@@@@@@@

Please note: the following essay  is very much a tongue-in-cheek work in progress

 

 

 

Original Dysology.org essay (16^th May 2011). All rights reserved.

 

Before the Crime Forecast: A tongue-in-cheek demonstration project exploratory exercise to examine what a Real Junk Crime Science  general theory of crime might look like and a serious examination of what current so called Crime Science is and is not

 

                                                                                 Mike Sutton ©

 

  

Abstract

 

In the Western industrialised world, national crime survey data and other official statistics since the mid 1990’s suggest that, with some exceptions, there has been an overall general crime drop.  This has baffled criminologists, not least because it bucks trends identified by post-hoc econometric research suggesting that, overall, acquisitive crimes rise during periods of economic recession and violent crime increases during times of prosperity. In an effort to make sense of the ‘ unfathomable crime drop’ (Young 2004), some criminologists are looking for best post-hoc explanations with an aim to inform future policy oriented criminology.  But this approach is irrational because crime is a social construct determined in large part by the ever changing affairs of people including: policymakers; lawmakers; enforcers; law breakers; victims and policy oriented criminologists. This means that seeking to predict the past and expecting that the future will resemble it is as irrational as the principle of induction it adopts. Predicting the past and inventing falsehoods is an easy and popular pursuit when there is a wide variety of competing and easy to vary explanations to choose from. But finding ways to justify our beliefs with poor explanations does not aid the growth of knowledge.  From this critical observation springs a most critical question: ‘Why were those criminologists who seek now to be wise after the event unable to predict the crime drop with the hard to disprove pet theories that they are now so easily able to vary to suit this unforeseen outcome?’  If, in theory, scientific principles might just be usefully employed in the pursuit of theory development leading to accurate prediction in criminology, then the only way to test that theory is to adopt a more rational way forward for knowledge development. Since science is the means we employ to stop us from deluding ourselves with falsehoods and certainties, a real 'crime science' would be founded on the principle stated by Karl Popper that we need to develop easily refutable, verifiable and hard to vary theories. A testable Dynamic Vulnerability (DV) theory that is hard to vary would better explain crime if it could forecast crime levels,  waves and drops accurately. Renaming it with something that sounds intellectually inferior - such as Robin Hood theory - would help to ensure we do not adopt it as a pet theory, but rather do our very best to neglect giving it any biased nourishment while seeking to objectively disprove or surpass it with better knowledge.

 

Introduction

For many years some social scientists have successfully argued that that their particular methods of inquiry are different from those of the natural sciences because they are concerned with the ever changing affairs of the free will and individualism of people, which do not conform to such predictable principles and laws as we find in the physical world, and that consequently different research methods need to be employed to understand the social phenomena of mankind (Krausz and Miller 1974). I do not hold any objection to those arguments with regard to areas such as, for example, cultural anthropology and cultural criminology. However, where we are concerned with understanding regularities and recurrent social events such as crime patterns, shifts, waves and drops, I am in agreement with Schütz (1970) that the need for a set of rules for scientific procedure is equally valid for all empirical sciences. Yet some criminologists (see Laycock 2003) have launched Crime Science on the premise that the criminological notion of opportunity is the most important cause of crime, which confuses the data they are seekng to explain with the explantion (see Sutton 2011a).

In science, facts are the properties of natural phenomena and the scientific method involves testing principles that we suspect might best explain particular natural phenomena. Tilley and Laycock (2002) and Laycock (2003) claim that: “The most significant and universal cause of crime is opportunity. If there were no opportunities there would be no crimes; sadly the same cannot be said for any of the other contributory causes”. Although Laycock and other crime scientists beg the question regarding what opportunity really means, from her statement it is also unclear whether the crime science notion of  'opportunity' is thought of as data or as an explanatory principle. Presumably, specific crime types are most certainly to be considered as data. For further discussion on this particular problem for crime science see Sutton 2011a.

 More recently, others allied to this so called crime science  (e.g. Farrell et al) seek to forecast crime by studying its past trends. Simply calling something science does not make it so and seeking to predict the future with reference to the past is both irrational (Sutherland 2007) and contrary to Enlightenment philosophy of science (Popper 1963). In response to this 'movement' away from social science explantions of crime, the sole purpose of writing this article  is to examine, as a theoretical excercise, what a more truly 'Real Crime Science' scientific method of inquiry might just be - in terms of embracing the principles of controlled inference and verification along with the ideals of unity, simplicity, precision and universality that underpin science.

If we really could examine crime trends and the possibility of accurately forecasting crime in this way, I believe it would bring considerable benefits for knowledge development in criminology. And yet, ironically, to have a chance of accurately predicting what populations will do next in their complex and ever changing social world, the new scientific criminological theory that is proposed and tentatively outlined in this article depends upon masses of the type of 'thick' (rich) data that is used in the necessarily less scientific research of qualitative criminology such as symbolic interactionism and the newer work of cultural criminology.

While critical of the limitations of SCP and RAT in helping us to understand, explain and predict crime trends, this new theory, like other good scientific explanations must explain what is wrong with the explanations it seeks to replace and, where appropriate, incorporate other explanations that are not 'disproven' by them. In that scientific tradition, the new theory outlined in this article relies in no small part upon the rules of thumb and theoretical high level frameworks of SCP and RAT's interpretation, classification and understanding in order to know what quantitative data to gather and how to understand it in order to accurately forecast crime and, also in part, to explain the principle behind the theory.

 

The crime drop

Bucking the trend predicted by Field’s econometric analysis of crime in England and Wales (Field 1990), crime has been falling, with some exceptions, pretty much across the board, in Europe and the USA since 1995. At the time of writing there remains no particularly compelling or well evidenced cause since Young (2004) described this as unfathomable.

Several months ago I wrote speculative peer-to-peer a blog post (Sutton 2010b) that asks whether the communications revolution, which includes recordable non-terrestrial television programming, DVD technology, mobile digital telephony, video games and the internet might have played a significant causal role in reducing non-internet facilitated acquisitive crimes such as vehicle crime and burglary and if it has, whether this should have been predicted and can be explained by RAT.

RAT is a theory of how crime shifts and changes in relation to changes in society - and according to Felson and Boba (2010)  the key to such changes is the technology of everyday life - then a significant amount of online activity and games console playing at home might be keeping people off the streets for significant periods every day and reducing the amount of hourly availability of victims and offenders in the offline world, while ensuring homes are occupied more, thus providing more guardianship against burglary. If it is, and yet that is not responsible for the currently unfathomable 15 year decline in violent crime, car crime and burglary in the Western world then RAT may be nothing like as good an explanation for high volume crime as its adherents believe.

That blog post goes on to speculate whether the 15 year crime drop in the offline world may be taking place in some kind of casually inverse displacement relationship with internet facilitated offending such as fraud, stalking, virus spreading, hacking and copyright theft,  which RAT would explain. On the face of it this appears to be an important issue with huge implications for criminology and crime reduction policy making.

 With the advent and increasing popularity - indeed, necessity - of the internet and the huge rise in mobile communications technology, we might have expected crime to rise dramatically from 1995 as a whole new environment called cyberspace opened up for people to exploit in criminal ways  old and new. Felson's RAT (Felson and Boba 2010 p.111) certainly sees that it should be that way after it has happened:

"The age of speedy Internet communications provides new options for youths to break laws, often operating out of their homes. They can produce their own pornography. They can view pornography by others. They can sell themselves as prostitutes. They can make sexual liaisons with those of their own ages or well beyond their own. They can send and/or receive threats via the Internet and buy or sell contraband goods. They can, at a young age, learn how to hack the computers of others or distribute computer harm in various ways. They can participate in cyber chat rooms to discuss all of this."

Is the RAT in the net, video console, TV and DVD?

While noting that teleworking might well keep more people at home as capable guardians against burglary and other daytime crime in residential areas, (Felson and Boba 2010: p. 204). Strangely, Felson appears not to have considered earlier than this that the internet was going to create a whole new environment for crime and perhaps stranger still he appears to have failed to see that his own RAT would explain that all this time spent online might equate to less time on the street for everyone - not just teleworkers - leading to less potential offending time and a smaller population of available victims of violence and robbery, car crime and burglary - which might explain the crime drop offline.

