Crime stoppers: evaluating Victoria's program

Foreword | Crime Stoppers Victoria (CSV) commenced in 1987 and is "a community-based initiative which encourages members of the public to provide information on unsolved crimes, wanted people and people they know are involved in criminal activity"(www.vic.crimestoppers.com.au). It is one of over 1000 Crime Stoppers programs around the world all of which provide a mechanism for citizens to pass information anonymously to the police by phone. Implicitly, that information would otherwise not be provided to the police.

The CSV website further states that it is "a unique program that is based on a joint effort between the community, police and the me dia". In practice the community provides information about suspicious and illegal behaviour and is represented by a voluntary Board of management for Crime Stoppers Victoria Ltd, an incorporated non-profit "company limited by liability". Victoria Police department staff operates the Crime Stoppers call centre, where they receive and process crime information then refer it to appropriate investigators. And the media publicise details of particular crimes and "wanted" people about both of which information is needed, as well as reporting arrests and providing crime prevention advice.

Toni Makkai
Acting Director

Evaluation of Crime Stoppers Programs

There are only three formally published evaluations of Crime Stoppers programs - from the USA (Rosenbaum et al. 1987, 1989), Canada (Carriere and Ericson 1989) and the UK (Gresham et al. 2001, 2003). Evaluating the success of a Crime Stoppers program is difficult because

"it is impossible to determine whether Crime Stoppers is actually solving crimes that would have otherwise remained a mystery, or if the organization is simply diverting calls from traditional channels of communication that exist between the public and the police" (Carriere and Ericson 1987, p. 81).

When a call to Crime Stoppers does provide information about a matter previously unknown to police, and the subsequent investigation leads to an arrest, there is no doubt that Crime Stoppers has been successful. The UK Crimestoppers Trust on their website (www.crimestoppers-uk.org) is probably referring to these sorts of calls when they note that "independent research has shown that nearly two-thirds of the offenders exposed by Crimestoppers were unknown to the police, or not suspected of the crime in question."

However an evaluation can include a number of measures that reflect the achievements of a Crime Stoppers program. They include:

• The level of community awareness of the program
• Media support for the program
• Participation in the program by the community
• Relative value of calls to the program
• Perception of the program by police
• The economic value of the program

Community Awareness of Crime Stoppers Victoria

A random telephone survey of 1008 Victorians in September 2002 revealed that 94 per cent of respondents said they knew what Crime Stoppers was. This level of recognition was not significantly different across the demographic characteristics of the sample that was equally split between city and country dwellers, males and females, and ranged across all ages. However there was a tendency for younger respondents to be more likely to indicate that they recognized Crime Stoppers.

Recognising the name Crime Stoppers does not necessarily mean that respondents know what Crime Stoppers actually does. Accordingly the 948 respondents who said they knew what Crime Stoppers was, were then asked to describe it in their own words. Not all people find it easy to give a quick and accurate description when suddenly asked to do so by a telephone interviewer, and that proved to be the case here.

Interviewers were asked to note which of the following four main features of Crime Stoppers were spontaneously mentioned by the respondents:

• An avenue for making reports to the police;
• A free telephone service;
• A completely anonymous service; and
• The possibility of a reward.

There were 225 respondents who did not mention, or give any indication of knowing, any one of the above four features. That left only 723 respondents who could safely be said to know Crime Stoppers was, resulting in an "awareness rate" of only 72 per cent, compared with the admitted recognition rate of 94 per cent. This awareness rate is arguably a more accurate indicator of the public's understanding of Crime Stoppers, even though it depended on respondents' ability to put into words their description of Crime Stoppers activities.

Only 19 respondents indicated that they had previously made contact with Crime Stoppers but the nature of those contacts was not pursued. More important was the question "would you call Crime Stoppers if you had information about a crime that you thought would help police locate the offender?"

All 1008 respondents were asked this question. Twelve respondents said they could not answer it, and 93 per cent of the remaining 996 said they would contact Crime Stoppers. By way of comparison, "independent research in Western Australia showed that...86 per cent (of the public there) would anonymously report crime or suspicious activity via the hotline"(Howard 2002, p. 16).

