Reports home
1988-1991 (1-23)
1992 (24-43)
1993-94 (45-66)
1995 (67-84)
1996 (85-106)
1997 (107-127)
1998 (128-143)
1999 (144-163)
2000 (165-177)
2001 (174-188)
2002 (184-195)
2003 (196-215)
2004 (212-231)
2005 (232-256)
2006 (245-261)
2007 (259-278)
2008 (274-
ATSB reports
Non-series reports
Peer review papers
Selected papers

Monash University > MUARC > Reports and brochures >

An Evaluation of the Effectiveness of Overt and Covert Speed Enforcement Achieved Through Mobile Radar Operations

Monash University Accident Research Centre - Report #187 - 2002

Authors: K. Diamantopoulou & M. Cameron

Abstract:

The effect of mobile radar enforcement and supporting publicity on road trauma during July 1995-June 1997 was examined in terms of the number of casualty crashes that occurred on undivided roads in 100 km/h speed zones in rural Victoria. The crash-based analysis compared crash frequencies for the period in which up to 73 mobile radar devices were operational (July 1995-June 1997) with corresponding two-year period before their introduction (July 1992-June 1994) when there was no mobile radar enforcement. The crash effects were examined according to the type of Police vehicle used during each session of mobile radar operation. The type of vehicle used was either a marked patrol car - signifying the overt operation, or an unmarked patrol car signifying the covert operation. Interactions of the enforcement with varying levels of mobile radar publicity awareness were also considered.

The analysis found evidence of casualty crash reductions in rural regions of Victoria, when the speed enforcement operations were either i) covert (unmarked patrol cars) or ii) a mix of overt and covert (marked and unmarked patrol cars). The strongest effects on casualty crashes occurred when a mix of overt and covert mobile radar enforcement was accompanied by high awareness levels of mobile radar publicity during July 1996-June 1997. Under these circumstances, a marginally statistically significant (p=0.0641) net 71.3% reduction was found for crashes occurring on the same day or up to four days after the enforcement was present. This effect was strongest on the day when there was a mix of overt/covert mobile radar enforcement in operation. A net 73.9% casualty crash reduction was found that was statistically significant (p=0.0450).

Crash reductions were also evident when interactions with the levels of mobile radar publicity were not considered. However the effects were smaller and weaker than those that were found when high awareness levels of mobile radar publicity accompanied the overt/covert mix of mobile radar enforcement. There was little evidence of statistically significant crash reductions during times when only marked cars (overt mobile radar enforcement) were in operation.

Executive Summary

During 1999, the Victoria Police suggested that Monash University Accident Research Centre (MUARC) undertake a project to investigate the most appropriate balance of covert and overt traffic enforcement activity. The Police indicated that they were particularly concerned with the efficiency of the two alternative methods of enforcement, the effect of each on unsafe driver behaviour, the relative effect of each on the perceived risk of detection, and the most appropriate balance of these enforcement methods. However, due to changes in enforcement practices, the project's focus has shifted to an examination of the potential benefits of increasing covert speed enforcement activities.

The study was conducted in two parts:

A statistical analysis of mobile radar speed enforcement data in which the type of patrol car (marked or unmarked) used in the Police operations was available.
A telephone survey examining public perceptions of the Police enforcement and perceived risk of detection in general, in relation to speed enforcement, and more specifically in relation to speed camera enforcement.

This report presents the research findings of Part 1 of the study, i.e. a statistical analysis of the mobile radar (also sometimes referred to as moving mode radar, MMR) enforcement data examined by the type of Police vehicle used during each Police session of mobile radar operation. The type of vehicle used was either a marked police car - this signifies the overt operation, or an unmarked police car - this signifies the covert operation.

The findings of Part 2 of the study are given in Senserrick (2000 & 2001).

1.1 Background and Aims

MUARC has previously completed a project that evaluated the effects of mobile radar speed enforcement and supporting publicity on road trauma for the period July 1995-June 1997. A documentation of the mobile radar enforcement and publicity awareness levels achieved, together with the estimated effects of the enforcement on road trauma can be found in Diamantopoulou, Cameron and Shtifelman (1998a & 1998b).