While the crime drop is not something that Felson or any other advocate of his theory used RAT to predict would happen, Felson and Boba (2010) support what is apparently an opposing view of Jacobs’ (1961) that it could. Jacob’s classic book included her intuitively evidenced contention that stable and likeable neighbourhoods have less crime because more people are actually out on the streets as good citizens providing capable guardianship against crime. For Jacobs, the US crime rates of the late 1950’s and early 60’s were to be explained by poor housing and neighbourhood design and in no small part by the invention of television that kept those capable guardians indoors – leaving the neighbourhood to delinquent types who did not appreciate the most popular US programming of the time that was typified by Texaco Star Theatre, Philco TV Playhouse, The Colgate Comedy Hour, The Lone Ranger and Hopalong Cassidy. But what if TV started producing programming that would keep potential delinquents and their young victims indoors as well?

On this theme, following the one hour time-slot in the 1960’s when the Beatles played live on US TV, a story took hold in the imaginations of many that not even a hub cap was stolen because every thief was indoors. No doubt they were believed to have been sitting alongside all those latter-day capable guardians, with their eyes not on the street but glued to the box in the corner.

For all we know the Beatles may indeed have reduced crime to some extent by gluing offenders to the TV. Perhaps we’ll never know so long after the event. However, if records are still available, it would be possible to look for an unexpected drop occurring in that one hour. Until then, we can only be sure that the widely believed story told that they actually did so is a complete fabrication. because it was born of no more than a sarcastic one-liner on a TV show (Snopes 2010).

 Serious research into the possible causal relationship between the Communications Revolution, that was  gathering startling momentum from the mid-1990s, and crime is today an increasingly explored area built upon a wealth of early pioneering criminological literature on the subject of Cybercrime (e.g. Landreth and Rheingold 1985; Clough and Mungo 1992;  Meyer 1989 Mann and Sutton 1998; Wall 1997, 2003, 2009) along with  many social psychology studies of the effect of various hi-tech media exposure on risky behaviour (e.g.  Fischer et al 2011).

The Communications Revolution arguably began in earnest with the 19th century invention of the telegraph, which was followed by the telephone, radio and television. But it really began snowballing in a way nobody foresaw following connectivity to the internet, the invention of the publically available World Wide Web in 1991 and the invention of Web browsers in 1993. This happened at the same time that cell phones were growing in popularity and sparking an unpredicted crime wave of supply by theft that began in the first half of the 90’s (Parliament Post Note 1995) and continued throughout the following decade (Sutton, Hodgkinson and Levi 2008; Farrell et al 2010).

The relationship between the design and infrastructure of the internet, offenders, victims and crime remains to date largely unexplored by criminologists, although Newman and Clarke’s (2003) pioneering work does make some useful inroads.

Meanwhile, traditional Environmental Criminology, from the work of Oscar Newman to the Brantinghams, is particularly concerned with linear explanations for the causal relationship of the built environment and crime. Felson and Boba (2010) typify this with the example of how a former student of the Brantinghams ensured that new construction in Vancouver seeks to minimise crime ‘opportunities’.

Beyond this relatively obvious approach, Environmental Criminology, and criminology generally, has not examined whether  the very idea of crime - as an example of what Dawkins (1991) refers to as a meme - might have shaped the environment in ways that are hard to see. One notable exception is Ekblom’s (2002) consideration of how crime reduction knowledge is transferred.

 Memes are ideas that take hold and become embedded in human culture, mutate and become culturally inherited in the same way we inherit genes. This is what has happened with knowledge about crime, how, why, when, where and against whom to commit it, and about criminals and victims. This cultural knowledge determined both early crime policy making, and all our individual delinquent and non-delinquent solutions to it. Crime then became physically embedded in and shaped the built environment, continually looping back to shape cultural understandings about crime, the economy, policymaking, criminality, offending and the criminal justice system reaction. Looking at it in this way we can think about crime as something that can be regulated by changes to the environment but also that the idea of crime is a meme (Dawkins 1991) that is regulating its own environment, which includes environmental criminologists, their students, their work in the environment, those such as Felson and Boba (2010) who write about them, me writing these words, and you reading them.

Taking this notion forward with a real example helps. To provide just one crime type example, that will be used later in this paper, people know exactly where to find scrap metal to steal from their environment. Aluminium can be got from road signs, bicycles and beer kegs. Copper can be stolen from electricity sub-stations, pipes in all buildings and railway electrical apparatus. Gold resides on our fingers and around our wrists and necks, inside jewellery boxes in houses and jewellers shops. Bronze resides in statues. Platinum is in motor vehicle catalytic converters. Cast iron forms manhole and drain covers. And lead covers the roofs and other weather protective flashings of old buildings. All of this metal is protected by the relative difficulty of accessing and removing it without being physically harmed or apprehended in the act of stealing, transporting and selling. But large enough shortages of scrap metal creates a demand that fuels enough motivation so that offenders override that previously capable guardianship that was provided by the laws of gravity (falling off high buildings) magnetism (dangers of electricity), physical effort, and people with eyes on the street. If theft of scrap metal continues for much longer that crime fact will simply have to shape the level of use of different metals (replacement with an alternative material) and/or protection in the environment, otherwise, I hypothesise, so long as demand is high enough the self-regulating socio-economic crime system would continue as follows: (a) scrap metal theft would pay so well that a major national police crackdown might be needed to ensure social and economic order if (b) there are no cost-effective suitable replacement materials and no effective security to trump the level of demand fuelling criminal motivation. Then a major law enforcement crackdown on offenders might run some (knowable in the future) degree of risk of stimulating outbreaks of civil unrest but (c) this would not lead to a nationwide breakdown in law and order, so long as theives were not culturally bound to scrap metal theft, because there would be a self-regulating displacement of a significant number of offenders to other targets until things die down just enough to make scrap metal theft a rational choice for all scrap metal thieves once again - so long as the demand remains. And (d) that continuing demand will have been shaped to some degree both internationally and nationally by 'tolerance' for stolen metal within the main marketplace for scrap metal. 

Meme's can be both rational and irrational. And collections of meme's that work together in some way are called a memeplex. When we refer to the idea of crime being a meme that is not incorrect, but this one meme has gone on to create narrower ideas about crime and crimes that form a long-standing yet evolving crime memeplex (Sutton 2011d). Long standing memes are those that are useful to people who for long periods in our history have had diverse purposes (Deutsch 2011 p. 388). The idea of crime is not new. Rather it is a long lived meme. Meme's have to compete for the recipients attention, perhaps by containing knowledge of some useful function (Deutsch 2011 p. 378) such as "Steal and sell or buy stolen goods." and: "Protect against theft." That crime has not destroyed our society may mean that it benefits us as a whole.

Perhaps crime is the selfish solution to our selfish society? What we are talking about here than is crime as "The Selfish Meme."

Crime, the Selfish Meme. Let's torture a junk science theory that is based on very little evidence

At this point it is important that I point out to criminologists that this hypothesis does not make the same mistake as Durkheim (1938) in that it does not work on any principle of simple naive functionalism. Nor can it be constrained within the limitations of Merton’s (1938) explanation of crime arising from anomie. If the crime memeplex does work at times as a boundary maintenance system then that is a beneficial by-product that plays a small role in the entire self-regulating social system. And if this memeplex includes delinquent solutions to the disjunction between goals and means it does likewise for crimes of the rich and powerful.  The explanation that crime is a self-regulating memeplex accounts for crimes as routine activities and it allows one to seek to examine how all of this came to be.  Likewise, this paradigm shifting hypothesis does not make the same mistake as Karl Marx's (see Marx 1964) naïve notion of evolution - because we now know that evolution is not at all about survival of the fittest. What genes do and what memes do to survive is first and foremost for their own benefit and that ensures their own survival, not necessarily their respective biological or social hosts.

This dynamic hypothesis explains predictable human behaviour regularities that are measurable as crime waves at the national level. At the local level, it is hypothesised that something similar happens with regard to displacement following extraordinary crime reduction and crack-down interventions in high crime areas, which is discussed in more detail below regarding how we can seek to refute the theory.

