Survey Biography Links News What do we Do? Downloads John Walker Crime Trends Analysis Running Man Logo

Crime Trends


Modeling Global Crime:

Illicit Drugs

Money Laundering

Justice System Resource Allocation:

Forecasting Trends in Crime

The Costs of Crime

Staffing the Police

Indigenous Over-representation in the Criminal Justice System




Interpreting Crime Trends

This page is intended to show how crime trends occur, and how easily they are misinterpreted, owing to the interplay of changes in demographics, social and economic trends, and technology, along with the power play between interest groups such as the police and the media.

These misinterpretations (almost inevitably exaggerations) of the "real" trends in crime help to explain the "ratchet effects" seen in recent years in numerous western countries, where politicians of both left and right have been obliged to outbid each other in an effort to convince voters that they have the toughest policies on crime. It is ironic that, if they had made a real effort to understand what was really going on, taxpayers might have had a genuine reason to vote for them, as we might have been spared the sorts of hysterical responses to crime that have been so obviously costly and unsuccessful in countries such as the U.S.A.  While crime rates have actually been going down in the USA for the last few years, they have fallen across the country, regardless of whether "broken windows", "tough on crime", "three strikes and you're IN" or other such policies have been implemented.  The answer was almost certainly a combination of economic, demographic and social change - and almost certainly wasn't "tough on crime".

The Key Players




Social, Economic and Technological Change


Interest Groups, including










Taken at face value, crime statistics in Australia during the 1980s and 90s showed increases of around 450% in rates of reported crime, but in more recent years they have also started to fall.  Theories range from the effects of the heroin "drought", the effects of increased police numbers and the effects of demographic change.  

We can show diagrammatically the wide range of things have to be taken into account in forecasting trends in crime and justice. Changes in social structure, economics and technology can all alter the age/sex profiles of offenders by making crime more or less accessible or more or less profitable for each community group.  Age profiles of offenders are remarkably stable over time although they can differ across cultures.  It's all to do with demographic life cycles.

"Traditional" property crime in developed countries is often a product of ill-supervised young people taking advantage of simple opportunities presented to them by our commuter-consumer lifestyles - plenty of stuff to steal and nobody home to defend it!  Most grow out of it, when it becomes clear that nobody likes a thief and it's time to get a "proper" job (see the excellent "Patterns and Precursors of Adolescent Antisocial Behaviour", Australian Institute of Family Studies and Crime Prevention Victoria, 2003).  For some, of course, it eventually becomes their "proper job" - either by choice (the best way for them to earn a high income) - or in desperation when jobs are scarce, and for others it becomes an apparently rational way to support a drug habit.

Violent crime rates are much lower than property crime rates.  They are highest in young adults, when, for example, abuse of alcohol or drugs coincides with independence from parental supervision, frequent social activity and the sometimes ill-judged attempts to form adult relationships.  Most soon learn to "behave like adults", because of the disapproval of their peers, although some continue their violence into adult life in forms that include violence against their partners and children.  It is interesting to speculate whether violence is "hereditary", in the sense that children of violent adults often develop violent traits of their own.  The "nature versus nurture" question is still not totally resolved however.  Is it in the genes or the upbringing?

More sophisticated property crime, such as major frauds, used to be the preserve of people with a position of economic influence - either within a business or in charge of a business, and has therefore been an older person's crime.  Currently, however, the emergence of various new forms of electronic commerce is creating opportunities for young computer-savvy people to commit new types of frauds by taking advantage of flaws in the software that drives e-commerce.


The first thing to look at if crime trends change - particularly if they change rapidly in a small region - is demographic change.

For example, consider the development of a new suburb, which is typical of the growth areas around many of our major cities. They growth through a "life cycle" of their own, beginning with the "nappy valley" phase, when many of their residents are young couples with young families.  A decade later, these suburbs usually have a reputation for juvenile crime - not surprising when the population of young people can easily double within a few years and developing suburbs rarely have the full range of social activities to cater for their social appetites.  It's only a phase, however, because they will grow up and many will move away.  With luck, the suburb will eventually age gracefully.

Undoubtedly, some crimes have increased in recent years, notably offences like household burglary which has been affected by our changing lifestyles (e.g. the "dormitory" suburb and the two-income household, which combine to leave whole streets empty during working hours, to become therefore a burglar's paradise).  Ironically, deregulation of business dealings over the same period effectively also reduced the community's vigilance over offences involving fraud, allowing real, but mostly unreported, increases in these types of crimes.  Enormously profitable but morally questionable business practices, such as those exhibited by the management of Enron in the USA and by Alan Bond and others in Australia diverted vast sums from shareholders and customers into the pockets of the offenders.  Businesses are very poor at reporting criminal victimisations - even quite serious ones, because [a] the time and costs of reporting crime are considerable (just think of the lost sales if the manager has to spend time reporting the crime, identifying offenders etc), and [b] they can simply pass on the costs to their customers, so long as their competitors play by the same economically rational (but not good for society) rules.