Both these Australian responses are more positive than the results of an on-line survey conducted on the UK Crimestoppers Trust's website. They asked "would you phone Crime Stoppers if you had information?" Seventy-three per cent responded 'yes', six per cent 'no', and the remaining 21 per cent responded 'maybe'. Whether this is a reliable reflection of the average Briton's view is debatable as the respondents were Internet users (from anywhere) who had actively sought out the Crime Stoppers website.

But what of the 73 Victorian respondents in this survey who had said they would not call Crime Stoppers? Forty-nine of them indicated that they would not call because they would call police directly in the first instance. It would distort the public's view of Crime Stoppers to leave these 49 as negative responses because they are actually showing a more direct willingness to pass information onto the police. (Three said they would call police because they knew the 000 number but not the Crime Stoppers number, which one described as "long".)

So it makes more sense to say that 923 of 947 respondents, or 98 per cent of the Victorians surveyed would be ready to use Crime Stoppers to report information about offending. It can therefore be said that Crime Stoppers Victoria has a considerable reservoir of public support.

Media Support for Crime Stoppers Victoria

As one of the three partners in a Crime Stoppers program the media make a vital contribution to its success. They broadcast the hotline number in news reports of criminal incidents, and they highlight incidents for which police are seeking assistance. This is centred around the "target crime" (formerly Crime of the Week) that is televised on Monday nights on the Nine Network. It is then featured on a special page - "Victoria's Most Wanted" - in Victoria's largest circulation daily newspaper the Herald Sun, and aired in radio broadcasts on three metropolitan stations later in the week.

The target crime is usually selected by CSV officers from incidents for which investigators have sought CSVs assistance. A journalist from the Police Media Unit then prepares a press release relating to the incident and that is used as the basis for the preparation of a television re-enactment that is screened as the target crime.

When CSV staff select a target crime, the availability of useful photographs or videotapes as well as their understanding of what makes good television, influences them. They know that the TV network would not be interested in vision if the network did not believe it was "entertaining" (see Carriere and Ericson 1989, p. 51). So in a subtle way the TV network influences which crimes are selected as target crimes and TV re-enactments tend to be more "dramatic" than a lot of crime.

This is confirmed by an analysis of 114 CSV target crimes appearing in the Herald-Sun since the inception of the "most wanted page" (see Table 1). Robberies - good television footage - account for 63 per cent of all target crimes, but only a little more than one per cent of major crime in Victoria. Overall 84 per cent of all CSV target crimes could be classified as violent, not so different from the 81 per cent publicised by Crime Stoppers in Canada in 1989 (Carriere and Ericson 1989, p. 56).

Later Canadian reports provide a different picture. In 1995 only 26 per cent of 332 Canadian Crime Stoppers' "news articles" reflected violence (Thomson 1995, p. 172). More recently, an analysis of 640 Crime of the Week "advertisements" found a rate of only 21 per cent (Lippert 2002, p. 482). This suggests a different approach to Crime Stoppers media coverage in Canada.

Table 1 also shows the average delay between an offence occurring and its being publicised as a target crime. The two publicised homicides were both quite old and their inclusion skews the average delay figure. The first, featured in the very first Crime Stoppers page in February 2000 described the murder of a young woman "whose death has puzzled police for almost 20 years". The second, reported in July 2001, had "police appealing for information regarding the fatal stabbings of two teenagers more than 10 years ago" and was run as part of Missing Persons' Week activities. Removing these two target crimes from the sample leaves an average delay between the offence and its publicity as a target crime of 5.3 months. These delay figures lend some weight to the suggestion that Crime Stoppers publicity is used as a last resort by investigators to solve cases.

In their Canadian research Carriere and Ericson note that "there is no way of knowing how many of the Crime of the Week incidents would have been cleared... if the program did not exist" (Carriere and Ericson 1989, p. 103). They found that publicising the target crimes did not frequently lead to apprehension of the advertised suspects and cited Canadian police coordinators who estimated that only "five to eight per cent of crimes advertised are cleared by the laying of charges" (Carriere and Ericson 1989, p. 83). That is similar to Victoria where six (or 5.3%) of the 114 Victorian target crimes under consideration, have been cleared as a result of information obtained through Crime Stoppers Victoria.

A further measure of the success of publicity in the Herald Sun is provided by a follow-up of those 'wanted persons' featured on the page. In most part those persons are wanted for failing to appear at court in conjunction with alleged criminal activity.