According to this hypothesis, as supported by an as yet unpublished (forthcoming) overview of the published research into displacement (Phillips 2011), crime displacement following situational crime prevention and other crime reduction measures is in fact the norm - despite the best efforts of SCP and RAT adherents to imply that the facts are otherwise. If that was not so then we would be living in a world of magic and miracles that would be a very different place from the one we know.

The fact is that that despite centuries of major crime waves and new crime harvests, often fuelled from shortages in supply or demand for products, materials and people from other nations, our lives in the social world and built environment and the way we humans are able to navigate and regulate it with expanding and declining empires has remained pretty much the same for centuries despite major world conflicts and social, political, economic and technological advancement. Crime rises and falls, but there never has been a golden age where crime did not significantly bother and enrich our lives (Pearson 1983).

This ability of criminal behaviour, and our responses to it, to adapt and survive suggests that the meme of crime might be in some way self-regulating crime by way of its influence upon offender reaction to the level and type of official and unofficial responses to it on much of our planet. This is not as tautological as it might appear. Because we can begin to test this tentative outline of an explanatory principle with an aim to understand why, and therefore predict, what crime does next.

To begin to do that, with an aim to forecast crime, we will need to develop this ambitious data hungry theory to help us seek to determine what generally constitutes enough criminal motivation, and how, where, when and why it is then realised. And we need to know why and how frequently offenders tend to deploy it - or not - and against what, before, during and after different measurable levels of crackdowns, replacement and protective measures. This is something I deal with below and it’s not simple because if it was it would not be such a difficult explanation to find for the complexity of crime. In the meantime, let's take a break with a slight mental detour and consider what would happen if we could accurately forecast crime.

Might accurate crime forecasting have derailed China’s economic boom and prevented the credit crunch?

 Research shows that prolific thieves do not tend to steal many items for their own personal consumption, instead  they steal what they know are valuable items that other people will buy because those items are in demand (Sutton 1995, 1998; Clark 1999) and what they really require is cash to buy something else that they want. This is why, to provide just on example, a shortage of lead after the second world war fuelled a crime wave of lead being stripped from the roofs of historic buildings throughout the 1950’s and 60’s to fuel the massive building programs of the time and the expansion in manufacturing. This could well be a regularity in that it has occurred many times before in our history and has reoccurred more recently, during the current overall crime drop, due to demand from China (Copping 2007; Kooi 2010).

What would have happened to China’s booming economy over the past decade, and to individuals and economies elsewhere depending upon the illicit supply of scrap metal, if the recent crime wave of metal theft that has swept the world had been forecast and so thwarted politically through a ban on scrap metal exports? We might speculate that it would hit the Chinese economy, and by consequence Western economies relying upon Chinese credit to buy cheap Chinese goods, with far more likelihood than the beating of a butterfly’s wings (Lorenz 2000) might cause a hurricane; because knowledge has infinate reach (Deutsch 2011). But such knowledge - such a forecast - leading to a political decision to crackdown on exports, might also have prevented the credit crunch and economic meltdown in the West that was caused by the reckless lending of Chinese fuelled credit in the first place.

Crime has phenomenal ‘reach’ into human affairs

I propose that it is a mere truism that physically imbedded crime knowledge must be self-regulating crime, albeit to some as yet unimaginable extent.  We know this not only from the scrap metal example provided above but because, at the most basic level, a lot of target hardening comprises physically imbedding the idea of crime in the environment using rules of thumb that we know reduce individual risk. If you doubt this, then during a crime drop or not, simply leave your car with the doors locked, windows closed and alarm turned on in a high crime area with a satellite navigation device stuck to your window or a laptop computer on the seat, retire a few hundred yards and observe for about ten minutes or so. Then try the same thing in a very low crime area. And then just to prove that things work differently in different places try it in both areas with the windows open, doors unlocked and alarm turned off. Or if you don’t own a car then don’t take regular solitary strolls in the evening along lonely footpaths that demarcate the very edges of the easy to recognise architecture and layout of notorious, less desirable, high crime areas that you care to avoid and the relatively safe environment with which you are more familiar (Felson and Boba 2010), unless you want to check out this truism by being robbed or worse. But you might want to do that in relative safety walking alongside a large or well-armed uniformed police officer.

By way of further evidence that the idea of crime is self-regulating crime is the fact that that some highly localised degree of displacement can be observed with the naked eye by those familiar with high crime environments. I am grateful for the comment from my colleague Roger Hopkins Burke who raises the fact that it is not rare to see teenagers walking along a row of parked cars and trying the door handles of each. When they find a locked one they pass on to the next and so on. Roger has seen this happen on many occasions and so have I. We were both born and brought up on British council estates. In this way, thieves adopt, for opposite reasons, the same strategy that beat police officers of old used when pushing on doors and tugging the padlocks of steel  shutters of high street shops as they progressed along their way. Residents of high crime neighbourhoods see this brazen behaviour by criminals on a regular basis. If they are not tugging on the car door handles then they are riding up to the cars on bicycles and stopping by each to look in to see whether it is worth breaking the car window to steal something inside it. You may only occasionally witness the window being broken (although I have seen this happen, and so has my colleague Mike Ahearne), but you will see them being displaced from car to car with the idea of crime in their mind responding to the idea of crime in the car owners mind. This is part of the irrational side of SCP and RAT adherent's reliance upon Rational Choice Theory applying to criminal decision making when the odds of finding an unlocked car, or a car with something valuable inside, may not be favourable to the effort involved in finding it.

 

The odds of winning the lottery are not favourable to the ticket buyer. So would SCP and RAT adherents argue that a punter's decision to gamble is rational within the limits of the time they have to make up their mind whether to buy a ticket or scratch card?

Whereas national lotteries are a tax of those who are bad at math, some crime is a tax collected by those who are both bad and bad at math. Looked at this way, simply using something called Rational Choice Theory to seek to explain crime and argue that it is not normally displaced does not make your use of it rational or good 'crime science' simply, by applying your own 'after theory' easy to vary caveat that the rationality of offenders is bounded (e.g. see: Laycock 2005. p.678).  

Underpinning DV theory, by way of contrast, is a hypothesis that is the opposite of one that underpins SCP and RAT. DV theory is based on the premise that offenders do not conform to Rational Choice Theory at all. And this new criminological hypothesis rationally removes SCP and RATs own  perversely irrational slippery requirement for an easy to vary ‘after theory’ bolt on fix that an offender's rational choice in weighing risks and rewards may in fact be rational even when it is objectively irrational.

 

 Furthermore, we can question the rationality of depending upon any straight use or hack customisation of Rational Choice Theory by asking where is the research evidence to support this assumption that offenders weigh risks and rewards? If they did that then why do so many get caught? The answer is surely because they make irrational decisions. If that were not so then would there not be more examples of successful criminals? Furthermore, research into Cognitive Dissonance (Tavris and Aronson 2008) shows us just how much of what we do is determined by our own irrational decisions and self-justification. If academics, as the Decline Effect confirms, do not rationally weigh the confirmatory and non-confirmatory evidence for their theories how can those same academics possibly build a theory on the assumption that offenders do better?

 

Dealing with this dilemma is not going to be simple. How for example do we explain why an abusive spouse tends not to assault their partner in public but chooses instead to hit them at home behind closed doors? Perhaps they are not weighing up the risks versus rewards of violent assault in public but rather sticking with some general rule of thumb that they have learned as meme as the best way to get away with their crime?  Clearly, more research is needed in this area to see whether what we need is an Irrational Choice Theory to take knowledge forward in this area.

 

Arguably, to find an alternative and good explanation for spur of the moment offender decisions as to whether or not to exploit a perceived vulnerability, we need to look for one that is hard to vary so that it is either right or wrong. I propose that the best explanation is to develop an Irrational Choice Theory that builds upon the Nobel Prize winning Arrow’s Theorem (Arrow 1951; Hunt 2007; Howard et al 1993). The natural conclusion to be drawn for Arrow’s Theorem is that alternative choices of human decision making are not weighed and outweighed by the individual at all. Instead they are argued, refuted and abandoned. In effect, potential offenders are responding to their potential victims, who by the laws of nature, and according to Arrow’s Theorem, are ‘dictators’ over their own lives, bodies and possessions (Deutsch 2011: p.342-343) by deciding which injustices and irrationalities they like best. Potential offenders then carry out their law breaking or conformity to laws accordingly.

