Queensland Opposition Deputy Leader Cameron Dick has challenged Premier David Crisafulli over his government's performance on crime, while Crisafulli says early figures show his policies are beginning to work. Their competing messages came as polling suggested many residents do not feel safer, despite a series of tough-on-crime reforms.

The dispute turns on two different measures: how people feel about safety in their community and what the government says official crime figures show. Treating those measures as interchangeable would conceal rather than explain the disagreement.

According to a DemosAU/PremierNational poll reported by The Courier-Mail, 39 per cent of respondents said they felt less safe than two years earlier, 18 per cent felt safer and 43 per cent reported no change. These figures describe respondents' perceptions. They are not an official measure of the number of offences committed.

Dick described the government’s midterm performance as falling short of its promises on crime, healthcare, housing and the cost of living. He used the poll results to argue that Queenslanders were not experiencing the improvements the government had committed to deliver.

Crisafulli defended the administration's approach, saying crime rates had fallen by nearly 10 per cent in two years following a decade of major increases. That is the premier’s description of the figures. A full assessment would need to identify the offences counted, the period measured, whether the figures are statewide and the statistical source.

The government has introduced several major changes to youth crime legislation, including expanded categories of serious offences for which young people can be sentenced under adult penalties. Crisafulli says the reforms are necessary to restore community confidence and consequences for offending.

Dick’s criticism should not be portrayed as proof that the laws caused crime to rise. Likewise, Crisafulli’s assertion of an overall decline does not establish that every suburb is safer or that victims have received adequate support. Perception surveys, police-reported incidents, court outcomes and victim services each measure different aspects of public safety.

The Courier-Mail also reported the experience of Brisbane resident Elena Parolin, who described the lasting impact of a home invasion and frustrations over the time taken to assess her victim-assistance claim. Her account illustrates why statistics can feel distant to someone dealing with a traumatic crime. It does not stand as a representative statistical sample of all Queensland residents.

There are questions both politicians must answer. For Dick, which specific policy changes would make communities safer while preserving appropriate legal safeguards? For Crisafulli, how are the government's claimed improvements being distributed geographically, and is assistance reaching victims promptly?

Queensland’s political contest is not confined to the headline number of reported offences. Residents want emergency responses, effective prevention, reliable courts and access to support after a crime. Some may recognise the need for strong laws while questioning the practical services surrounding them.

A fair report distinguishes the poll’s methodology from official recorded crime data and clearly attributes the competing political statements. No finding about the effectiveness of the legislative changes can be drawn from the sentiment poll alone.

The polling result is politically significant because it measures how residents evaluate their personal security rather than asking them to approve an individual law. The published figures were 39 per cent feeling less safe than two years earlier, 18 per cent feeling safer and 43 per cent reporting no change. That is a clear indication of public unease among respondents, but it does not identify which crimes they have experienced, which sources of information shaped their judgment or whether anxiety differs by region or age. Interpreting it as a measured rise of 39 per cent in crime would be entirely wrong.

The two-year comparison also spans more than one administration. Crisafulli's government took office in October 2024, so the question asks respondents to assess a period including the government's tenure and the circumstances immediately before it. A poll can be a fair political test of whether a government has improved confidence, but it cannot establish that a particular legislative change caused people to feel more or less secure. That causal question requires separate analysis of trends and influences.

Crisafulli has cited a reduction in crime rates approaching 10 per cent over two years, and the government says its youth justice changes are part of a sustained response. Dick disputes the adequacy of that performance and has broadened his criticism to government delivery in healthcare, housing and the cost of living. Neither side should have the other’s claims converted into an endorsed statistic without checking the timeframe and categories used.

Official Queensland crime statistics distinguish offences reported to police from offences detected by police, and report different categories separately. Assault, unlawful entry, theft and property damage do not necessarily move in the same direction. Trends can also shift when victims are more willing to report a crime or when police target a specific offence with enforcement activity. A policy claim about crime rates needs to identify whether it refers to an aggregate rate, a count or an individual crime category.

The Queensland Government Statistician’s Office publishes the state's annual recorded-crime report and supporting tables. Those figures can help readers examine longer-term changes, but they may cover a different reporting period from a political claim made in October. A careful comparison should therefore avoid mixing an annual statistical series with a fresh, unspecified government number. The release date and relevant years should be visible beside any chart or interpretation.

Youth justice adds another layer. The LNP government has expanded the range of offences for which young people can face penalties comparable to those imposed on adults. Supporters argue that tougher consequences are necessary for serious repeat offending and to reassure victims. Critics question whether punishment alone prevents future offences or whether prevention, education, mental health support and diversion can reduce reoffending. This is a policy disagreement rather than a simple referendum on whether crime should be addressed.

An arrest or charge is not proof of guilt; an offender’s sentence concerns a case that has proceeded through a legal process. Public debate can lose these distinctions when numbers of incidents, alleged offenders, convictions and sentences are discussed as if they describe the same thing. Crime statistics also need population context. A growing population can generate higher total counts even if rates per person fall, while individual neighbourhoods may experience a worsening trend that does not appear in a statewide average.

The experience of victims deserves attention without turning one account into a measure of the whole state. Reporting featured Brisbane resident Elena Parolin, who described being threatened during a home invasion and feeling lasting anxiety. She also said delays in assistance left her dealing with costs for treatment and security. Those personal impacts are real elements of her account, but they cannot prove the prevalence of similar experiences. They highlight questions about victim support that may remain important whatever happens to the overall offence rate.

That is why the debate should extend beyond penalties. A person who has reported a serious incident may need immediate safety measures, access to medical treatment, psychological care, help replacing locks or other security upgrades, and guidance through compensation procedures. Delays in any of these can leave a victim feeling unsupported even if police ultimately record fewer offences. A government’s public safety record must be assessed across response, prevention and recovery, not only through sentencing legislation.

Crisafulli faces a challenge in explaining how the reported reduction in crime relates to people’s lived experience. The government could publish a plain-language breakdown by offence, region and timeframe, with separate figures for reoffending, victim assistance and response times. Dick faces an equally important test: he can criticise the governing party, but voters are entitled to know which changes Labor would make, what those changes would cost and how their results would be assessed.

There are limits to what any poll can tell readers without the underlying survey methodology. Sample size, question wording, mode of data collection, weighting and margin of error all affect interpretation. The reported percentages are worth analysing, but a rounded headline cannot reveal the distribution of every demographic group. That is an additional reason to avoid declaring that the entire state is moving in one direction based on a single finding.

The sensible question is how public confidence and measured crime trends can both be improved. That requires data on the kinds of harm residents fear, services available after incidents and the efficacy of prevention programs. It also requires patience with evidence: policy interventions may be evaluated over years, while fear can respond immediately to a traumatic local event. Dick and Crisafulli are contesting different interpretations of the current position, and the public deserves to see the numbers behind each.

It will also matter whether local crime trends align with the statewide claims. A state average may mask improvements in some police districts and deterioration in others. Disaggregated official figures would let residents test the politicians’ competing messages against the incidents affecting their area, rather than relying solely on a national-style political slogan.

Dick says the government is failing its own tests. Crisafulli says his administration has begun reversing previous trends and must continue its approach. The unresolved issue is how those competing claims will be measured against transparent data and the experiences of communities most affected by crime.