Other data, however, including crime victim surveys, show that most offences (and violence in particular) have increased little in these twenty years.  What appears to have happened is that people have been encouraged to report crimes, either by specific campaigns (such as in respect of domestic violence) or because of the increased vigilance of insurance companies, which require victims of crime to report the incident to police as a condition of the policy.

So why did the majority of the public - at least as measured in public opinion polls - think that crime was so out of control that it needed drastically "tough" measures?  In order to understand what drives crime and the public response to crime, we need a "dynamic modelling" approach (including feedback loops reflecting the way the real world works).  This particular illustration was constructed in 1984, and was based on observation of the effects of the strong cyclical demographic trends resulting from the post WW-2 baby boom and immigration to Australia.

Feedback Loops

Feedback loops are everywhere in the criminal justice system, but everyone acts as if they don't exist.  Have a look at the model below - it's quite complicated but you'll see three clear "loops" in the logic that illustrate what happens when important groups in the community - e.g. the media, politicians or the police - misinterpret the trends they see or the statistics they read.

Phase 1 of the model begins as a cohort of "baby-boomers" are just reaching their early teens. Because of high immigration and birth-rates after the second World War, Australia enjoys a roller-coaster demographic cycle, with outbreaks of fecundity around every twenty years. There was a very noticeable peak in the mid-1970s, and then a less noticeable one that started in the late 1990s. The reason why the 1990s wave was less noticeable than the 1970s one is because recent generations of young adults have been delaying having families, and even not having families at all.  Even so, during the late 1990s you may remember that our school-leavers were in the news regularly because of the difficulties they were facing finding university places or employment.  Their high rates of offending were also noted by the media, which goaded the politicians into such absurdities as "three strikes and you're in" mandatory sentencing, which resulted in cases of incarceration for children who had stolen bags of sweets.  This demographic cycle is particularly evident in the more economically accessible suburbs around each of the capital cities, because these places are where young parents can afford to live.

Currently in Australia, rates of arrest/caution for offences like burglary, vandalism and theft increase by around 400-500 per cent between the ages of 10 years and 15 years, and peak around 15. The inevitable increases in these petty crimes and minor burglaries, as the cohort enters its teens, quite understandably raise public perceptions and fears about crime. There are naturally calls for "something to be done"! When the media also highlights (and often sensationalises) the size of problem, politicians and police are obliged to respond with bold promises to eradicate crime.

The shrillest calls are always in favour of tougher police crackdowns or longer sentences, because few people are ever really convinced by the less spectacular options such as community crime prevention, or by the idea of fixing the boredom and lack of supervision which leads young people to commit offences (e.g. by providing affordable social or sporting facilities for young people). The police sometimes have a bit of a vested interest in playing up the size of the problem, because they can then claim they are being thwarted by lack of resources or even that poor pay limits their ability to attract good quality recruits. (This is just the way public affairs work!) The media also have just a little self-interest, since people apparently flock to buy "Shock, Horror" headlines, and economic survival comes before journalistic balance every time.

So, even in those communities which go for the "crime prevention" options, there will be a police "crackdown" and they will specifically target known problem areas and arrest a few extra offenders.  So far, so good. The interplay of community interests is reacting understandably and reasonably responsibly to an increasing crime problem.

But the extra arrests are then interpreted by the media, by some pressure groups, and sometimes by the police as demonstrating that a second escalation of crime has taken place. The demands for tougher sentences are based on the fact that crime is soaring out of control. The fact that most of the offenders are young children is ignored by those calling for longer prison terms. This "feedback loop" is found in "bubbles" 2-6 in the diagram above, and at its most frightening can lead to "moral panics" and witch-hunts, in which the innocent are as much at risk from the criminal justice system as the guilty.

Around about this stage in the cycle, Phase 2 commences. The leading group in the demographic cohort is reaching their late teens (bubble 7). By this stage, their offending patterns include violence, (mostly amongst themselves, often involving alcohol and often in public places), appalling driving habits, and possibly drug abuse. The very public and threatening nature of these crimes belie their comparative rarity, and people begin to feel that the streets are being taken over. It is no longer safe to walk around the streets after dark. Well, to some extent that is quite true, although the purpose of this article is to show that it is only a temporary phenomenon.

The obvious conclusion drawn by the public is that the earlier crackdown didn't work, and that even tougher measures are required. Newspapers are always keen to put "Violence Soars!" headlines on the front page, and one can always find a politician anxious to win votes, so there is once again strong public support for even more police, even greater powers of arrest, and even tougher sentencing, or preferably all of these, and we embark on the second feedback loop - bubbles 8-12.

Meanwhile, prisons are filling up with young offenders, mixing with hardened criminals and learning the finer points of crime. At some point, there will be a crisis; perhaps the prisoners riot over excessive overcrowding, perhaps the civil libertarians raise human rights issues. Either will lead to plans for more prison accommodation, but it takes several years to build prisons. In the meantime, the courts may begin to feel that it is not appropriate to send young people to such overcrowded prisons, and this may slow the growth of the prison population. But prison statistics will reach the public arena, showing the growth in prison populations and the rise in numbers of violent offenders.