 

A new understanding of crime and environment

For thousands of years, the physically imbedded idea of crime will - to a currently unfathomable extent – have been self-regulated itself and crime. To see further how this must be so, we need only consider the impact on human movement, all other behaviours and culture of  the billons of  locks and bolts, railings, alarms, grills, shutters, gates, walls, doors, street lights, turnstiles and other security apparatus that has been invented,  multiplied and taken hold in the visible Western world since the iron age. Then add to this the multitude of prominently placed police stations, courts of law, prisons, university criminology and psychology departments, libraries, government buildings, and associated vehicles, official uniforms, guard dogs, warning signs, property marking, books, other printed media and movies and TV programming.

Entertainment centers such as the London Dungeon, The Torture Museum in Amsterdam and Scotland Yard's own private 'Black' Museum (lately renamed the Crime Museum) are full of stuff like old torture devices, official devices of cruel physical punishment such as stocks and pillories, gallows, garrottes, guillotines, leg irons, kneck irons, scolds bridles. They are stocked with locks and keys, and non-military weapons. Listed buildings have thick, iron braced oak doors; and old coins have milled edges to prevent scrap metal theft. These objects, like hundreds of others, all contain the public and the policymaker's idea of crime and what to do about it. The idea of crime and its influence upon us is so deeply embedded in the fabric of our built environment that it exists in abundance in the archaeological strata below it as well.

The internet and crime

By way of contrast to our built environment, the internet is a new environment that has been described as a virtual reality that exists not in real space, certainly not in the built environment that we physically inhabit to protect us from our hostile planet, but in a cyberspace we inhabit in our minds. While the physical properties of the internet, the computers and the communications infrastructure needed to connect them are to an extent an addition to our new built environment on Earth and in its orbit, cyberspace was born and has grown not only separately from the physical world that people have inhabited since the emergence of mankind, but as an environment within which the idea of crime has not – yet at least - played such a large part.

Crime and cyberspace is relatively new but is no longer a brand new phenomenon. The Victorian telegraph system, for example, appears to have had an early form of interconnected cyberspace community complete with hackers and wire fraud (Standage 1998) and the blind community in the USA played a large part in facilitating and spreading phone phreaking information in the 1960’s (Clough and Mungo 1992). But in terms of the physical infrastructure required to support it, the sheer number of people using it and the proportion of their daily lives spent on it, mankind has never before experienced anything approximating the scale of the virtual environment of the internet. And with more people comes more crime.

Unlike the traditional built environment of our societies, which  evolved with the idea of crime, the internet was designed primarily to be anarchic so that it could survive as a self-re-routing military communications network in the event of nuclear strikes on US cities (Denning 1989) and from that cause, whereby unitary control and ownership of the system was avoided,  its early post-military development as an educational resource was very much influenced by an undercurrent philosophy of open access, horizontalization of knowledge and freedom of speech that led to the development, and early facilitation, of a criminal hacker-counter culture when the idea of internet crime began to take hold.

Given such rich technical and cultural explanations for online deviance and the wealth of information available from a multitude of lifestyle surveys (e.g. Harris 2009) that show year-on-year how much more time people are spending indoors online, watching TV and playing video games, I could make a powerful, intuitive and compelling post-hoc case in this paper that RAT would explain how and why the Communications Revolution is an important causal factor in the rise of relatively uninhibited and unhindered internet crime, the crime drop offline, and the rise of some offline crimes such as mobile phone theft that have bucked the overall trend.

How might I do that, if I so wished?  Well, I might begin with social psychology research findings that found computer games can be addictive due to immersion and unreality factors. I might also use RAT to speculate that an increase in computer gaming might feasibly lead to a rise in the illicit market for stolen PC's and games consoles. But hypothesise that there might be fewer thieves to supply it if, for example, fewer potential offenders are getting addicted to opiates and other drugs or misusing alcohol out of boredom because they have escaped boredom in the real world by entering a more exciting cyberspace to play and interact with others. RAT could be used to hypothesise that it is because so many video game players and other people are viewing media and video gaming at home for so many hours that there are fewer potential offenders on the streets and fewer potential victims, while homes are occupied for longer and so less susceptible to burglary. I could distil this idea into a simple hypothesis that I might call Game Substitution, which would state: “A significant rise in Routine Activities indoors leads to a significant rise in crimes facilitated by indoors communications media such as the internet and regulates a consequential reduction in crime outside.”

Armed with my new pet hypothesis I could set about looking for confirmatory evidence to support it. I might include some compelling gems such as Hagell and Newburn’s (1994) study, the conclusion of which implies that there is a need for more to research to determine whether a wider choice of TV viewing might reduce crime among young offenders. I could back that up with my own analysis of secondary data sets and published research, of the year-on-year increase of hours people are spending engrossed in media and video gaming (e.g. see: Media Literacy Clearinghouse 2011), in order to demonstrate the correlation between the increased time spent viewing various forms of media and video gaming and the crime drop offline.

I could make a compelling post-hoc case that the Communications Revolution is a significant cause of the crime drop and apply for research funding to test my hypothesis by way of a demonstration project to measure whether providing those at-risk of offending and/or victim groups with more choice of TV shows, video gaming and high speed internet access reduces offending and victimisation. And yet, after all that, despite my best endeavours to examine threats to theory, it is most likely that my efforts would become another statistic representing the notorious Decline Effect  (Lehrer 2010) when others conduct similar research to test my ground breaking conclusions. Why? Because I suspect that my research would inevitably be far too biased towards confirmation of my pet theory as opposed to my best attempts at falsification (Sutherland 2007; Popper 1963).  Further, building theories on observations of the past is not conducive to the development of crime forecasting knowledge because it falls foul through irrationality of the problem of induction, which is the expectation that the future will resemble the past (Popper 1963); an approach that gives us no more than rules of thumb, which are by their inherent nature poor predictors in a social world of unforeseen changes bought by the affairs of people.

Taking forward this criticism of the irrationality of those seeking to explain the past with pet theories and other untestable notions and just how easy it is for us to invent plausible falsehoods to explain things we cannot know, what other explanations could I propose for the fall in acquisitive motor vehicle crime?

I might propose that an increase in security is the most compelling explanation (Farrell et al 2010) or I might propose that for the past 15 years cars have not contained hugely expensive and easily transferable entertainment systems, with a healthy illicit customer base for stolen items, as they did at the time of my earlier research into stolen goods markets (Sutton 1998) where good radio cassette players were retailing at between £800 and £1200 in high street shops; and that portable satellite navigation systems – because they are portable by their owners - have not filled the demand void in theft from motor vehicles. I might, of course, claim that the cause is my Game Substitution theory and that young men have stopped stealing cars, and from cars, for pleasure and peer status because they have found a more convenient and legal indoors equivalent in the form of Grand Theft Auto on their games console. Alternatively, I might make the case that the upward spiralling cost of motor insurance for  those under 25 years has resulted in the eradication of ‘real’ ‘car culture’ for this age group so that stealing and enjoying them is no longer ‘in demand’. Linked to that last explanation might be the fact that the prohibitive cost of car insurance for young men over the age of 25 (well over £1000 per year at the time of writing) makes them likely suspects if seen driving on the roads so that the risk of detection is now much greater than in the past. Linked to that last explanation might be that detection and arrest is enhanced by CCTV vehicle registration recognition systems including fast police response vehicles and helicopters. The list of competing explanations from which I could choose does not end here. I might also hypothesise that those from groups that might have committed motor vehicle crime in the past are now committing less risky more profitable offences instead – such as dealing in cannabis or stealing scrap metal.

Which, if any, of these explanations is the single most important cause of the decline in motor vehicle crime? Does debating them post-hoc in journal articles and book chapters take us closer to the truth about crime or just closer to winning a debate, or agreeing to disagree, on the issue?

What if I wanted to make an even more compelling argument? I might claim that more than one of these causal factors are having a combined influence, or that  one or more those others that Farrell et al (2010) rejected in favour of their own, are operating together with one or more of these to reduce car crime.

