Fact check: Have robberies, assaults and burglaries increased in Victoria since Labor was elected? – Fact Check
Victorian Opposition Leader Matthew Guy has tweeted that certain categories of crime have increased by more than 30 per cent since Labor was elected. (AAP: Joe Castro)
Crime rates in Victoria continue to be hotly debated in the lead up to November’s state election.
Earlier this year, Immigration Minister Peter Dutton weighed into the state-based issue, claiming Victorians were too “scared to go out to restaurants” because of “African gang violence.”
The comments prompted the hashtag #MelbourneBitesBack to trend on social media as Melburnians mocked Mr Dutton’s claim.
More recently, both Victoria’s Premier and Opposition Leader have quoted seemingly contradictory crime statistics for the state.
Premier Daniel Andrews tweeted on March 14: “Here are the facts: crime is down 10 per cent. The biggest drop ever…”
But on the same day, the Opposition’s Matthew Guy tweeted, “Since the Andrews Govt was elected: aggravated robbery up 30%; robberies up 35%; assault up 37%; aggravated burglary up 48%. Only the Liberal Nationals will make our state safe again.”
Are Mr Guy’s figures correct? Has there been an increase in these types of crimes in Victoria since the Andrews Government came to power in November, 2014? RMIT ABC Fact Check investigates.
Mr Guy’s claim is overstated.
Data compiled by the Crime Statistics Agency (CSA) — a unit within Victoria’s Department of Justice and Regulation that analyses crime statistics independent of Victoria Police — does appear to support Mr Guy’s claim.
However, Mr Guy is referring to crimes by count alone (that is, by simple numbers), which doesn’t account for population growth; in some cases, he also cherry-picks types of crimes.
For example, the CSA classifies aggravated robbery as a sub-set of the broader category of robbery, which falls within the offence division of “crimes against the person”.
Mr Guy is also selective when citing figures for common assault, which is a sub-category of the broader assault division and which has seen a 37 per cent increase in offences since 2014. But CSA data for the broader “assault and related offences” category, shows only a 16 per cent increase over the same period.
Experts told Fact Check that focusing on crime rates per 100,000 of population — rather than by raw numbers alone, as Mr Guy appears to have done —would give a more accurate picture of crime across the state.
They also pointed out that crime numbers fluctuated from year to year, and there were different ways of measuring crime.
Data collated over 10 years, for example, helped provide a clearer understanding of crime than a particular set of offences over a shorter period.
Although the categories of crime cited by Mr Guy did show increases in offence numbers of the magnitude he suggested, experts described the claim as misleading.
About the data
Fact Check contacted Mr Guy’s office seeking the source for his claim, but didn’t receive a response.
In March, Victoria’s Crime Statistics Agency released its 2017 annual report, showing the number of crimes committed in the state had dropped by 9.6 per cent over the previous year.
But there are a number of ways of unpacking these statistics; for example, by the number of offenders or victims of crime, by the number of offences committed or by the number of criminal incidents recorded (a recently introduced measure).
Criminal incident data is designed to provide a measure which takes into account that a criminal event can involve multiple offences, with a number of victims or offenders involved.
This is why the number of offences recorded will necessarily be higher than the number of criminal incidents recorded.
Although the two measures — offences and incidents — both provide accurate measures of crime, the CSA has applied the incidents methodology to make the data more meaningful.
The agency defines “criminal incident” as an event that may include multiple offences, alleged offenders and/or victims recorded on a single date or at a single location.
It defines “offences” as criminal acts reported to or detected by police, where a penalty could be imposed.
Adam Halliwell, the CSA’s Data Capability Senior Engagement Officer, told Fact Check that simply relying on the number of offences was too simplistic when analysing crime data.
“We calculate rates per 100,000 [people] and that’s designed to account for population growth over time,” he said.
While Mr Guy’s figures are correct in terms of the number of offences recorded, Mr Halliwell noted that “you can say something’s gone up 30 per cent in the last four years, but you’re not actually controlling for the population growth over that time.”
He said there was no right or wrong in terms of using offence or incident data; the new measure of recorded incidents was just another way of measuring crime.
“One incident of crime has a number of different things within it; a number of different offences recorded.
“The two measures are both accurate; they just show slightly different counts or measures of crime.”
But Karen Gelb, a consultant criminologist and lecturer at the University of Melbourne, warned against relying on offences by count alone.
“That’s not an accurate representation,” she said. “It doesn’t take into account Victoria’s population [which] has increased during that period. That’s why you have to use a rate and not an offence count.”
She added: “It may well be accurate statistically, but its not honest … If you’re using a straight count rather than a population-based rate, then that’s completely misleading.”
Victoria’s crime trend
Trend data produced by the CSA shows a moderate rise in criminal incidents over the last decade — during both Liberal and Labor terms in office — rising 1.2 per cent between 2008 and 2017.
