unsafe out

  Table 5 is discussed in FARRALL, Stephen and DITTON, Jason (1999) "Improving the measurement of attitudinal response: An example from a crime survey", International Journal of Social Research Methodology, Vol. 2, No. 1, pp. 55-68.

  The conclusion is that around 8 per cent of the subjects have an 'unsafety' problem—not really the stuff of moral panics.

  Clearly rows 5 and 6 are either responding to special circumstances (domestic violence, neighbour problems) that would be difficult to attack with rational policy. Row 7 and 8 do have a major problem but they fall within the prevalence range for General Anxiety Syndrome and therefore may not be an appropriate case for Criminal Justice System treatment. Membership of these groups is unrelated to age, gender or previous experience—as we would expect if these people are in fact suffering from generalised anxiety syndrome.


  The consensus of the Police Foundation seminar was that the concept of "Fear of Crime" has become by slow steps not only intellectually disreputable but practically absurd. That this is now so is just as much the fault of researchers as politicians, civil servants and police, all of whom have some vested interest in keeping the poor idea alive.

  Up until the publication of the White Paper on Police Reform and the associated statement of Ministerial Priorities, there was no particular harm in the concept and its surrounding academic political and Criminal Justice System flummery. However, it was felt that when an honourable profession is placed in danger of being seriously assessed on its ability to reduce fear of crime across all sectors of society, then it is time to call a halt.

  The seminar recognised that it might be difficult at this late stage and in view of the Police Service's determined pursuit of a "reasonable agenda" based on the "Fear of Crime" concept, to jettison this Ministerial priority entirely.

  It was also thoroughly appreciated that behind "Fear of Crime" lay a real and complex concern to which the Criminal Justice System and the Government ought to respond.

  It was agreed, therefore, that the ACPO and seminar participants through the Police Foundation ought, as a matter of urgency, to prepare an alternative strategy based on firmer philosophical and methodological foundations. The support of the Home Affairs Committee for this endeavour is sought.


  There is one encouraging strand of research in this general field that focuses on using local informants from a variety of backgrounds to identify physical sources of risk and threat in time/space. Geographers have known for some time that establishing settled connection with an area involves learning not only about local amenities (opportunities for benefit) but also about risks and threats that need to be avoided or protected against.

  The seminar agreed that the relatively new partnership arrangements and the availability of local crime management infrastructure presented an unparalleled opportunity to map commonly identified sources of threat and risk at ward level.

  Mapping threat/risk in this way would help to define the relative burden of responsibility falling on various agencies: (Housing, Social Services, Police, National Health and Education). While some of this threat mapping would echo crime/incident mapping—several new dimensions would probably be added.

  The Foundation's own experience as consultants on data sharing in a London Borough suggests that Housing, Education and Social Services might have very considerable roles to play in diagnosing and treating threat hot-spots that now tend to be seen solely in terms of police responsibility.

  Mapping at ward level would nevertheless feed directly into the police reassurance enterprise. However, hard data on threat would do much to specify, prioritise and energise local police efforts.

  The Police Foundation, for its part, will continue to work with the ACPO, the members of the seminar and geographers at University College London to try and develop an appropriate methodology for this enterprise.

Barrie Irving, Police Foundation

February 2002

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2002
Prepared 7 May 2002