In looking back, without enough evidence to even come close to knowing what really happened, my motive in claiming what probably worked to reduce car crime cannot really be to find out and share the truth or uncertainties about the truth. My motive would really be to make a plausible case that is believable by others that my pet theory did it. In effect, I would be guilty, not of lying, but of what the moral philosopher Frankfurt (2005) defines as bullshitting.

Towards a scientific Dynamic Vulnerability theory to enhance knowledge about crime

The dynamic and ever changing affairs of people are a reality that needs to be taken account of by a good theory of crime that can forecast it. As Deutsch (2011: p.26) points out, we are good at inventing falsehoods to explain the past and they are easy to vary once found. By comparison, discovering good explanations is much more difficult because they are hard to find. But the harder they are to find the harder they are to vary once found, which makes them useful explanations for knowledge progression because this allows us to see if they can survive our best attempts to falsify them.

Despite the impact of crime on society, the availability of funding for criminology research is shamefully miniscule compared to that available to the natural sciences. Hence most criminologists tend to be self-limiting by stoically attempting to get on with the limited research that can be afforded or currently known or imagined to be affordable by current funding budgets. But the fact of our inability to predict or explain the crime drop with these realistic compromises questions whether we need to think much bigger if we wish to do better than completely fail to ever come close to providing a better explanation for crime and accurately predicting what will happen next.

Crime as part of an unbounded self-regulating social and economic system

Building on Felson’s (2006) innovative work on lessons for crime from nature my  hypothesis that the  idea of crime and corresponding behaviour of people is having a regulatory impact on the business of people which is impacting on crime in a non- linear way, is an idea that I have very broadly developed from Gaian theory (Lovelock 1989).

Gaia theory, which has had a huge impact on thinking about climate change, is an evolutionary theory that overturns the previous Neo-Darwinist theory that all forms of life on Earth evolved to survive the environment. Lovelock (1989) postulates that life on earth is in fact regulating the environment.

Taking some lessons from Lovelock’s top down approach to theorising about causes in complex systems, might help to develop a good explanatory and predictive theory of why crime rates change, rather than why some targets such as homes, cars, security vans and banks become more or less victimised. Such a theory would need to go beyond parochial and empirically measured appearances such as simple post-hoc explanations of whether or not it was security that improved enough to make a difference to overall crime rates.

To forecast crime with something more than wishful thinking, we need a theory to account for the fact that crime outcomes that are produced by the idea of crime, and the varying and new business of people regarding crime, are likely to be hard to predict unless a theory can take account of them (Pepper 2008). In such a complex social system as ours it is likely that effects will be greater than would be expected from considering the impact of parts of it separately (Lorenz 2000; Lovelock 1989).

To deal with this complexity we need to develop explanatory knowledge of crime that can account for things that are difficult to detect with our senses and existing tools. Such a theory will be ravenously data hungry for all relevant things that are new, and relevant things that vary among individuals and for groups over time from a variety of different causes - such as economic and other cultural ‘demand’ for certain products or behaviour, strength of criminal motivation, numbers of those so motivated and their proximity to potential crime targets, commitment and capacity to offend, imminent drug epidemics, changes in attitudes towards existing potential crime targets and the emergence of new targets and criminal methods. Alongside all these constantly varying and shifting causal influences, we need an up-to-date measurement of how vulnerable targets are to them.

The type of data proposed for a theory that is based on the premise that crime is self-regulating is likely to change as predictive power and different statistical tests are tried out (Deadman 2003). This fine-tuning of the predictive process should not to be confused with the weakness of having a theory that is easy to vary, which is discussed in greater detail later in this article. A theory that crime is self-regulating in our current social systems is easy to disprove by demonstrating that it is not self-regulating at all because, for example, an experiment implementing extraordinary major security and social behaviours changes among potential victims in small local 'target' areas - such as on a geographically and cognitively clearly defined - housing estate along with cracking-down on known current offenders does not lead to crime being significantly displaced to other local or more distant areas (where such hands-on crime reduction measures are not being deployed) and/or victims, times and offence types. And that in the target area, or more distant areas, offenders are not as a consequence 'soon' replaced, joined or competing with more of the same.

Better explanatory knowledge about crime should account for the incidence and prevalence and place of victimization  or - where the crime is a so called ‘victimless’ offence - of incidences and prevalence and place of offending that occurs as a consequence of the dynamic impact of  variation and shifts in the regulatory systems of human affairs. And, because crimes are such diverse social constructs, a theory is needed that can be applied to individual crime types.

With regard to theft, for example, economic, ethnographic, some kind of risk assessment data, and police data would be needed as a minimum and values or weights assigned to three key factors and their theft-relevant characteristics:  public, industry and national demand (the international, national and local demand for particular products, raw materials or behaviour - which includes cultural influences on the demand for deviant behaviour and existing and new law and policy making influences on the demand for conformity to law) offender motivation (strength: determined by short and medium term commitment and actual capacity to offend or not offend, which also takes account of economic conditions, equality, man made and natural disasters, national and civil conflicts, numbers of potential offenders, RAT and cultural influences) and victim protection (degree of vulnerability of target in terms of SCP notions of security, RAT, proximity to potential offenders/proportion of potential offenders in population and efficiencies in offender detection, apprehension and subsequent social processing of offenders - which also takes on board man made and natural disasters, national and civil conflict and cultural influences). Since many offenders are also victims of certain crimes it will be necessary that we understand more about that so we can take account of the fact that people will have simultaneous influences upon them in terms of all three and at other times may experience clear shifts from factor to factor.

As part of our analysis of demand, motivation and protection factors, for crime forecasting, we have to gather rich ethnographic data in order to know what other people are 'knowing'  and believing. To do that we must learn to understand and see what Hayek (1955: p.34) saw:

'It is only by the systematic and patient following of the implications of many people holding certain views that we can understand, and often only learn to see, the unintended and often uncomprehended results of the separate and yet interrelated actions of men in society. That in this effort to reconstruct these different patterns of social relations we must relate the individual's action not to the objective qualities of the persons and things towards which he acts, but that our data must be man in the physical world as they appear to the men whose actions we try to explain, follows from the fact that only what people know or believe can enter as a motive in their conscious action."

Limitation of space in this article prevents me from providing more examples, but it does not take a great deal of imagination to see how the characteristics of these three factors could be modified and considered differently for (possibly) all other crimes ranging from driving above 20-70mph speed limits on motorbikes and sports cars marketed and loved for being capable of doing 0-60 in under 4 seconds, with top speeds of 130 to180, to internet facilitated fraud, trading in abusive images of children and even genocide. Here then is scope for merging the concerns, ideas and understandings of cultural criminology (e.g. Haywood 2004) with methods of ethnography and other scholarship and those of quantitative analysis of crime data.

To unknown, but perhaps not unknowable, varying degrees, Demand, Motivation and Protection are likely to influence one other with varying regularity. If they each could be measured within representative populations at suitably short periods of time - to limit the effects that man will not hold still and is constantly shifting social dynamics (Nicolaus 1969)  - the data could be modelled to explain and predict criminal victimization as an emergent property.

Thinking about crime in this way enables us to imagine a Dynamic Vulnerability (DV) theory based on the principle that crime is a self-regulating social phenomenon.

The general DV hypothesis is that the factors classified under the broad headings demand, motivation and protection are in a state of flux and that their relative variability to one another influences them as a whole. The premise is that this effect is non-linear and so acts to self-regulate crime. Measurable changes in, and therefore between, these three key factors explains crime trends in terms of who does crime where, when, why, how often, to whom to such an extent that it can be forecast because it is not unique and is recurrent enough to be an identifiably predictable trend in the official crime data. Such an easy to verify refutable outcome would go beyond the usefulness of SCP and RAT, which are our best current explanations for why crimes takes place where they do. Because SCP and RAT can only be used with any degree of accuracy to attempt to explain trends after they have happened – including crime on the internet (Yar, 2005). We know this is so because if it was otherwise they would have been used to predict the crime drop and they would be used today to let us know what crime will be doing tomorrow, next month and next year.

The reasons for choosing the three key factors and the reasons for the classification of variables under each is important because an explanation is needed for the thinking, assumptions and premises behind the theory. This issue will be explained at a later date, because the explanation required is too long and detailed for this initial exploratory paper.

