Discussion of Findings
Searches
-
Victoria Police’s ability to be transparent and accountable about its engagement in racial profiling is adversely impacted by the large number of missing police records for ‘ethnic appearance’ in its search data. The amount of missing ethnic appearance records (16.9% in 2024), remains poor given that Victoria Police made completing these records mandatory in 2019. Indeed making ethnic appearance a ‘mandatory’ field appears to have made very little difference to its completion. This indicates poor internal enforcement and minimal, if any, external scrutiny of this policy.
In 2024 Use of Force forms are missing ethnicity data in 12.5% of cases. What is the explanation for the slightly better recording of ethnicity in Use of Force forms?
Our data is missing the searches performed by Victoria Police in 2020 and 2021. During these years Victoria spent many months locked-down due to COVID-19 restrictions. We don’t know if data from those years is likely to be consistent or quite different to other years. We do have access to VicPol COVID-19 fine data from 2020.
-
Despite poor record keeping practices, some trends are apparent in the available data. Victoria Police search without warrant data from 2018, 2019, 2022- 2024 consistently reveals that when police perform reasonable grounds searches on people they perceive to be African, Middle Eastern/Mediterranean and Asian they are less likely to find contraband than when they search people they perceive to be Caucasian (see Figure 3). This indicates that they are being influenced by race rather than reasonable grounds in the initiation of many of these searches. This provides a source of evidence that the police are engaged in racial profiling against these communities. Over this period the hit rates of people police perceive to be African compared to people they perceive to be White has been low by international standards. This indicates that combating racial profiling in Australia is deserving of far greater attention that it currently receives. This analysis draws on police perception of ethnicity and is unlikely to capture the full experience of racial profiling experienced by non-visually identifiable communities. For example, many Aboriginal people in Victoria appear White, and yet may be targeted due to being known to police, have families known to police or live or spend time in locations where police engage in saturation policing (such as Smith St Fitzroy).
Over all, hit rates for reasonable grounds searches seem to be subtly improving from a low point in 2022 of 16.4% to a high point in 2024 of 19.4%. This could be result of improved practices, or reporting differences. The improved hit rates seem to also reflect a 33% reduction in the number of reasonable grounds searches from 15K in 2022 to 10K in 2025. What this probably indicates is that Victoria Police are now conducting more searches that don’t require reasonable grounds such in designated areas or a where there is Fire Arms Protection Order.
-
Hit rates vary dramatically depending on the basis on which Victoria Police are choosing to search a person. According to Liberty Victoria’s Unreasonable Grounds Report, Victoria Police searches in designated areas produce a find in just over 1% of searches. Our 2024 data shows that when Victoria Police search where these is a Firearms Protection Order in place, hit rates are 8.4%. Both FPO and Designated Area Searches do not require the police to have reasonable grounds. This data provides a sound basis for the elimination of these types of searches as a waste of time and unnecessary breach of human rights (privacy, freedom movement and security of the person). For the searches that require the police to have reasonable grounds, Firearms searches (s149) have a hit rate of 15.2%, Drugs Poisons and Controlled Substances Act Searches (s82) have a hit rate of 16.6%, Control of Weapons Act searches (s10) have a hit rate of 31% and Graffiti Act searches (s13) have a 54.7% hit rate. Of all searches Graffiti Act searches seem to be carried out most reasonably. In 2024, Drugs searches were the most common type of search carried out by Victoria Police (61%) and yet only 16.6% of these searches are resulting in a find. The data indicates that many firearms and drug searches are being carried out on spurious and unreasonable grounds requiring tighter judicial oversight and internal accountability. Only Graffiti Act searches are meeting a 50% or higher hit rate benchmark.
-
The data we have suggests that the hit rate of searches involving people perceived by police to be Aboriginal has increased over the five years of data (Figure 3) from 10.9% to 23.4%. Furthermore the total recorded searches of Aboriginal people (this is people perceived to be Aboriginal added to people who are recorded to be of Indigenous Status) dropped by 13.5% from 852 in 2022 to 737 in 2023 (see Table 3 and 5).