Curiously, these figures are rarely interpreted as showing some success in combating crime; more usually, they are used as if they are crime statistics, and confirm the public's worst fears.

Phase 3 is reached when most of the cohort is over 18 years of age (bubble 13), when rates of arrest begin to decline. The lower numbers of young teenagers means that their "trademark" crimes, including burglary, theft and vandalism, begin to reduce. This is the sort of good-news story that makes a brief footnote on page nine of the local daily, however, and never makes the "9 O'Clock News" at all. However, violent crime is now at a peak, so there are many more interesting stories for the media to frighten us with. Delays in court processes ensure that prison populations continue to grow, and violent offenders fill the courts. The greater powers of arrest and the new police targeting tactics ensure a lasting enmity between the police and a large segment of the youth population.  This third feedback loop (bubbles 14-18) is perhaps the most turbulent, and decisions made in this phase about the nature of criminal justice will probably set the tone for the next decade or more.

Finally, the cohort passes into the relative tranquillity of its twenties (bubble 19), where arrest rates for the more "public" sorts of offences are much lower than for teenagers. Offences typical of this age group are those associated with work, with economic activity, and with domesticity. i.e. offences like frauds, stealing from employers, and domestic violence. While these offences may be just as damaging as the offences committed by teenagers (and often very much more so), they are not nearly so public and often treated by indifference. Employers often regard staff thefts as just a business cost; spouses sometimes regard domestic violence as part of adult life or may be so intimidated that they dare not complain. Either attitude leads to very low rates of reporting of such crimes. Round about now, the new prison cells will be completed, and the "human rights" considerations are no longer important. Courts will lose what reluctance they may have had to imprison "borderline" offenders, and rates of imprisonment will increase in spite of declining rates of crime.

Reported Crime rates are now low. We have come full circle. But in doing so we have ratchetted up the arrest powers of the police, increased sentence lengths, prison capacities and police numbers. The cost to the taxpayer has increased also. The actual effect on crime is impossible to determine, and could be positive (if "deterrence" actually works) or negative (if the antipathy created between youth and "authority" is translated into higher rates of offending).

Having achieved all this, we are now ready for a re-run with the next generation of "baby-boomers" about to reach their teens.

This model was constructed in 1984, and was based on observation of the effects of the strong cyclical demographic trends resulting from the post WW-2 baby boom and immigration to Australia.

During the 1970s, we adopted the two-income household, commuting to work from a dormitory suburb, as our predominant way of life. The baby boomers from that decade reached their teens during the 1980s, and no-one who has looked at the age-profiles of arrestees would be surprised to find a rise in petty crime in this period. The moral panics which exaggerated the response to this wave of petty crime were based on the commonly held viewpoint that society was completely falling apart. Yet it was perfectly foreseeable that crime rates would fall again after this wave of young people had passed the peak ages of arrest, and they did so. Property crime rates all around Australia began to fall in 1991, with violent crime rates starting their decline about three years later.

Unfortunately, the ratchet effect has left us with unnecessary increases in public expense due to longer prison sentences, a generation which largely distrusts the Police, and the "real" criminals just keep sending their money to Swiss banks or off-shore tax havens!

If we had access to consistent and comprehensive data on crime and offenders in Australia - particularly the ways that criminal careers develop and how much we can change them by effective sentencing and treatment, we could construct an operational model of this type, and find out what is really happening.  But in too many jurisdictions in Australia -like many overseas - we have been saddled with "heritage" technology that cannot link data about offenders with data about sentences.  Our courts and legal system exist and operate in ignorance of whether the sentences they impose actually do any good to anyone.

The Interplay between Demographic cycles and Interest Groups is overlaid by changes in social, economic and technological "drivers'.

Social, economic and technological changes affect offending in different ways for different demographic groups
Interplay of demography and interest groups

Relationships between Demography and Interest Groups are dynamic and cyclical.

Then have a look at these diagrammatic representations of the sorts of societal, economic and technological factors (below) which overlay this entire complex and dynamic process, ensuring that the lessons we learn from history and from the present never quite prepare us for the future or equip us with the knowledge to forecast it!.

But if we had access to more consistent and comprehensive data on crime in Australia, we could at least construct an operational model of this type, and develop some ability to anticipate trends in crime, and therefore to develop sensible policies to tackle them. It has to be better than knee-jerk reactions and moral panics!

The responsibility for preventing crime doesn't just lie with the police.  People who plan our cities, software developers who design our computers and the ways they communicate with each other, bankers who develop new forms of making transactions, all bear considerable responsibility to ensure that what they do doesn't create opportunities for crime.  Crime happens when there is an opportunity and a willing offender.  The demography tells us that people are potential offenders at all ages and regardless of gender.  Politicians and community leaders also bear considerable responsibility to ensure that the communities they lead do not develop built-in conflict between different community groups.  These conflicts become rationalisations in offenders' minds, so that they become willing to commit offences and uncaring of the consequences to victims.


[Home] [Up]

Send mail to with questions or comments about this web site.
Copyright 2006 John Walker Crime Trends Analysis
Last modified: 09/13/06