Whilst the above evaluation did not discriminate between the type of patrol car used in the mobile radar operations, the project reported here has considered the enforcement activity levels achieved when marked cars and when unmarked cars were used. In addition, the effects on road trauma of the marked car and unmarked car enforcement were estimated separately. Interactions with the levels of mobile radar publicity awareness were also considered.

1.2 Documentation of the Overt and Covert Mobile Radar Enforcement

The mobile radar activity that occurred in rural(1) Victoria during July 1995-June 1997 was documented and assessed in terms of the following deterrence factors:

The number of mobile radar hours of operation conducted by Police;
The number of speeding offences detected through mobile radar;
The number of speeding offences detected per 100 hours of mobile radar operation.

For each of these deterrence factors, the variation between the overt (marked patrol car) and covert (unmarked patrol car) operations was examined. Overt and covert enforcement was documented separately for the following periods of increased mobile radar enforcement:

Period A (July 1995 to June 1996) - when there were 48 mobile radar devices in operation;
Period B (July 1996 to June 1997) - when there were 73 mobile radar devices in operation.
Periods A & B (July 1995 to June 1997) - when up to 73 mobile radar devices were in operation.

1.2.1 Police Hours of Mobile Radar Operations

In rural Victoria during the combined periods of increased mobile radar enforcement more Police hours of overt mobile radar operations were achieved than covert hours (77% compared with 23%). However during July 1996-June 1997 when there were 73 mobile radar devices in operation, the proportion of the mobile radar hours conducted via unmarked cars increased to 28% compared with 18% during July 1995-June 1996 when there were 48 mobile radar devices in operation. In terms of actual number of hours of enforcement, in rural Victoria there was an increase in the number of hours of covert mobile radar enforcement during the second period of increased mobile radar enforcement, increasing from 7,391 hours in July 1995-June 1996 to 11,587 hours in July 1996-June 1997.

1.2.2 Speeding Offences Detected

For the combined periods of increased mobile radar enforcement (i.e. July 1995-June 1997) when there were up to 73 mobile radar devices in operation in rural Victoria, the majority of speeding offences (79%) were detected via overt mobile radar operations, i.e. 49,120 offences compared with 13,261 offences detected through covert mobile radar operations.

When considering the two periods of increased mobile radar enforcement separately, the proportion of speeding offences detected via covert mobile radar operations (i.e. unmarked police cars) increased considerably to 29% during July 1996-June 1997 when there were 73 devices in operation. The corresponding proportion during July 1995-June 1996 when there were only 48 mobile radar devices in operation was 12%.

1.2.3 Speeding Offences detected per 100 Hours of Mobile Radar Operation

During the first period of mobile radar enforcement, July 1995-June 1996, an average of 73 speeding offences was detected for every 100 hours of overt mobile radar enforcement conducted in rural Victoria. In comparison, for the same period, 45 speeding offences were detected for every 100 hours of covert mobile radar enforcement. However, during July 1996-June 1997 there were increases in both of these detection rates. For every 100 hours of overt mobile radar enforcement, 82 speeding offences were detected. The corresponding rate for covert mobile radar operations was 86 speeding offences detected per 100 hours of enforcement.

Hence when there were 73 mobile radar devices in operation the rate of detection of speeding drivers was more effective/greater via covert (unmarked car) mobile radar operations than overt (marked car) operations compared to the period when there were only 48 mobile radar units in operation in rural Victoria.

Similar trends in the above deterrence factors were found for mobile radar operations in all of Victoria (i.e., including the mobile operations that occurred in the outer Melbourne Police Districts).

1.3 Effects on Road Trauma

The findings in Diamantopoulou, Cameron and Shtifelman (1998a & 1998b) indicated evidence of a four-day residual enforcement effect on casualty crashes for mobile radar operations in rural areas of Victoria, with the effect diminishing on the fifth, sixth or seventh day after the enforcement's presence. Assuming a four-day residual effect, Diamantopoulou, Cameron and Shtifelman (1998a & 1998b) also found that the strongest effects on casualty crashes in rural Victoria, occurred when the mobile radar enforcement was accompanied by high (2) levels of specific mobile radar publicity awareness during November 1996-June 1997. The effects were strongest one to four days after the enforcement was present in each Traffic Operations Group (TOG) region (3) of rural Victoria.

