Between 2009 and 2017, the annual prevalence of assault fell by a third and attempted assault fell by almost a quarter.

Professional fearmongers like Alan Jones may refuse to acknowledge it, but this drop in crime is crystal clear in the data. It is evident in all states and territories, it can be seen across almost two decades of reporting, and it appears in multiple independent datasets, including successive national crime victim surveys conducted by the Australian Bureau of Statistics.

And before you ask: no, people aren't less likely to report crime. The public's willingness to report crime has barely changed; in many cases it has increased.

It's not all good news: reported sexual assaults have increased by 10 per cent since 2017, and internet fraud - perhaps not surprisingly - has doubled since 2007. And there's been a small uptick in reported crime since the book's data were compiled. But the statistics leave us with an obvious question: what caused crime to fall?

This is the question Don Weatherburn and Sara Rahman explore in their new book, The Vanishing Criminal, released this week by Melbourne University Publishing. They pick apart decades of research on the causes of crime - from lead in petrol to a relatively youthful population - to see which of these theories best explains the fall in crime in Australia.

Their findings, based on rigorous econometrics and extensive data analysis, make the way many journalists and politicians talk about crime, and the best ways to reduce it, look a bit silly.

Weatherburn and Rahman start by ruling out the theories that make little sense in the Australian context. There's no credible evidence internationally, for example, that the death penalty or the right to carry guns reduces crime in the first place. But they certainly haven't reduced crime in Australia since we have neither.

Other theories - that crime has fallen because of increased immigration or the strengthening of families, community groups or other social institutions - can be dismissed based on the sheer size of the fall in crime. Changes in immigration have been too small and too gradual to explain the big changes in Australian crime rates, and the fall in crime was too fast and too sudden to be explained by changes in social institutions that take decades to evolve.

American studies have found that the removal of lead from petrol eventually reduced crime by preventing the delinquent behaviour that comes from exposure to lead as a child. But the timing doesn't work in Australia: lead was removed from petrol well before crime began to fall.

The same is true for abortion. American studies famously found that legalising abortion reduced crime around twenty years later (the idea being that unwanted children are more likely to commit crime). But given abortion was legalised in Australia almost thirty years before rates of theft and robbery started to fall - and almost forty years before assaults started to fall - Weatherburn and Rahman question whether it has played much of a role.

Improvements in psychiatric medications, shown in American studies to reduce rates of assault, face the same problem. Not only is there little evidence that these drugs reduce income-generating crime like burglary, but they began being prescribed in Australia well before assault rates started falling in 2008.

So, what has caused the drop in crime?

It turns out that the teenager in your family is about eight times more likely to commit a crime than grandma. Weatherburn and Rahman find that Australia's ageing population explains about 22 per cent of the fall in theft and assault, 21 per cent of the fall in homicide and 18 per cent of the fall in robbery. Fewer young people means less crime.

Improvements in security and technology have helped. The fourfold increase in the number of cars with immobilisers in New South Wales from 2001 to 2017 explains much of why motor vehicle thefts fell 97 per cent (along with new technologies like "microdots"). The effect of improved security and technology on overall crime is less clear, particularly since strengthened security in one area can shift criminal activity towards less secure targets.

READ MORE OPINION:

Alcohol played a big role for assaults. The decline in alcohol consumption among young people - the cohort most prone to assaults - appears to be the primary, if not only, reason for the decline in assault. And while the use of hard drugs like crystal meth has been shown to increase violent behaviour, the increased usage of these drugs has (so far) coincided with a fall in crime rather than an increase in it.

Improvements in the economy have also been shown to reduce crime. In surveying the literature, American economist Steven Levitt reckons that a one percentage point fall in the unemployment rate sees a one percentage point fall in crime: too small an effect to be driving the fall in crime we've seen in Australia, but nevertheless an important part of the story.

What about politicians' favourite: being "tough on crime"? The available evidence suggests that prison has little, if any, specific deterrent effect for potential reoffenders. The relationship between crime and rates of imprisonment is weak. The threat of mandatory minimum prison terms for assault also has no general deterrent effect on the assault numbers. And while an increase in the number of police has been shown to reduce crime, the size of this effect is disputed and depends heavily on the tactics and strategies those police use and the prevalence of police corruption.

What does this mean for policymakers and those in the media?

First, they need to get their facts straight. Overheated political rhetoric and media reporting of crime should have at least some grounding in reality. The fact that the rhetoric and reporting has increased significantly over the very years crime has plummeted suggests that some of our journalists and politicians are, at best, deeply ignorant of what they are talking about or, at worst, deliberately whipping up fear in the community.

Second, when we talk about reducing crime, we should be talking less about jail, police and laws and more about reducing alcohol consumption, reducing drug use, advancing and improving access to safety technology, creating jobs and generating economic growth. The future will hold new challenges when it comes to crime. If we are to solve them, we need to start prioritising reality over rhetoric, and information over ideology. .