DV theory improves upon SCP and RAT in that it moves beyond the simply obvious fact that people commit crimes such as theft, for example, simply because they want something that they can get because it is not adequately protected from them (see: Tilley 2009 p. 121). DV theory seeks to allow us to understand precisely, not vaguely, what is making a target desirable at any point in time, when and at what particular level 'target desirability' influences criminal motivation to overcome existing levels of protection - no matter how low or high those current protection levels might be. DV theory simultaneously aims to take account of the proportion of potential offenders to potential victims. And DV theory, therefore is potentially capable not only of forecasting crime levels but of forecasting displacement as society responds via crime's self-regulating nature to any particular crime regularity. DV theory does this by seeking to assign valid measurable values to demand, motivation and protection factors.

 

What DV theory will need in order to that scientifically (as opposed to the current, predicting the past or else prophesising craft-based guesswork employed by SCP and RAT adherents) is to use its self-regulating explanatory principle to develop powerful techniques of error detection and error correction. In other words, DV theory will need to seek to explain all possible errors in its measurements and interpretation of data. Doing this will enable a more accurate assessment of such things as displacement, which, for example, will improve upon the current unstable situation where false results, and false interpretation of results, has concealed the truth (Phillips 2011). In short this is what makes DV theory Real Crime Science as opposed to the current craft-based rules of thumb, unevidenced guesswork about rational choice, and spurious examples, that that SCP and RAT have always depended upon when considering how best to dismiss the 'threat to theory' issue of displacement (see below). This is no light task however. For example, DV theory will have to develop not only a valid way to assign values to measure the strength of essentially qualitative assessments of cultural influences, it will need also to know how and where to look for and detect errors in those measurements and know how to correct them. And it will have to do the same for every other influence deemed to be operating within demand, motivation and vulnerability explanatory factors. 

The explanatory principle of DV might be hard to falsify by linear mathematics, however, and that is a problem for theory that needs to be addressed. Further, we need to address the issue that an accurate long to medium range crime forecast would have the ability to be a self-defeating prediction if acted upon effectively.

Are ironic unintended consequences discovered in criminology evidence of a self-regulating social and economic system?

A partial understanding of how the 'idea of crime', as a meme, has led to crime creating the environment for a sometimes self-generating and at least always partially self-regulating complex social system can be had from earlier research, which reveals that ironically unintended consequences of crime have been noted in several discreet areas of criminology: such as Durkheim’s (1970) Social Darwinism theory that crime serves the function of  boundary maintenance between socially acceptable and unacceptable behaviours; Wilkins’ (1964) Deviancy Amplification; Tannenbaum (1938) and Lemert's  (1951) Labelling Theory; Young (1971) and Cohen's (1973) concept of Moral Panics; Clark and Weisburd's (1994) Diffusion of Benefits hypothesis; and my own (2011a) discovery of Braced Myths. But, to date, we have not properly considered the idea of crime as part of an unbounded self-regulating social and economic system, the parts of which can be measured and theorised about to better understand the whole. If we had then we might be wary of solving the problem of crime forecasting in light of the example of what might have happened to China's economy if scrap metal exports were banned in order to reduce its theft.

Could DV theory forecast genocide and terrorist attacks?

A DV theory would make the leap to predict before the event that crimes may take place with greater frequency in known risky places or elsewhere against the same or different targets. By way of a contrast, RAT, which is designed specifically as a framework for explanation rather than prediction, explains why, and can be used to roughly predict where, robberies are most likely place in certain areas of the city, but it cannot predict the probability that it will be wrong. DV theory would be able to predict otherwise unknowable, ironic and other unintended consequences of policy making as well as predict what will happen when events change so that we could calculate and forecast an increase in violent attacks happening in broad daylight in previously safe areas – such the 9/11 and 7/7 atrocities and the fatal violence against the Jews in Nazi Germany and the Tutsis and moderate Hutus in Rwanda in 1994. By way of contrast, RAT’s weakness for use as a predictive theory is that it is an easy to vary explanation, which means that it can be used to seek to explain, after the event, why crime events happened due to the convergence of offenders in the presence of ‘suitable’ targets in the absence of capable guardianship (Felson and Boba 2010). RAT does not assign any refutable measurable values to the factors it claims make crime events happen. RAT, therefore, cannot be used to accurately forecast them. This means that we never know with RAT what it is that makes a target suitable enough, an offender ‘likely’ enough and a guardian not capable (or present) enough in the presence of the other two.

Without any measurable predictive scale, RAT’s ‘enough’ values can always be varied by those wishing to use it to explain crime post hoc.

‘Enough’ is too simply not enough!

Exactly the same criticism of RAT as a poor explanation and predictor of crime waves and drops can be applied against the so called ‘Opportunity Approach’ to crime that is deemed by Tilley and Laycock (2002) and  Laycock (2003) to be the  most universal and significant cause of crime. For a fuller discussion of this issue see Sutton (2011a).

 I start my criticism of the ‘opportunity’ concept within SCP and RAT by asking for a definition rather than allowing proponents of these ‘opportunity’ approaches to get away with begging the question of what exactly is opportunity anyway. The Collins dictionary (1979) that I have on my office desk defines it as the loose concept that it is: "(1) a favourable, appropriate or advantageous combination of circumstances (2) a chance or prospect."

Given such generally accepted definitions of opportunity, crimes from theft by finding to the most audaciously planned and executed armed bank robbery, indeed every crime that is committed must have had some kind of opportunity - otherwise how else could any of them have occurred?

Crime as 'opportunity' then must be true, which means it is self-evident in that opportunity - post hoc is always there once a crime has been committed. This makes it a truism and, therefore, a very poor explanation for, or predictor of, crime because it is both easy to vary and impossible to refute (Popper 1963).

Explanationless  theories are bad science (Deutsch 2011: pp.319-320). Because the very concept and meaning of  ‘opportunity’ is not adequately defined or explained by SCP and RAT it cannot be an adequate explanation for crime.  Arguably, a theory with a grossly inadequate explanation is no better at knowledge progression than one with no explanation, which makes SCP and RAT bad science.

Further, opportunity related explanations for crime trends, which are at the core of the SCP and RAT thinking, can be easily varied which means they are not falsifiable. When attempts have been made to test them in a local context through demonstration projects - comprising target hardening attempts - evaluations of impact and displacement they have been implemented by their own proponents, with all the problems of bias that has been shown to arise from such practices (see: e.g. Sutherland 2007; Lehrer 2010). Further, these demonstration projects are at best exceedingly limited with regard to measuring displacement and diffusion of benefits effects not least because they have strangely steered well clear of studying offenders in the field, preferring instead to seek confirmatory evidence in flawed geographically and temporally limited crime data.

Sticking with theft as just one example among many crime types that could be picked to make the point, ‘opportunity’ is used in SCP and RAT as a general broad post-hoc label that is fixed to any target that any offender deems - through a process explained in Rational Choice Theory (Clarke 2005) as worth preying upon because they are saleable enough, portable ‘enough’ and vulnerable enough (Clarke 1999), rather than being explained in terms of a measurable causal factor to be accounted for in predictions of target vulnerability.  Conveniently vague ‘opportunity’ explanations therefore, cannot be tested in advance of observation, experiment or post hoc irrationalisation to seek to forecast, or scientifically determine, whether ‘opportunity’ is a causal factor in theft.

 Farrell et al (2010) do a very thorough scholarly job of explaining the limitations of past attempts to explain the crime drop. They then move on to propose that it was an increase in security – so called ‘opportunity reduction’- to such things as homes and cars that is the most likely explanation for the decline in acquisitive car crime and burglary. But if this were a good hypothesis, Farrell et al could make predictions of past, present and future (Deutsch 2011 p.13) crime trends so that if they are wrong their hypothesis can be contradicted by the outcome of some possible observation or experiment. The fact that this cannot be done with ‘opportunity’ related explanations such as SCP and RAT is a dilemma for those such as myself who believe they are the best explanation we have for crime and understanding national and international crime trends. This then is an issue worthy of in-depth consideration.

My friends and colleagues Farrell et al (2010) write in their paper that their critical work of others stands on the shoulders of giants and, in the spirit of The Enlightenment, they expect others to be critical of them.

