Abstract

This article briefly sets out some of the significant threats to our freedom and human rights, in the broadest sense, that are facing us now and are likely to increase in the near future.  These threats will be underpinned by the rapidly developing technologies and their potential for abuse by criminals, terrorists, and oppressive states. It then defines crime science, a relatively new approach to the control of crime, which promotes science, crime prevention and security, together with practicality and ethics. It suggests that crime science offers a more rational and appropriate methodology for the future control of crime, including disorder at one end and serious violence and terrorism at the other. The article considers the difficulty of getting the right balance between protecting the public from threats and infringing their human rights and freedoms. It argues there is a need for constant vigilance and for support for those individuals and agencies that can monitor state and other threats to unethical and unwanted intrusions.

Why do we need saving?

This article addresses crime, violence and terrorism, from which citizens expect protection by the state. It is reasonable to expect an increase in these events in the near future, driven by global warming (bringing food and water insecurity), political instability and particularly the rapid development of new technologies such as artificial intelligence, drones, synthetic biology, and so on. Suleyman has characterised this as an impending wave that will hit humanity with an almost irresistible force comparable to the invention of the wheel, or the industrial revolution (Suleyman, 2023). We are ill prepared for it. That said, we can and should think carefully about the implications of this coming wave and what can be done to mitigate the potential threats, whilst acknowledging the enormous positive opportunities it promises. To answer the question posed in the title of this article – no, crime science cannot save the world, but as is argued below it can offer a radically different approach to crime control, which can contribute to better preparing us for the future.  

What is Crime Science?

Crime science as a discipline was first taught at University College London (UCL) in 2001, when the Jill Dando Institute of Crime Science (the JDI) was established. In the Handbook of Crime Science, it is defined as the application of scientific methods and knowledge from many disciplines to the development of practical and ethical ways to reduce crime and increase security (Wortley et al 2018:2). Traditional criminology typically focusses on the criminal, crime science focusses on the crime and its reduction. Just as medical science draws on physics, pharmacology, genetics, biochemistry etc., crime science draws on the social, physical, and computer sciences in looking to reduce crime. It takes the criminal law as a given, although individual crime scientists may have personal reasons for excluding some crimes from study. 

Perhaps surprisingly, the Department of Security and Crime Science at UCL is in the Faculty of Engineering Sciences. Engineers solve problems. For example, an engineer might be asked to build a bridge across a wide river. Engineers have learned to build complex structures like bridges after many years of learning from trial and error, which has built up a body of knowledge from mathematics, physics, material sciences and other disciplines. In getting to this point, they have had to learn from their mistakes. This is what crime scientists also do – in trying to control crime they define the problem in detail (domestic burglary on a specific housing estate); analyse the characteristics of the problem (entry point, time, date, location of burglaries in the area over time etc); think about possible responses that might reduce the problem (typically a series of situational measures – better security, neighbourhood watch, etc); implement the measures and check if they worked. Essentially this is the application of scientific method to the problem of crime reduction – it is experimental. If it doesn’t work, try again with something else. 

What have crime scientist done so far?

Lots! In this section we briefly discuss the international crime drop and then consider the work of crime scientists around predictive policing and its relevance to future crime control.

Over the past 30-40 years volume crimes (vehicle related crime, burglary, theft) have all dropped significantly in those jurisdictions where crime can be reasonably accurately measured (Van Dyke, et al, 2012). Crime scientists argue that this is because opportunities cause crime and the opportunities for crime in the latter part of the 20^th Century were high. Once those opportunities were blocked, crime dropped – generally through better security such as deadlocks and immobilisers on vehicles (Farrell, et al, 2011). Reducing crime opportunities is related to the redesign of goods, services and management systems; it is not related to dealing with offenders, although some crime scientists work closely with forensic scientists in attempting to increase the risk of detection. 

One of the significant insights of crime science is the importance of repeat victimisation as a focus for crime reduction. Historically, in England and Wales, about 4% of victims suffered 44% of crime. Domestic violence is the obvious example of repeat victimisation, but it is also common in burglary, car crime, racial attacks, school bullying, assault and other offences. We also know that when the repeat occurs it is often relatively quickly after the first incident (Farrell and Pease, 2001). Additionally, as Bowers and Johnson have shown, if a burglary victim has been identified, particularly in a relatively high crime area, then their immediate neighbours are at increased risk of victimisation for a few days following the initial incident and it is therefore worth increasing police patrols to deter the offenders (Bowers and Johnson, 2004). Studies such as those carried out by Bowers, Johnson and colleagues have formed the academic basis for what became ‘predictive policing’. But in this case, it is predictive policing based on the crime, not the characteristics of the offender. As such it avoids the dangers of stereotyping or systemic bias; crime science is concerned with reducing crime ethically.

According to the RAND Corporation there are now four different types of predictive policing: methods for predicting crimes, methods for predicting offenders, methods for predicting perpetrators’ identities, and methods for predicting victims of crime (Rand, 2013). Their report makes the important point that this work would be better described as forecasting, being based on data akin to attempts to forecast the weather. It is not an exact science by any means, but for now we will stick with the more commonly used label of ‘predictive’ policing.

It is the methods for ‘predicting’ crimes and the victims of crimes that have formed the focus of the work of crime scientists at UCL. Attempts to predict who will offend in the future are fraught with statistical and ethical problems, including the use of algorithms that are vulnerable to systemic bias. Predicting the characteristics of perpetrators, also sometimes known as offender profiling is also problematic although geographic profiling (where predictions are made on where offenders might live or work based their criminal activities) can support detection. 