The change in hit rate over this period is remarkable and suggests one of two things. Either police practices are improving over the period towards people police perceive to be Aboriginal OR there has been a change in police recording practices. It is conceivable that the increasing hit rate of searches of people police perceive to be Aboriginal over the four years of data is the result of police responding to increased scrutiny through the Yoorrook Justice Commission and a number of critical inquests (including those into the deaths of Yorta Yorta elder Tanya Day, Gunnai, Gunditjmara and Wiradjuri man Raymond Noel and Gunditjmara, Dja Dja Wurrung, Wiradjuri and Yorta Yorta woman Veronica Nelson), by reducing the unreasonable targeting of people they perceive to be Aboriginal for searches. At the Yoorrook Justice Commission in 2023, Victoria Police’s Chief Commissioner of Police stated that ‘we’ve changed our search policies because we were made aware that Aboriginal youth were searched more frequently.’ (Transcript 8 May 2023, Yoorrook Justice Commission, p. 503.)
However, while the increase in hit rates over time could be the result of improved police practices on the ground, they may alternatively be a reflection of a change in police recording practices. Knowing that they are under scrutiny in their policing of Aboriginal people, police may be reducing their recording of searches of Aboriginal people that don’t result in a find. Alternatively they may be recording Aboriginal people as having an ‘unknown’ ethnic appearance or leaving ethnic appearance blank. This could explain the 13.5% reduction in recorded searches between 2022 and 2024 of Aboriginal people as well as the increasing hit rates.
Changes in recording practices have frequently been identified as the reason behind what may otherwise appear as evidence of improvements in organisational practices. Sociologist Andrew Hopkins writes that once indicators of a particular activity are ‘made to matter, they…are likely to be manipulated’ Hopkins, A. 2016, p.39. In other words, rather than addressing the problem, the organisation ‘manages the measure’, Ibid, 44-46; see also Chan, 1997, p.79; Sabino, 2024, Russo-Lennon, 2024). This does not necessarily mean there is a policy directive to reduce recording, but rather that officers are responding to organisational interest by changing what they record.
In order to determine which of these possible explanations is behind the changing hit rate data, further research is required. This could include an independent qualitative study into the police themselves. As the Stop Data Working Group note in their 2017 Monitoring Racial Profiling report, this could involve conducting interviews and ‘ride alongs’, with the police and the scrutiny of body worn camera footage. It may also be necessary to conduct a survey of the public’s experiences of being stopped by the police. [Here are the results of a such a survey in 2018/2019].
However, while the overall numbers of searches of people perceived to be Aboriginal appears to be dropping, in 2024, the rate of searches of First Nations people was sixteen times greater than the rate for white people (See Figure 7). If the police are indeed reducing their recording of searches, sixteen times may well be an under-estimate. While the sixteen times greater search rate is a preliminary conclusion, it nonetheless reveals a huge difference in the treatment of white people with Aboriginal people. Even with potential issues in the recording of searches, systemic racism against First Nations people is evident in the present data.
-
The reasonable grounds search hit rate for people perceived to be Pacific Islander increased in 2019 (Figure 3). As changes in police data suggests changes in policing practices (either operational or administrative), interpreting what has occurred requires research beyond the data we currently have.
-
In 2022, for the first time, police records separated people perceived to be Middle-Eastern from people perceived to be African. Unfortunately, Victoria Police still combine 'Middle-Eastern' with Mediterranean which may dilute the findings. Despite these caveats, If the data can be relied on, both groups continue to experience lower hit rates than White people - indicative of racial profiling. It is notable that the hit rate for people perceived to be African increases dramatically in 2023 from a base of around 10-12% to about 15.4% and then 16.9 in 2024 (Figure 3). This appears to reflect an overall improvement in the reasonableness of police searches across the board towards all groups (other than people perceived to be Asian). Alternatively, it may reflect a change in police reporting practices. Further research is required. The present data reveals that people police perceive to be African and Middle-Eastern continue to attract lower hit rates than White people and disproportionately more searches than White people. The evidence from multiple sources provides clear evidence of widespread, ongoing racial profiling against these groups.