A total of 189 'wanted person' articles had appeared in the Herald Sun pages up to September 2002 and they involved 157 distinct persons. Up until the end of October 2002, 62 of these wanted persons had been arrested, but in only 36 (or 23%) of those cases could the arrest be attributed to the publicity on the Crime Stoppers page. In fact no less than 14 of those 36 arrests occurred within a day of that publicity, and the average time from publication to arrest for those 36 was 13 days. By contrast the remaining 26 arrests, which could not be said to have been a result of the Crime Stoppers publicity, averaged 173 days.

There is therefore no doubt that publicising in the Crime Stoppers page those who breach bail conditions and fail to appear at Court when required is a very useful practice. The widespread media coverage of a bail jumper's photograph seems to be the key ingredient in their surrender.

Community Participation in Crime Stoppers Victoria

The volume of calls that are made to the hotline provides the most obvious measure of participation in CSV. All mainland Australian Crime Stoppers programs use the same hotline number (1 800 333 000) and Telstra connects calls from within a state to that state's Crime Stoppers program. In the 12-month period October 2001 to September 2002 Telstra reported 139 375 calls from around Australia to the hotline. Just over 14 per cent of them (19 880 calls) originated in Victoria and were therefore directed to CSV, and 17 872 of those were successfully connected to a CSV call-taker. Telstra also logged 12 649 calls from mobile telephones which they could not allocate to any state.

Excluding unallocated calls there were 66 Crime Stoppers calls per 10 000 population for mainland Australia over the 12 months. Victoria's rate was only 42 per 10 000 but Victoria has a low crime rate so this might have been expected. However the mainland rated 86 calls per 1000 reported crimes, and Victoria only 68. Both these rates suggest that Victorians may be less likely to make calls to Crime Stoppers than Australians in general.

The Telstra records also show that the average call to CSV lasted 172 seconds, a duration less than the Australian average of 198 seconds. Of course longer does not necessarily mean better. Victorian callers might simply present their information more rapidly or clearly. In general the length of a call depends on the content of the call, the detail that the caller wants to impart, and their opportunity and ability to do so. Obviously the length of any call can be impacted by the behaviour of the call-taker.

The raw number of calls to CSV is only one indicator of its performance. The value of the information provided in those calls is arguably a far better measure and that is discussed next.

The Value of Calls to CSV

Trained staff, as with most Crime Stoppers programs, assess calls to CSV, and if the information from the call is assessed as potentially valuable to investigators, it is forwarded to them. From the above 18 782 successful calls to CSV, 5293 Information Reports (IR) were prepared for and passed to investigators. The remaining 13 489 calls exist only as call-takers' notes and it is not possible to say anything further about them.

The distribution of the 5293 IRs by offence type appears in Table 2 and is compared with the recent British evaluation. Those distributions are quite different. Crime Stoppers calls in Victoria are far more likely to involve drug offences and crimes of violence against the person. The latter appear to be impacted by the inclusion of calls made in response to police requests for information relating to particular homicides that occurred during the 12 month period.

CSV asks investigators to provide feedback and let them know if the Information Reports were:

• "of value and led to an arrest";
• "of value though no arrest was made";
• "retained for intelligence purposes only"; or
• "of no value".

Despite regular reminders, CSV remains unaware of the value of many IRs it sends to investigators. This feedback problem is not unique to Victoria - it also impacted upon the British evaluation. In that study, higher rates of successful calls were found after careful tracking of a sample of calls in three separate Crime Stoppers regions, than were disclosed in the national statistics. One reason put forward for the lack of feedback was that investigators were overstretched and feedback was seen as "just one more, apparently inessential piece of form filling" (Gresham et al. 2001, p. 4).

Despite these feedback difficulties Crime Stoppers programs around the world still publish the number of arrests they have achieved as an indication of their success. Those arrest statistics are generally published cumulatively, that is, when an arrest is notified to them it is added to their current statistics irrespective of how long ago the actual call was made to Crime Stoppers. CSV follows this practice and over 15 years they have achieved a (cumulative) ratio of one person arrested for every 11 Information Reports they have distributed. Over the last five years CSV have distributed 27 845 IRs to investigators and been advised of 2531 arrested persons resulting from CSV information.