I am grateful to Farrell et al for such a clear and scholarly explanation of the limitations of the main theories and hypothesis put forward to explain the crime drop. Nevertheless, it is their explanation to which I now critically turn this essay.

The crime as opportunity/security hypothesis is a poor explanation that cannot forecast crime

The best we can say of ‘opportunity’ is that it is a necessary but insufficient and very poorly defined condition for theft - or indeed for any crime to occur. No wonder proponents of the 'opportunity' or so called ‘security hypothesis,’ such as Farrell et al (2010), did not use it to predict the crime drop. And equally, we should not wonder why those now seeking to predict the past with these too-fuzzy explanations cannot do so with any degree of mathematical confidence of the probability that they might be wrong (Sutton 2011b).

 

Arguments by analogy are fallacies and therefore bad explantions

"Arguments by analogy are fallacies. Almost any analogy between any two things contains some grain of truth, but one cannot tell what that is until one has an independent explanation for what is analogous to what and why." (Deutsch 2011: p371)

 Another self-limiting weakness of ‘opportunity’ approaches to the explanation of crime is that the examples used as evidence that they are good explanations are frequently no more than weirdly parochial analogies that mistake appearance for reality or for regularities of some universal principle or law in an attempt to accumulate confirmatory evidence. But the analogies are fallacies because the explanation is "opportunity," which is not a good explanation at all because it is too vague and easy to vary. By way of example, three all-time favourite examples - that are repeatedly cited by Clarke and others in their support and promotion of SCP are:   

(a)    Clarke’s initial discovery of crime and ‘opportunity’ after discovering that lax regimes led to more absconding from borstals,

(b)    compulsory crash helmet laws reducing motorcycle theft and

(c)    domestic natural gas installation reducing the British suicide rate

Let’s look at these three all-time ‘opportunity reduction’ favourites in turn to see if they stand up to my criticism that they are parochial and have no explanatory reach beyond themselves and things very closely resembling them.

Firstly, escaping from borstals with lax regimes may have the appearance of offenders exploiting some extra ‘opportunity’ to offend that can be used as evidence for some kind of principle to support theories about the wider world. But what young man who is locked up for months or years while in the testosterone fuelled flower of his youth would not constantly screen their imprisoning environment for weakness? In the world outside prisons, things are very different and more complex. In this wider world a far smaller percentage of the population is constantly thinking about, surfing and probing for such weaknesses. And yet there is weakness all around us that most young men choose not to exploit. In our wider world, something more than lax regimes and physical weaknesses are required for people to exploit them criminally.

Secondly, the nationwide switch from coal gas to non-poisonous natural gas in British homes and the consequent fall in the suicide rate is not about people losing an ‘opportunity’ to prey upon others – it is about them losing an immensely convenient, culturally acceptable, effective and popular and painless facilitator in their own demise for which there was no immediately widely known substitute. Coal gas was a tool to be applied by people against their own physical vulnerability – not the vulnerability of others. As a tool rather than target, therefore, it is nothing like a home or the contents of a car, or a person, or any other criminal target.

Thirdly, the introduction of compulsory crash helmet wearing legislation for motorcyclists in the UK did not reduce the physical vulnerability of motorcycles to theft. Instead, it simply made crash helmets an additional facilitator that was necessary for lower risk of apprehension in any taking and driving away motorcycle theft once the thief is riding his two wheeled booty on the open road. This cannot rationally be used, therefore, to support wider arguments for SCP beyond those where the law changes to require thieves to bring along their own crime facilitators.

Both SCP and RAT are self-limiting explanations in that they have very little, if any, reach from behaviour to prediction on the scale required for a crime forecast.

 In addition to the limitations of SCP and RAT discussed so far, another limitation is that they do not adequately seek to explain the wider complexity of reasons for offender motivation in terms of (a) being an outcome of a social/cultural or economic ‘demand’ for their offending and (b) demographic influence on offending sub-cultures of the number of offenders and potential offenders on victimisation levels and on size of potential and actual offender ‘group influence’ effects (see Sutherland 2007) on both rational and irrational offender behaviour and decisions to commit crimes and continue offending. Put more simply, beyond RATs brief common-sense commentaries on group violence (see Felson and Boba 2010), and Rational Choice in general offender decision making, RAT and SCP say nothing about what social psychologists know, from experimentation, about the peculiar effects on behaviour of safety in numbers. One thing we really need to seek to understand more about - alongside the influence of demand for crime - is the degree, if any, of likely influence upon crime and rational and irrational decision making of the number and proportion of young males (aged 15-25 are those most at risk of being offenders and victims) in international, national, regional and local populations.

If the explanation for crime was evident in its appearance as an exploited vulnerability or a happenstance pure opportunity (as in the case of theft by finding) then empiricism would be true and we would have no need for a more scientific method to understand crime. But, to repeat the point already made, as we know since it is such a poor explanation for crime, the so called opportunity/security hypothesis that underpins both SCP and RAT cannot be used to explain or predict general trends in crime before they happen. Consequently, our considerable and valuable knowledge gained about crime through SCP and RAT reasoning, hypothesising and research does not, unfortunately, provide us with the means required to forecast crime.

We might one day have a reliable regional, national and international crime forecast much as we do with the weather. Currently we are not even close. The criminological response to the crime drop ranges from mass indifference, through incomprehension promotion (Young 2004) to scientifically irrational methods of explanation and prediction (Farrell et al 2010). With regard to the latter, it’s as though we are currently working down the rabbit hole in Wonderland where the March Hare announces over tea:

 “For the crime forecast today, tomorrow and the week ahead:  we simply don’t have a clue. But that’s alright since predicting the future is fraught with professional dangers. Never mind then, because the crime forecast for the past 15 years will be sunny, after crime fell due to a pet theory belonging to the Dormouse. Today and tomorrow things will improve further as we predict the past and present with an even better explanation that I prepared later.”

Where does this healthy scepticism leave criminology and a crime science?

In the healthy spirit of scientific criticism that is encouraged by my good friends and colleagues Farrell et al (2010), and all proponents of SCP and RAT I conclude that their 'opportunity' or ‘security’ explanations for crime are poor explanations. Because when crime rates inevitably rise and then fall again, regardless of regression to the mean effects, those who support these crime as ‘opportunity’ ideas can simply shift their explanation - post hoc - with too fuzzy biased rhetoric to seek to account for crime rates rising by saying that enough new opportunities must have been discovered or stumbled upon by enough offenders, old and/or new.  And when crime rates fall again they can easily claim that somehow enough new security now worked enough and – on the basis of biased interpretation of the very little known and largely unknown facts about what offenders actually do when faced with security seeking to deter them from any given crime – that not enough crime has been displaced to pose a threat to theory and future theoretical axe grinding.

Criminologists should not self-limit the development of better explanations for crime on the basis of current paucity of external funding since that amounts to a self-fulfilling prophecy (Merton 1936).  Further, we should not shy away from the search for universal explanations because of the degree of difficulty involved. There will be problems, but the history of mankind’s knowledge development proves that problems are solvable.

The need for a solution to the problem of crime forecasting is neatly put by Pepper (2008):

“…little serious attention has been devoted to crime rate forecasting, and there is no well-developed research program on the problem. Effective forecasts of social processes that evolve over time would seem to require a scientific process that evolves as well. Certainly, periodic efforts to forecast crime or analyse forecasting models cannot hope to provide meaningful guidance.”

I propose a solution to this problem. Namely, the development of Dynamic Vulnerability (DV) hypotheses as both a general and a universal explanation for crime rates that within each specific crime type to which it is fine-tuned and applied, is hard to vary before experimental testing and easy to falsify after. This movement towards a DV theory is rational in that it: (1) does not rely on past observations in the ever changing social affairs of man to predict the future, (2) can allow us to calculate the probability that it is wrong and (3) is easily refutable because it is hard to vary. 

