Analysing the Data
Methodology 1 - Hit Rate Analysis of Victoria Police Warrantless Searches
One way that racial profiling by police can be identified is through examining the reasonableness of police searches. Under legislation — other than in designated places or where there is a Fire Arms Protection order in place (both excluded from this analysis) — police are not authorised to search a person unless they hold a suspicion, based on reasonable grounds, that a person is in possession of a contraband item. When police perform a ‘reasonable grounds’ search we can infer (it is only an inference) that it is more likely the police had reasonable grounds for suspicion if they ‘find’ a contraband item following the search. This inference is not reasonable in individual cases and there are many examples of police finding a contraband item following an unreasonable search. However, over a large number of searches, the find rate provides a widely accepted mechanism to assess the overall reasonableness of police searches.
In this section we examine the reasonableness of police searches in Victoria through analysing their find rate. The analysis is focussed on reasonable grounds searches under the following legislation:
Section 82 Drugs Poisons and Controlled Substances Act 1981 - search without out warrant powers;
Section 10 Control of Weapons Act 1990 - search without warrant powers;
Section 13 Graffiti Prevention Act 2007 - search without warrant powers;
Section 149 Firearms Act 1996 - search without warrant powers;
1.1 What is Victoria Police’s search find rate over time for reasonable grounds searches?
Figure 1 below contains the total number of searches police in Victoria conduct per year (orange column), the hit rate of those searches (i.e. the number of finds) (yellow column) and the percentage of searches that resulted in a find (green line). Please note these are searches made without a search warrant and exclude searches where there is a Firearms Protection Order.
Figure 1 above shows that that the number of searches Victoria Police performs annually increased by about 1, 799 searches between 2018 to 2022. Between 2022 to 2023 it fell by 3,399 searches. The data indicates that the total annual reasonable grounds hit rate has fluctuated from 16.4 to 17.4%. In other words, less than 1 in 5 reasonable grounds searches carried out by Victoria are resulting in a find. In contrast, in 2023, 9.8% of searches where a firearms protection order existed, resulted in a find (see Figure 2).
Click here to read a comparison of Victoria Police’s hit rate with other jurisdictions.
1.2 What is Victoria Police’s hit rate by racial appearance group over time?
To examine evidence for racial profiling using hit rates, we need to compare the find rates when police perform reasonable grounds searches for different racial appearance groups. Victoria Police records the ethnic appearance of the people that they search without warrant. Analysing police search decisions in response to police perception of a person’s racial background is a reasonable method for detecting bias in people whose ethnicity corresponds to appearance. However police perception data will not detect the racialised police-targeting of Aboriginal people recorded as being White. The results for Aboriginal people therefore must be read with caution. For more information about the way Victoria Police categorises racial groups including further limitations in that process click here. Figure 3 below sets out the hit rates (find rates) of Victoria Police searches for groups of people perceived by police to be Aboriginal, African/Middle Eastern, Pacific Islander, Asian and Caucasian over time.
Figure 3 above reveals that when Victoria Police members search people of Middle Eastern/Mediterranean (light blue) African (red) Indian (darkest blue) and Asian (grey) appearance their ‘find rate’ is consistently lower than than when police search people of Caucasian appearance (green). If the data provided by Victoria Police is an accurate reflection of Victoria Police practices, it provides evidence that over these periods the police were disproportionately using search powers without reasonable grounds, against each of these groups. This is evidence of racial profiling. The data indicates that while in 2018, the find rates for searches of people the police perceived to be Aboriginal was very low (10.9%) and indicative of racial profiling, since then, the find rate of police searches against people perceived by police to be Aboriginal has increased to 18.9%. [What is going on?] The data indicates that the raw find rate for Victoria Police searches of people they perceive to be Pacific Islander has been fluctuating over this period.
Figure 3 was created using the data from Table 1 below. Table 2 below is a table of the hit rate disproportionality ratios for different groups compared to people perceived to be Caucasian. It is calculated by dividing the hit rate (%) for each racial appearance group by the hit rate for people perceived to be Caucasian. Where the hit rate disproprotionality ratio is 1, it means that the find rate of this group is equal to the find rate for people perceived to be Caucasian. Where the ratio is above 1, it means that the find rate is greater than for people perceived to be Caucasian. Where the ratio is below 1, it means the find rate is lower than for people perceived to be Caucasian. This ratio is useful for making comparisons.