Tracking Information Reports

A better way to measure CSVs performance is to track the 5293 disseminated IRs and see what resulted from them up until the end of December 2002. There is a small methodological problem here in that some IRs had only a three-month follow-up period and others a fifteen-month period. However a greater problem is the fact that there had been no responses from investigators for 1660 (or 31%) of those IRs - a stark reminder of the feedback problem. The pity of that is, as the British researchers note, the "lack of feedback does lead to an underestimate of call usefulness". (Gresham et al. 2002, p. 4).

Notwithstanding those difficulties the "attrition" of calls to CSV, starting with the 19 880 calls reported by Telstra appears in Figure 1.

Investigators had provided feedback about 3633 IRs and the majority of them (67.5%) were described as useful. As a percentage of all the 5293 actionable calls, these useful calls comprised 46 per cent, exactly the same rate as reported in the British evaluation when investigating officers who were interviewed stated that 46 per cent of actionable calls "provided some useful information" (Gresham et al. 2001, p. 2)

The British evaluation provides a contemporary benchmark against which the performance of CSV can be compared. In short a comparison shows that CSV:

• successfully answered more calls (94.5% compared with 84.6% in Britain)
• actioned more calls (issuing IRs for 28.2% of calls compared with 12.5% in Britain) and,
• almost doubled the UK's ratio of arrests to calls (1.9% compared with 1.0% in Britain).

These are good results for CSV but the completeness and accuracy of record keeping has a considerable impact on the figures. There are a large number of "disseminated" or "pending" IRs in the Victorian figures - 1660 in all. If some of these pending matters conclude with an arrest, that would raise the success rate when CSV are told of it. The possibility of "unreported" arrests is high if the British situation is true for Victoria. While the overall British rate for arrests, as a percentage of actionable calls was 9.6 per cent, careful tracking of IRs in three regional areas produced rates of 11, 13 and 24 per cent (Gresham et al. 2001, p. 3). As with crime statistics themselves, the accuracy of successful CSV calls is dependent upon diligent recording practices by police.

The majority of the Victorian arrests, 196 (or 68%) related to drug offences. This figure is much higher than that proudly reported in the British research as representing "a high success rate in solving drug-related offences with 41 per cent of all successful resolutions relating to drug offences" (Gresham et al. 2001, p. 2).

Not only did CSV receive more calls about drug-related offences than Britain more of their subsequent arrests were for drug-related offences. Together those facts suggest that it is reasonable to conclude that some part of the reason for Victoria's better success rate of 1.9 per cent of all calls lies in its successful investigation of these drug-related IRs.

Police Perception Of Crime Stoppers Victoria

A random sample of 998 Victoria Police staff were sent a brief questionnaire about CSV in early December 2002. The sample was split evenly between staff dedicated to investigation (working for a Criminal Investigation Unit (CIU) or a specialist investigating squad like Fraud, Homicide, Organised Crime etc.), and uniformed staff based in police stations across Victoria.

A 42 per cent response rate was achieved with clear differences between CIU members (with an average 18 years police service and 10 years investigation experience) and uniform members (26 years and two years respectively). Notwithstanding that, 95 per cent of all respondents said they would recommend their colleagues seek assistance from Crime Stoppers. That indicates a high "approval" of Crime Stoppers activities.

This positive view of Crime Stoppers extends to the analysis of members' interactions with CSV. Just over half the respondents had received information from Crime Stoppers in relation to an incident they had been investigating. Collectively they stated that they had received over 11 000 Incident Reports from Crime Stoppers.

Just under half of all those IRs were retained for policing intelligence purposes suggesting that the information was useful. That is similar to the UK situation where police officers stated that 46 per cent (of actionable calls) provided useful information (Gresham et al. 2001, p. 5).

Overall respondents saw Crime Stoppers as "a useful source of assistance" in over 60 per cent of cases. Yet, only around a quarter of the investigator-respondents had actively sought Crime Stoppers help with an investigation, mostly because they thought it would lead to a breakthrough or because the incident was of major interest to the public.

Police support for CSV can thus be said to be positive overall although it was clear that many respondents were not completely aware of the ways in which CSV could actively assist them.

The Economic Value Of Crime Stoppers Victoria

The annual costs of running CSV are around $1m comprising the costs of Victoria Police Department staff working at CSV (around $690 000) and around $275 000 for CSV staff and operating expenses. This last amount comes from sponsorship of, and donations to, CSV.