The Andrews Government was elected in November, 2014. Data for the three years to the end of 2017 shows a 4.4 per cent jump in the rate of criminal incidents.
The rate rose by 2.8 per cent between 2010 and 2014 under the previous Liberal government.
The decade included a sharp jump in criminal incidents in 2016 and a return to the previous pattern in 2017.
The reason for the spike appears to have been a 7.8 per cent rise in property and deception offences, although the raw statistics don’t explain why that occurred.
Dr Gelb, of Melbourne University, says the rise in crime in 2016 could be an outlier given the overall trend, and could have been the result of any number of things, including policy, legislative and practice changes (for example, changes to police reporting practices or a crackdown on specific types of crimes).
“Based on CSA data of the criminal incident rate over 10 years, it’s actually stable,” she noted.
“The key thing about trends is that there can be a short-term increase or decrease that is based not so much on what’s happening in the real world, but our responses to it.
“So, policy changes and practice changes can lead to blips up and down. It’s not necessarily going to have anything to do with the nature of the offending that’s happening; it could have to do with the nature of policy and practice changes.”
She added: “You need to look at trends to allow those sorts of interventions to settle.”
Taking offence data alone, while not adjusting for population growth, gives a different picture and shows an upward trend that appears to have started in 2010, the year the Liberals last came to power.
The offence rate rose from 6923 that year (2010) to 8186 in 2017 — an increase of more than 18 per cent.
What the data shows us
Fact Check has investigated the data for each of the categories that Mr Guy listed in his tweet (robbery, aggravated robbery, common assault, and aggravated burglary).
Robbery and aggravated robbery
In his claim, Mr Guy referred to robbery, which the CSA defines as the unlawful taking of property.
He also mentioned “aggravated robbery” — the taking of property while in possession of a weapon.
The number of offences for robberies rose by 35 per cent over the three years to the end of 2017.
However, the rate of robberies (that is, the number of offences per 100,000 population) rose by 28 per cent over the same period.
This same pattern is evident in Mr Guy’s claim for aggravated robbery, where the number of offences rose by 30 per cent and the rate by only 22.7 per cent.
Mr Guy referred to common assault, which is included within the broader category “assault and related offences”.
The number of offences for the common assault category rose by 37 per cent between 2014 and 2017, which is similar to the incident rate — up 32.5 per cent.
But the offence rate for the same period was 29.5 per cent higher.
This is different for the broader “assault and related offences” category, which is defined by the CSA as a direct and confrontational use of force, injury or violence.
The number of offences shows a 16 per cent increase between 2014 and 2017. But the increase in the rate for the same category is 10 per cent.
This is similar to the incident rate for the same period (up 11.5 per cent).
Mr Guy also referred to “aggravated burglary”, defined by the CSA as the unlawful entry of a property while someone is present, such as the homeowner, or involving the use of a weapon.
The number of offences for aggravated burglary rose by 45 per cent between 2014 and 2017.
However, the offence rate for the same period was up 38 per cent.
What do the experts say?
Experts cautioned that there were myriad factors affecting crime rates.
Adjunct Research Professor Colleen Lewis, of Monash University’s Faculty of Arts, said: “Crime rates can be influenced by a number of factors including increased police numbers, police targeting particular types of crime, the number of people reporting crime and increases in the state’s population.
“If the crime rate trend line is increasing each year, then crime is increasing. However, to understand the nature of the problem you need to examine the trend lines for particular categories of crime as there will be variations; some crimes will have decreased while others will have risen.
“Just looking at trend lines in isolation is a somewhat simplistic approach to a complex issue.”
Dr Sarre, of the University of New South Wales, said the numbers being referred to in Mr Guy’s claim were relatively low.
“It’s difficult to make an assumption; with low numbers, things do tend to jump around. That’s why you have to look at the long-term trend. Otherwise, you can’t make a true judgment.”
Principal researcher: Natasha Grivas
- Crime Statistics Agency, data table for Criminal Incidents, year ending December 2017
- Crime Statistics Agency, Recorded Criminal Incidents, year ending 2017
- Riley Morgan, SBS news, #MelbourneBitesBack: Dutton mocked after African youth crime comments, January 2018
- Crime Statistics Agency, Offence visualisation data table, year ending December 2017
- Crime Statistics Agency, Historical Crime data by recorded offences, year ending December 2014
- Opposition Leader Matthew Guy, Twitter, March 2018
- Victorian Premier Daniel Andrews, Twitter, March 2018
- Paul Karp, The Guardian, Peter Dutton says Victorians scared to go out because of ‘African gang violence’, January 3, 2018
- Greg Barns, Sydney Morning Herald, The Libs’ tough-on-crime policies failed Victoria, December 5, 2014
- John Silvester, The Age, How to break the crime cycle, September 30, 2016
- Peter Martin, The Age, Victoria the safest it’s been for 10 years, says ABS, February 8, 2018
- ABS, Recorded Crime – Offender, 2016-17, February 8, 2018