In view of these findings, the examination of crash effects in this study assumed a residual enforcement effect of up to four days prior to a crash.

The crash effects were examined at times when there was:

i. only OVERT mobile radar enforcement occurring in a particular TOG region on a particular day (i.e. only marked car operations);

ii. only COVERT mobile radar enforcement occurring in a particular TOG region on a particular day (i.e. only unmarked car operations);

iii. both OVERT and COVERT mobile radar enforcement occurring in a particular TOG region on a particular day (i.e. both marked and unmarked car operations).

The effects of the mobile radar enforcement (covert, overt or both) on casualty crashes were examined for:

a. The combined period of increased mobile radar enforcement (July 1995-June 1997);

b. The first period of increased mobile radar enforcement, PERIOD A (July 1995-June 1996);

c. The second period of increased mobile radar enforcement, PERIOD B (July 1996-June 1997);

d. Period B considering interactions with mobile radar publicity4.

1.3.1 Crash effects during July 1995-June 1997

During the combined periods of increased mobile radar enforcement (i.e. July 1995-June 1997), reductions in casualty crashes were more evident when the mobile radar enforcement was covert (i.e. unmarked cars) than overt. This was particularly the case 1 to 4 days after the enforcement was present in each TOG region of rural Victoria. Under these circumstances, when up to 73 mobile radar devices were in operation, a net 22% reduction in casualty crashes was found. Reductions were also evident when a mix of both overt and covert mobile radar enforcement was in operation. However, overt (marked cars) mobile radar enforcement on its own was not as effective.

1.3.2 Crash effects during July 1995-June 1996

A similar result was found during the first period of increased mobile radar enforcement, July 1995-June 1996, when there were 48 mobile radar units operating in patrol cars. Casualty crash reductions for this enforcement period were more evident during times and in rural regions where the mobile radar enforcement was covert (i.e. conducted via unmarked Police cars). A net 20.7% reduction for crashes occurring 1-4 days after the presence of the covert enforcement was found.

1.3.3 Crash effects during July 1996-June 1997

During the second period of increased mobile radar enforcement (i.e. July 1996-June 1997) when 73 units were in operation, crash reductions were most evident 1-4 days after the mobile radar enforcement was present in rural Victoria. This was the case when only overt or only covert enforcement was operating (i.e. net 6.6% and 17.5% reductions, respectively, were found).

The largest casualty crash reductions, however, were found during times when both marked and unmarked Police cars were in operation in rural TOG regions, especially on the day of mobile radar enforcement, i.e. a net 40% reduction in casualty crashes was found.

Although none of the above findings were statistically significant, they are indicative of weak casualty crash reductions occurring during times and in rural TOG regions, particularly when there was either:

Covert enforcement (i.e. unmarked mobile radar cars), or
A mix of covert and overt enforcement (i.e. unmarked and marked mobile radar cars).

1.3.4 Interactions with Mobile Radar Publicity

When the enforcement was accompanied by high levels of mobile radar publicity awareness, the strongest (and largest) crash reductions occurred when both marked and unmarked Police cars were conducting mobile radar operations on a particular day and in a particular TOG region of rural Victoria during the second enforcement period, July 1996-June 1997.

Under these circumstances, a marginally statistically significant (p=0.0641) net 71.3% reduction5 was found for crashes occurring on the same day or up to four days after the enforcement was present. This effect was strongest on the day when there was a mix of overt/covert mobile radar enforcement in operation. A net 73.9% casualty crash reduction was found that was statistically significant (p=0.0450). However, as indicated by the 90% confidence limits placed on this estimate, this net % reduction could range from as low as a net 15% reduction to as high as a net 92% reduction.