-
Hit rate data for 2024 show that when police search people they perceive to be Aboriginal, they find an item 23.4% of the time. When they search a person they perceive to be White, they find an item 20.7% of the time. When they search a person they perceive to be African, they find an item 16.9% of the time.
In other words, searches of Aboriginal people appear to be slightly more justified than searches of White people, while searches of African people are slightly less justified than searches of White people. But it is also evident that these rates are not substantively different, making it reasonable to conclude that the hit rate data provide no basis on which to target (over-police) any one of these groups. In all cases, less that 1 in 4 searches result in a find.
Taking the argument a step further, it follows that any substantive disproportionality in the search rates (disproportion based on census data) will be evidence of the unjustified targeting particular racialised groups.
Figure 6 shows the disproportionality search rates of different racialised groups based on census data.
These data show that in 2024, Aboriginal people are 16 times more likely to be searched than white people. African people are 9 times more likely to be searched than white people. These are extreme figures.
Given the previous conclusion, that these is no basis on which to police groups differently, there can be no justification for these extreme differences. These extreme differences are unambiguously due to unjustified targeting or racial profiling.
Finally, the fact that most searches do not result in further police action does not mean they are harmless. Even encounters with the police that do not result in arrest are stigmatising, traumatising and harmful.
-
In Victoria, Criminal Investigation Units, Crime, Divisional Response Units and Uniform branches are making finds in less that 1 in 5 searches. For some units, less than 1 in 7 searches are resulting in a find. These low hit rates indicate that these operations are likely to be performing unlawful searches (i.e. searches without reasonable grounds.) PSOs have the best search hit rates of all operational groups, reaching a high of 42% in 2023. This dropped back to 34.8% in 2024.
-
Police hit rates in Australia are disturbingly poor by international standards: (See comparative hit rate analysis). The low Australian hit rate suggests that police in Australia are not subjected to the same accountability measures (external review, court challenges including exclusion of evidence) as comparable international jurisdictions (UK, Canada, US). Alternatively it could suggest that police in other jurisdictions are better at 'managing’ their data. International studies show that it is quite possible for police to improve their reasonable grounds search rates to achieving a ‘find’ in at least 1 in 2 searches.
-
A preliminary analysis of police search data from 2024 reveals that the police in Victoria disproportionately search Aboriginal people, and people they perceive to be African, Middle Eastern and Pacific Islander. Aboriginal and Torres Strait Islanders are sixteen times more likely to be searched than White people, people perceived to be African people are nine times more likely to searched than White people, people perceived to be Middle-Eastern/Mediterranean are six times more likely to be searched than White people and people perceived to be Pacific Islanders are five times more likely to be searched than White people. This is evidence of systematic over-policing of these communities. Because the overall hit rates are low for all searches (less than 1 in 4), there is a question about their legality. This brings sharply into question Victoria Police’s claims in 2015 to have conducted significant work to ensure police do not racially profile in any form.
-
While, overall, Victoria Police search people they perceive to be Asian at lower rates than White people, when they do search people they perceive to be Asian, they are less likely to find contraband than when they search White people. This means that when police search Asian people they are more likely to do so without reasonable grounds. There is evidence that in 2023, police operations including the Major Drug Squad and Richmond Uniform were targeted at people perceived to be Asian. This effect is not visible in 2024. Nonetheless the 2024 data reveals that police searches of Asian people are the least reasonable than other groups. This means that racial profiling against Asian-Victorians continues to be a concern in searches despite their under-representation in the overall data. We are aware of concerns raised by service providers about other ways in which Asian-Victorians may experience racial profiling such as through under-policing where they are victims of crime or in missing person’s reports and through their mis-identification as perpetrators of family violence. Victoria Police must expand the data it collects and publicly releases to ensure it is possible to monitor racial profiling in all these activities.
-
Like Asian Victorians, the present data reveals that people who the police perceive are from the Indian sub-continent are more likely to be subject to an unreasonable (unlawful) search than White people, but are less likely to be searched overall than white people. In 2009, Indian students raised concerns about the police blaming them for being victims of crime. Further research is required to understand how police are responding currently to their crime reports.