There are strong theoretical grounds for suggesting that by protecting victims, further victimisation can be prevented. This is the case across a wide range of offences reflecting the observation that repeat victimisation is far more common than is normally assumed (Farrell and Pease, 1993). The first evaluation of the effect of protecting crime victims from further victimisation resulted in a decrease in domestic burglary of almost 70% on a particularly high crime estate in the north of England (Forrester et al, 1988). Since then there has been further evidence of the reduction of domestic burglary following the protection of existing victims (UK College of Policing, 2023), although more experimental evidence is needed in relation to other offences. Focussing preventive effort on victims, with the intention of preventing a repeat offence, is a form of forecasting, and as such it too can be seen as ‘predictive policing’. 

Assessing the effectiveness of an intervention requires the careful comparison of crime data from say, 12 months before an intervention, with 12 months after. It also requires a comparison, or control area of some kind. This is not the place to go into the various methods of evaluation but suffice it to say that crime science sees opportunity as a major cause of crime. Blocking opportunities reduces crime regardless of social inequalities, poor parenting or any of the other risk factors for criminal behaviour. 

What works in crime reduction?

There are many ways in which opportunities for crime can be reduced which have been classified under five headings: Increase perceived risk, increase perceived effort, reduce rewards, reduce provocation and remove excuses. These activities are intended to affect the potential offenders’ decision processes. So, introducing CCTV into a car park, for example, would be intended to increase an offender’s assessment of risk (I might be seen) and effort (if it involved finding a part of the car park that was not under surveillance), thus reducing the likelihood of the offence being committed. Research has shown that CCTV in car parks does indeed reduce offending (Tilley, 1993). In contrast, there is no evidence that it reduces disorder on the streets when pubs and clubs close, because many of the potential offenders are drunk. In this context, CCTV might help the police to detect offending or deploy resources, but not necessarily reduce crime. The results of several systematic reviews on the effectiveness of CCTV in crime control are available on the UK College of Policing website (College of Policing, 2021). 

In deciding on what works, there are no universal answers: You need to consider its expected effect size; how it is expected to work (the mechanism); where or in what context it might work; how to successfully implement it, and whether it is economical. These elements have been summarised using the acronym EMMIE (effect, mechanism, moderator (e.g. context), implementation and economics), which is used to assist practical decision making on the UK College of Policing’s Crime Reduction Toolkit. 

Most of the work described above originates in the UK. But context matters in crime control. What works, or what is relevant in the UK is not necessarily right for other jurisdictions, although it may have been in the past, or could be in the future. What lessons have we learned that might be useful to others? First, on the question of ‘what works’ in crime reduction – nothing works everywhere, lots of things work somewhere, nothing happens because someone said it should, sometimes initiatives are expensive and sometimes the same initiatives are free and time changes almost everything. Having said that, if you are not concerned with civil liberties, human rights, police legitimacy and ethical crime control, then increasing surveillance and all aspects of predictive policing are extremely powerful and could be used to great effect as a means of controlling members of society. But the technologies that are emerging – artificial intelligence and biological/genetic markers to mention just two – are going to offer even more powerful opportunities to apply coercive control over citizens, but there will be errors associated with such prediction or forecasting since, like weather forecasts, it depends on data. If ethical issues are of no concern, however, then these errors might well be disregarded. 

The question for democracies, which at least claim to be concerned with ethics in the broadest sense, is how to maintain a balance between what is possible and what is right. The application of the law has always been biased. As Shakespeare said, “Through tattered clothes great vices do appear; Robes and furred gowns hide all.” The rich can quite literally get away with murder. But for those who care about justice, and hope for its even social application, how might they approach the coming technological opportunities for oppressive surveillance and more? 

How can we avoid abusive methods?

In the UK there are strict laws about the collection and legitimate use of data (originating from membership of the European Union). There is at least an awareness of the biases and errors that can arise from the thoughtless analysis of large and error prone data sets. These biases and errors are not condoned, but neither are they always identified. There is also a free press, which protects journalists and contributes to exposing injustices (of which there are many); there are laws against libel; freedom of speech and human rights are protected, and there are regulatory systems and NGOs – all alert to the dangers of drifting from what is legitimate to what is not. But it is undoubtedly a difficult judgement to make – the difference between legitimate and illegitimate use of power is not black or white; there is a significant grey area in between. There thus needs to be constant vigilance against the potential abuse of new technologies by states and other actors, together with a clear position from the various governments on what can and cannot be regarded as legitimate. Perhaps the best that might be claimed by an honest politician is that we tried. 

So, who is responsible for ensuring the legitimate use of the massive and increasing amount of personal data, surveillance systems, our DNA, existing and, more worryingly, developing technologies? We all expect to live in safety and security with our right to privacy and respect. But it is a difficult balance to get right. Perhaps the bottom line is that we are all responsible for contributing to the kind of society in which we live. 

What is crime science addressing in the future?

In 2016, UCL established the Dawes Centre for Future Crime in the Jill Dando Institute. The Centre is concerned with the developing technologies that offer opportunities for committing new offences, or new ways of committing old offences  (Johnson, 2021) . It is a vast agenda given the speed with which technologies are developing and the lack of thought being given to their potential effects on crime, including the potential for abuse by rogue states. Current research includes studies on social robots and anti-social behaviour; climate technologies and future crime, and AI based solutions to analyse online fraud (Dawes Centre for future crime, 2025).

It is already clear that with the arrival of the Internet, the opportunities for transnational crime are considerable and the possibility of arresting and prosecuting offenders who may be resident in another country is problematic. This situation highlights the importance of controlling criminal opportunities and reducing them, rather than relying on the operation of the police and criminal justice system for crime control. It reminds us of the importance of seeing crime as a problem to be solved by whatever ethical means available, and increasingly will imply that commercial organisations, industrial concerns and a wide range of disciplines will have an important role to play in maintaining safety and security.