From Table 1 and 2 you will note that from 2022 onward Victoria Police separated people they perceived to be African from people they perceived to be Middle Eastern/Mediterranean in their data. We can observe that both groups continue to have amongst the lowest hit rates (experience the greatest amount of unjustified searches) of all the groups. In 2022 people perceived to be African had a lower hit rate than people perceived by police to be Middle Eastern. However in 2023, people perceived by police to be Indian had the lowest hit rate, indicating that during 2023, people police perceived to be as Indian (or as having ancestors from the Indian subcontinent) were experiencing the highest rate of unreasonable searches in Victoria.
1.3 What is Victoria Police’s hit rate for different operational groups?
Victoria Police have over 400 different operational groups that conduct searches without warrant across regional and metropolitan Victoria. These groups can be roughly classified into seven different types. The first group is the Uniform police who perform foot and divisional van patrols in local areas. Most local areas also have a Criminal Investigation Unit (‘CIU’). These are colloquially known as ‘detective’ or plain clothed police. There are also ‘DRU’ or Divisional Response Units. These units respond to crime reports across larger areas. Victoria Police also run Highway Patrol police whose dominant task is traffic enforcement, and alcohol and drug testing. ‘Transit’ Police work across Victoria’s public transport network and PSOs (Protective Service Officers) operate with more limited police powers in designated areas. Finally there are individual crime units who perform a highly varied functions from family violence and sexual assault investigation to anti-gang and public order police functions. Figure 4 below describes the average number of total searches each operational type performed across 2018, 2019, 2022 and 2024.
Figure 4 reveals that the Uniform police perform the vast majority of searches without warrant by police in Victoria. Highway Patrol perform the least number which is to be expected given that their dominant focus is on traffic law enforcement. Figure 5, below, sets out the find rates of each of the operational groupings for reasonable grounds searches.
Figure 5 indicates that the Highway Patrol and PSOs have the highest hit rate of the operational groups. Close to 1 in 3 of the searches performed by Highway Patrol results in a find while in 2023, PSOs were almost averaging a find every 2 searches. Highway Patrol and PSOs conduct the lowest number of searches, so that small changes in the reasonableness of searches by these groups will have a noticable effect. The lowest performing groups are the CIU, Crime, DRU and Uniform branches. In 2022 the DRU dropped to a search hit rate of 7.9%.
1.4 In 2023, did any particular police units disproportionately target any racial group?
Here are some examples that stand out in the data (these figures include all recorded searches performed in 2023):
People perceived as Aboriginal were:
16% (4/25) of all searches by Lake Entrance Uniform police.
Lake Entrance Uniform police had an overall hit rate of 4%.
7.9% (5/28) of all searches by Red Cliffs Uniform police.
Red Cliffs Uniform police had an overall hit rate of 3.6%.
33.3% (6/18) of all searches by Echuca Uniform police.
Echuca Uniform police had an overall hit rate of 11.1%.
19.4% (55/283) of all searches by Mildura Uniform police.
Mildura Uniform police had an overall hit rate of 12%.
People perceived as African were:
42.9% (9/21) of all searches by DRU Caroline Springs.
DRU Caroline Springs police had an overall hit rate of 9.5%.
19.5% (8/41) of all searches by NWN Crime Squad.
NWN Crime Squad police had an overall hit rate of 9.8%.
People perceived to be Asian were:
38.9% (7/18) of searches by the Major Drug Squad.
The Major Drug Squad had an overall hit rate of 16.7%.
25.9% (28/108) of searches by the Richmond Uniform Police.
Richmond Uniform police branch had an overall hit rate of 8.3%.
People perceived to be Middle-Eastern/Mediterranean were:
75% (3/4) of searches by the Anti-gang Taskforce
The Anti-gang Taskforce had an overall hit rate of 0%.
41.6% (32/77) of searches by Broadmeadows Uniform
Broadmeadows Uniform police had an overall hit rate of 13%.
50% (6/12) of the searches by CIU Merri-bek.
CIU Merri-bek had an overall hit rate of 16.7%.
People perceived to be Pacific Islander were:
38.9% (7/18) of searches by The Major Drug Squad.
The Major Drug Squad had an overall hit rate of 16.7%.
50% (3/6) of Southern Metro Crime Squad.
The Southern Metro Crime Squad had an overall hit rate of 0%.
3% (4/17) of the Eastern Region Crime Squad.
The Eastern Region Crime Squad had an overall hit rate of 0%.