-
Since 2018-2019, Victoria Police’s use of Firearms Protection Order search powers has increased 18-fold. In 2024, Victoria Police conducted 2789 Firearms Protection Order searches. These searches do not require reasonable grounds. Consistent with an absence of threshold for these searches, in 2024, the hit rate for these searches was 8.4%. The absence of a reasonable grounds threshold for these searches leaves them wide open for abuse. This is particularly the case when it is the police themselves who determine who should be on a Firearms Protection Order. Furthermore, police can make these orders if the police don’t like a person's behaviour or the people a person associates with. It appears these threshold-less search powers may be starting to be used in place of searches that require reasonable grounds. In 2024, the hit rate Disproportionality Ratio ('DR') for Firearms Protection Order searches (that is, the hit rate of the racialised group compared to White people) was as follows: people perceived to be African (DR: 0.75, hit rate: 8.7%), Asian (DR: 0.63, hit rate: 7.4%), Middle-Eastern (DR: 0.52, hit rate 6%), Pacific Islander (DR: 0.47, hit rate 5.5%), Aboriginal/TS (DR: 0.66, hit rate: 7.7%) and White (DR: 1, hit rate: 11.6%). While these searches don’t require police to have reasonable grounds to suspect a person has a firearm, the disproportionate find ratios indicate these searches are being used less reasonably on all racialised people when compared with white people.
Use of Force
-
It is no accident that Aboriginal people are 15 times more likely than White people to be searched by Victoria Police and 10 times more likely to have force used against them. These patterns of excessive investigation and brutality in treatment are consistent with the findings of Yoorrook Justice Commission and the Royal Commission into Aboriginal Deaths in Custody and with the Chief Commissioners acknowledgment in 2023 of systemic racism against Aboriginal people. We can now also see that systemic racism in policing extends beyond Aboriginal people. African people are 9 times more likely to be searched than White people, and 7 times more likely to have force used against them, Pacific Islanders are five times more likely to be searched than White people and five times more likely to have force used against them. People who are perceived to be Middle Eastern are five times more likely to be searched and twice as likely to have force used against them. The high rates of search coupled with force levels indicates that it is likely that the increase in stop, question and search of these groups is leading to higher rates of force. Where police have higher rates of contact with particular communities, that contact results in a higher rate of force against those communities. Consequently, in addition to increasing oversight and accountability of the police in their use of search and force, reducing police contact with these communities is vitally important. This means increasing alternatives to police such as Aboriginal Night Patrols and other alternative forms of first response. As the protesters of Abdifatah Ahmed’s death in custody state, the community needs ‘care’ not ‘cops’. The fact that Aboriginal women are 10.6 times more likely than White women to have force used against them is an important data point that is consistent with other data pointing to the higher rates of child removal of Aboriginal children and misidentification of Aboriginal women in family violence. The link between excessive search and excessive use of force, also lends weight to the case for removing unfair police powers such as searches in designated areas and the power to search without reasonable grounds where there is a Firearms Protection Order. Furthermore, search hit rates must be greater than 50% to be seen as acceptable.
-
Uniform Police use the most force of all operational groups across Victoria Police. Uniform police also engage in the most searches where they have among the lowest hit rates. This supports the link between searches and the use of force and clarifies that the use of force is an everyday tactic and not merely a response to protesters or crime gangs as media reporting might suggest. However, PSOs, while well behind Uniform police in the numbers of use of force incidents, are the group using the second highest amount of force, and yet PSOs perform the second least number of searches. This indicates that in addition to searching, officers with high public facing operations are using force in their everyday activities. Hotspots for the use of force includes the CBD, but also in regional areas such as Ballarat, LaTrobe, Shepparton, Bendigo and Mildura. While the use of force is a generalised phenomena across Victoria Police, there is one unit that stands out in the data. This is the Critical Incidents Response Team. This team has been subject to criticism by IBAC, but is this enough? What the present data shows is that the widespread use of force by the Uniform police also requires attention. This data supports calls from across communities, CLCs and academia for a re-commitment by Victoria Police to Project Beacon’s guiding principles on the minimisation of force.