A smaller net reduction of 50% was found for crashes occurring 1-4 days after the enforcement, however this reduction was not statistically significant.

Non-statistically significant crash reductions were also found during times when:

only overt enforcement, or
only covert enforcement

was accompanied by high mobile radar publicity awareness levels. This was especially evident 1-4 days after the presence of the mobile radar enforcement in rural TOG regions.

1.3.5 Limitations of the crash analysis

It should be noted that the effects on road trauma estimated in this study may be somewhat conservative in nature because of the following decisions that were made to facilitate the evaluation:

All casualty crashes were examined rather than a subset that, presumably, was speed-related (assuming such crashes can be defined).
All undivided, 100 km/h speed limit roads in the regions thought to be covered by mobile radar operations were included rather than those where the enforcement actually took place.
The control group was restricted to crashes that occurred on the same roads and regions as the mobile radar enforcement but during days that were considered to be uninfluenced by the enforcement. Because the operations were so frequent, this limited the number of crashes that could be considered as 'control' crashes.
Although the study compared crashes that occurred during the period of mobile radar enforcement with the period when there was no mobile radar enforcement, other types of speed enforcement (such as speed cameras or hand-held radars) may have been operating on the same roads during the pre-period that may not have been accounted for by changes in the control crashes.
The analysis assumed that crashes were exclusively influenced by the most recent mobile radar enforcement present in a particular region. This most recent enforcement was considered to over-ride any enforcement that occurred one to four days earlier, which also may have influenced the crashes. Given that a four-day residual effect on crashes was found in the analysis, the most recent enforcement may not necessarily have been the most important in influencing crash frequencies.

1.4 Conclusion

The evaluation of mobile radar enforcement and supporting publicity found evidence of casualty crash reductions during July 1995-June 1997 in rural TOG regions of Victoria, particularly when the speed enforcement operations were either:

covert (unmarked patrol cars) or
a mix of overt and covert (marked and unmarked patrol cars).

The strongest effects on casualty crashes occurred when a mix of overt and covert mobile radar enforcement was accompanied by high awareness levels of mobile radar publicity during July 1996-June 1997. Under these circumstances, a marginally statistically significant (p=0.0641) net 71.3% reduction was found for crashes occurring on the same day or up to four days after the enforcement was present. This effect was strongest on the day when there was a mix of overt/covert mobile radar enforcement in operation. A net 73.9% casualty crash reduction was found that was statistically significant (p=0.0450).

Crash reductions were also evident when interactions with the levels of mobile radar publicity were not considered. However the effects were smaller and weaker than those that were found when high awareness levels of mobile radar publicity accompanied the overt/covert mix of mobile radar enforcement.

(1) Mobile radar operations occurred predominantly in rural Victoria Police districts during the evaluation period, July 1995-June 1997. There was also some enforcement in outer Melbourne Police Districts (i.e. Districts E, F, H, J & K). However, because the Melbourne operations constituted only about 10% of the total mobile radar activity, the focus of the documentation and the crash evaluation was primarily for rural Victoria.
(2) Achieving weekly advertising awareness levels of at least 250 Adstock (a function of television Target Audience Rating Points) units.
(3) The locations of the mobile radar operations were limited to the Police Traffic Operations Group (TOG) stations that conducted the operations. Estimates of the likely areas of a Police Division thought to be covered by each TOG station were made in terms of groupings of Local Government Areas, LGAs. It was these estimated groups of LGAs considered to be covered by their respective TOG stations that defined a TOG region for the crash analysis.
(4) Interactions with mobile radar publicity awareness were only considered during Period B (July 1996-June 1997) because a television advertisement specific to the mobile radar technology was first launched by the TAC in November 1996. There was no mobile radar advertising during Period A.
(5) It should be noted that with 90% confidence, the true value of the estimated net 71.3 % reduction lies in the range of a net 91.2% reduction to a net 6.9% reduction, as indicated by the confidence limits placed on this estimate.

Sponsoring Organisation: Baseline Research Program - Department of Justice, Transport Accident Commission, Royal Automobile Club of Victoria (RACV) Ltd, VicRoads.