The value of the benefits that attach to the work of CSV are less easy to quantify. These comprise the value to the community that flows from CSV activities and the value of services that CSV would otherwise have to buy if it were not a community-based entity.

The first category of benefits includes the value of recovered drugs and property, police savings, and the benefits of criminals desisting or being deterred from offending. None of these can be easily measured and it is necessary to make assumptions to derive realistic estimates of them.

The value of drugs and stolen property recovered as a result of Crime Stoppers activities is globally used to show how well a Crime Stoppers program is performing. CSV recovered drugs valued at $7.6m, and stolen property worth $880 000 in the 12 month period under review. These figures cannot automatically be counted as benefits.

The drugs have no value in that they cannot be sold to realise that amount. Rather, the benefit of recovering the drugs may arise from its impact upon criminal activity that might have occurred in the trading of those drugs. And much of the recovered property would have been returned to those from whom it was stolen saving some of the costs of replacing it.

It is also difficult to estimate the benefits resulting from some offenders desisting from further offending, or being deterred from it, because of CSV activities. The data gathered in the evaluation suggests that all these benefits, along with police and government savings from CSVs activities, is in the vicinity of $2m. (For full details see Challinger 2003).

The second category of benefits comprises the value of the television and print media publicity for CSV that is provided to them at no cost because the media outlets concerned see their generosity as part of their community service activities. These media costs in 2001/2002 amounted to around almost $3m.

Overall then, the cost of running CSV amounts to about $1m and the benefits (as described) amount to almost $5m. That indicates that CSV achieves value for money for the community.

Conclusion

Collectively the various assessments of Crime Stoppers Victoria above show its overall soundness and value.

• CSV achieved a most acceptable level of success as measured by the percentage of calls that led to arrests.
• It was financially cost-effective returning a total benefit almost five times its direct costs.
• It registered substantial support from the public almost all of who are willing to report matters to CSV in the future.
• It received significant support from criminal investigators and operational police with about half of the CSV Information Reports assessed as containing useful information.
• And CSV received effective and attractive media coverage in Victoria leading to arrests for five per cent of its "target crimes" and the location of 23 per cent of the "wanted" persons featured in print media coverage.

Crime Stoppers Victoria's successful operation does not mean that it should now be complacent. It should work towards:

• Increasing the level of feedback from investigators concerning the value of Crime Stoppers information. (The current evaluation was based on only the 69% of Information Reports for which feedback had been received.)
• Continuing its community education programs to increase awareness of the considerable value to the community from using the Crime Stoppers hotline.
• Better informing all police members of the ways in which Crime Stoppers can assist them with investigations.
• Increasing the volume of calls to Crime Stoppers possibly through special phone-in days focused on particular offences of concern to the community.

References

• Carriere K.D. and Ericson R.V. 1989, Crime Stoppers: A Study in the Organization of Community Policing, Centre of Criminology, University of Toronto.
• Challinger D. 2003, Crime Stoppers Victoria: An Evaluation, Australian Institute of Criminology, Canberra.
• Gresham P., Stockdale J., Bartholomew I. and Bullock K. 2001, An Evaluation of the Impact of Crimestoppers, Home Office Research Development and Statistics Directorate (Briefing Note 10/01), London.
• Gresham P., Stockdale J. and Bartholomew I. 2003, Evaluating the Impact of Crimestoppers, Home Office Online Report 22/03, London.
• Lippert R. 2002, "Policing Property and Moral Risk Through Promotions, Anonymization and Rewards: Crime Stoppers Revisited", Social and Legal Studies, 11 pp. 475-502.
• Rosenbaum D.P., Lurigio A.J. and Lavrakas P.J. 1987, Crime Stoppers: A National Evaluation of Program Operations and Effects, National Institute of Justice, Washington.
• Rosenbaum D.P., Lurigio A.J. and Lavrakas P.J. 1989, "Enhancing Citizen Participation and Solving Serious Crime: A National Evaluation of Crime Stoppers Programs", Crime & Delinquency, 35 pp. 401-420.
• Thomson T. 1995, "Crime Stoppers and the Police: The Ideological Construction of Community", in D. Beall (ed), Constructing Danger: The Mis/Representation of Crime in the News, Fernwood Publishing, Halifax NS.

About the authors

Dennis Challinger is a consultant criminologist in private practice.