As said, there will be other problems for theory, for which we will need to find a solution or else abandon it when a better explanation comes along. There are two big problems that I can foresee. Firstly, gathering sufficient and necessarily up-to-date representative data of the kind required for a data hungry scientific DV theory would be an enormous and expensive task that may prove prohibitive to initial bids for funding a research study with real data. Further, even though DV is a better idea for understanding crime and, therefore, forecasting crime and crime waves and drops than RAT and the security hypothesis, it may – as an explanatory top-down theory of why crime rates change within an unbounded self-regulating social economic system - prove just as resistant as is Lovelock’s theory of our self-regulating planet to falsification tests based on linear mathematics (McDowall and Loftin, 2005; Lovelock 2006). We should nonetheless build on what we can learn from understanding the limitations of the pioneering work of SCP and RAT to develop a theory, testable by observation and experimentation, in order to better understand and forecast crime. We should, therefore, doggedly seek funding to provide the data that will be required to test its ability to forecast crime. Or should we? Let us think a bit deeper about the data that will be required to test this proposed explanatory theory and how we intend to use and understand it.

Deutsch (2011: p. 313) explains, that in real science we measure a quantity of something that we need to explain, which is testable in that the way we have measured it is likely to reveal its objective value within a known degree of error.  Doing that is likely to prove most difficult with measurements of cultural influences gained from the kind of mass ethnography required by DV theory. Until, that is, we can develop a good testable explanatory theory of what cultural influence on crime is.

Clearly stated, the problem for theory here is that we hypothesize that cultural knowledge influences criminality and our responses to it. The problem we have now is in capturing that knowledge and measuring its quantities and qualities and discovering, or disproving, its contribution to crime.  No wonder then that cultural criminologists and SCP and RAT adherents are at loggerheads (see:  Hayward 2007;  Farrell 2010).

 

The best test of the truth of a theory, even a junk science theory, is to attempt to disprove it with a good scholarly kicking

At this point in this paper I have an important admission to make.

The truth of the matter is that I hate this DV theory. I wish I'd never thought of it. All I really wanted to do this summer was to sit and write a book about stolen goods markets that fits nicely within the explanatory frameworks of SCP and RAT. My research led me to this point, but I don't want to be here.

Marcus Felson (originator of RAT) and Ron Clarke (originator of SCP) have been extremely kind to my work over the years, describing my Market Reduction Approach to theft as important (Felson 1998; 2010; Felson and Boba 2011) and building upon it to take note of offender motivation in SCP (Clarke 1999), all of which was instrumental in the US Department of Justice commissioning me to write an international policing guide for tackling stolen goods markets (Sutton 2010). And yet this article is critical of their work and it is critical of the research of my friends and colleagues Nick Tilley, Graham Farrell and Machi Tseloni. Whatever their attitudes are to the spirit of the Enlightenment and the need to question orthodoxy and all explanations for phenomena, I may always feel that at some level I have been treacherous to them in my own attempts at loyalty to the principles of good scholarship expressed in this article.

Those very principles that I hold dear might be abandoned by me if I were to fall in love with my DV theory of crime and make it my pet. Being blind to failings may be the best romantic strategy in courtship (Pinker 1999) but it is possibly the worst for knowledge progression. Falling in love with my own theory would affect my judgement so that I too might be inclined to spend most of my time searching for and writing about confirmatory examples of it with reference to such things, for example, as the failure of mankind to reduce prostitution and trafficking in people. I could seek to support it further with the unintended crime consequences of alcohol prohibition in the 1930s in the USA. I could cite also the more recent history of the failure of anti-drugs and liberal drugs policies and I could cite the many examples where crack-downs on criminal networks merely results in new ones stepping in to fill the vacuum and I might use the example of China as a country with a known history of extreme intolerance to internal offending and deviance as the world’s largest exporter of fake medicine, alcohol, luxury goods, mechanical service components, music and software and the largest importer of stolen metal.

But I'm not going to let that happen. Instead, I'm going to nurture an allergy to this theory since owning pet theories is harmful to the healthy scepticism that is essential for knowledge progression. Unlike good explanations, a poor pet theory is analougous to those virutal pets know as Tamagotchi, which soon die simply because their owner neglects to feed them. 

So as not to become a tamagotchist, I need something help me to keep my theory at arm’s length and so I'm going to give it a ludicrous name. I'm going to rename DV theory as Robin Hood theory to replace the more impressive sounding Dynamic Vulnerability. I hope that doing this will encourage me to do my best to give it a good scholarly kicking, and at other times to neglect to nurture it, not least because I don't want to make a fool of myself creating and supporting a ludicrous sounding theory that is also wrong.

Besides the fact that it is a deliberately ludicrous name, the notion of Robin Hood fits the idea that crime is a self-regulating system. According to the fictional legend, Robin Hood became a problematic outlaw as an ironic unintended consequence of his initial criminalisation. According to legend, he redistributed goods to those who had a demand for them because they were in short supply. According to the stories written about him, his general modus operandi was to prey upon targets that were relatively easy to overcome. When faced by measures to defeat him, Robin Hood never stole less overall or went into early retirement. Robin Hood, our Prince of Thieves  also had a band of merry men with a number of other characters who could step in and fill the void in the event of his capture or eventual retirement. Robin Hood, like the 'idea of criminals' today is glamorised in books and films, was glamorised in the medium of song and later in plays, books, and film. And research suggests that Robin Hood is a name that derives from what was merely a general term applied to problematic criminals in England around the 13th Century (DeVille, O. 1999). Confirmatory evidence to support the theory can be found in the fact that the rap group Gang Starr propose their own unchallenging Robin Hood Theory in verse and music. 

If the theory that my research led me to propose as an excercise in real crime science for this article is one that turns out to be worthy of developing and testing stands up, after our best attempts to disprove it, then there might just be something in it. If it does stand up, then I'm still going to keep trying to disprove it because that's what good scholarship involves.

Who knows, I really doubt it will, but at some point Robin Hood theory could turn out to be the best explanation we have for crime, even before it is tested as a means to forecast crime. But such an outcome would never mean it is right and will never mean it is the best explanation for crime that we will ever have. And hopefully its ludicrous name will lead others to try to disprove and surpass it rather than nuture it as a pet theory.

The first test of Robin Hood theory's hard to vary explanatory power is to see what offenders actually do when faced with police crackdowns and SCP security measures. To repeat the point made earlier, the Robin Hood theory that crime in western industrialised societies is self-regulating demand, motivation and protection can be refuted if it can be shown that an extraordinary increase in protection in an area of high criminal predation significantly reduces crime there and then and does not result in any kind of significant displacement of crime along the lines described by Reppetto (1976) anf Hakim and Rengert (1981) and does not result in a rapid or medium-term return to it becoming a high crime area once again.

In short: crime cannot be self-regulating in the face of all our past and current efforts to prevent and reduce it if extra security measures simply make prolific offenders reduce or stop their offending.

A proper test of this, that goes further than previously inadequate tests, would be to conduct an ethnographic research study, triangulated with extensive, comprehensive official and independent crime data to determine what offenders really do when faced with an increase in security measures and a police crackdown on their offending in an area they are extensively preying upon.

Some ideas for useful research

Further useful research of the kind that criminologists and their funding bodies are more accustomed to dealing with might be conducted in the short term while Robin Hood theory is being developed, tested to destruction, and funding to test it is being sought.

For example, with regard to my RAT Game Substitution hypothesis for the role of the internet in the crime drop; rather than fall foul of the problem of induction, we need to gather data of the kind necessary to conduct rational research. Even in this relatively less ambitious area we need to know much more about a number of things about which we currently know next to nothing:

• How much time is spent video gaming, online and watching TV by groups that research predicts are at greater risk of becoming offenders/victims – compared with other groups?
• We need ethnographic information to help us decide whether, and if so to what extent, video gaming and other media is used as a substitute for offending and/or acts as a deterrent for risky activities in the offline (real) world. This should be considered in relation to both potential offending and victimisation.
• Examine how peer status might be achieved offline and online by different groups and whether video gaming serves as a substitute for deviant ‘edgework’ offline.
• Conduct research in the field to test two hypotheses: (1) The Rational Choice Hypothesis that offenders weigh risks and rewards (2) The Irrational Choice Hypothesis that offenders choose between competing irrational choices and then self-justify their decision making.

• Examine US crime records to establish whether the Beatles did cause an otherwise unexplained crime drop during the 1 hour they played live on TV in the 1960s.