-
Victoria Police are 13 times more likely to use tasers against Aboriginal people, 6 times more likely to use tasers against African people, 6 times more likely to use force against Pacific Islanders and twice as likely against people perceived to be Middle-Eastern than White people. Police are also more likely over the last two years of data to initiate a pursuit against these groups than a white person. The use of techniques such as ‘Grab/hold/push/swarm’, ‘Wrestle/grapple’, handcuffs, takedown techniques and two person 3 point holds were all more frequently used than communication techniques in police/citizen use of force incidents. Tasers were the seventh most frequently used/produced use of force technique.
-
Victoria Police members do not create ethnic appearance records for vehicle only searches. Yet accounts from individuals indicate that these searches are just as likely to be as biased as personal searches and that records ought to be collected. Indeed, the term racial profiling arose out of 'driving while black' cases first reported by The Los Angeles Times in 1986. It is critically important that ethnicity data is collected in all driving cases.
In addition to vehicle only searches, police must be required to collect ethnic appearance data in relation to all stop and question activities, deaths in custody, arrest, bail, caution, missing person notifications, family violence calls, family violence arrests and complaints. The Met Police in London have just agreed to collect data on all traffic stops and subject them to external monitoring.
While from 2022, Victoria Police is separating people they perceive to be African from people they perceive to be Middle Eastern, the ‘Middle Eastern’ category continues to include people perceived to be Mediterranean. (How is ethnicity classified by Victoria Police?) However, data from NSW and from surveys indicates that Middle-Eastern appearing people are more likely to over-policed than people who appear to be Mediterranean. It is critical that Victoria Police separate these groups in its data collection.
-
The Census should record people’s ethnicity in order for systemic racism to be accurately tracked in Australia. While some people will have multiple ethnicities, rather than double counting them, for the purpose of investigating the effect of racial appearance on police behaviour, we need to know the most 'minoritised' of those ethnicities. This will provide clearer answers to the question “which institutions in Australia engage in systemic racism?” The present analysis provides a preliminary indication, consistent with other data sources, that Victorian police engage in systematic racism including racial profiling against a variety of racialised groups.
-
This Project is an experiment. It is an intervention, intending to create transparency in reaction to long standing government failures to hold the police to account and monitor racial profiling. However, the public release of data through this Project may have a number of unintended consequences. Firstly, we speculated that it may result in Victoria Police attempting to deny future data releases, however we can positively report this does not appear to be happening. We also speculated that it may also result in Victoria Police denying the preliminary findings their data currently suggests. Alternatively, it may result in Victoria Police putting more effort into ‘managing the measures’ of racial profiling. Ideally, we hope that the public release of this data will put pressure on Victoria Police to improve the thoroughness and accuracy of its data collection practices, improve the reasonableness of its searches and uses of force and to reduce its targeting of particular communities. However this will require independent scrutiny and external review to ensure that the data reflects what is actually happening, and does not simply represent what Victoria Police want the public to see. Victoria does not currently have independent monitoring and external review of police stop, question, search data or other every day practices. This is a critical issue for the government to address. Further independent research including a second properly funded survey of the public’s experience of policing in Victoria, under the guidance of impacted communities, is critically important in these circumstances.
The data available through this Project covers a far wider ranger of topics that described on this website. For example, in 2017 and 2018 the data reveals that about 83.5% of Victoria Police searches are recorded as being performed under the authority of section 82 Drugs Poisons and Controlled Substances Act 1981. In 2024, 61% of searches are Drugs Act searches. While the percentage of Drugs Act searches are dropping, the data suggests that decriminalisation of drug possession could still dramatically reduce the use of police search powers.
The reduction in the percentage of Drugs Act searches is likely to be because of a massive increase in Victoria Police’s use of Firearms Protection Order Searches. FPO searches have very low hit rates overall. However in 2024 the hit rates of all racialised groups was lower than White people indicating they are being used as tool for racial profiling (the unjustified searching of racialised people).
