Safe cities

The reduction of crime is not primarily to be expected from effective police work, despite the best technological support, but from a society with more equality and sufficient income, adequate housing and good education for all.

A pleasant and safe environment to live. For many people this desire is at the top of their wish list. Safety has many aspects, from protection against natural hazards and epidemic diseases, to safety in traffic, fire protection and safe internet. All these forms of safety are covered in one of the essays in this series. Below, the emphasis is on protection against crime: violence (including sexual assaults), robberies, burglaries, blackmail and theft. It is also a subject where ‘smart solutions’ are often mistakenly seen as a solution. 


Safe Cities is the fifth of a series of short essays on how cities can become more humane. That means finding a balance between sustainability, social justice and quality of life. This requires far-reaching choices. Once these choices have been made, it goes without saying that we use smart technologies to realize them.

The essays already published can be found here.


According to The World Health Organization, interpersonal violence is a leading cause of impaired quality of life and mortality in the world, especially among people between 15–44 years. In 2000, homicides resulted in half a million deaths worldwide; almost twice as many as in wars during the same year.

Data from the European Union (EU) show moderate reduction in crime. Car thefts fell by 36 % between 2008 and 2016, robberies fell by 24 %. Both trends flattened after 2010. The number of intentional homicides in de EU was about 5200. Overall, there has been a downward trend since 2008. Yet, police-recorded sexual violence in the EU shows a steady rise of 26 % between 2013 and 2016.

Causes of criminality

Before discussing solutions, I look at what research has to say about the causes of criminality.

It is important to realize that factors that are seen as the causes of criminal behavior do not invariably result in it. Instead, they increase its risk. A distinction must be made between proximate causes, which immediately precede criminal behavior and factors that are distal in a varying degree, though not necessarily less influential or important.

The table below summarizes the essence of research publications from recent years, ignoring details.

Crime is related with inequality, poverty, poor housing, unemployment, and use of alcohol and drugs. These are therefore the most important characteristics of neighborhoods with a lot of crime. For residents of these neighborhoods, these circumstances impede a decent life. ‘Economic stress’ influences the quality of the parent-child relation. Poor parent-child attachment, insufficient parental supervision also with regards to alcohol and drugs and inconsistent discipline all increase the risk of involvement in crime of young people and lower the prospect of a successful career at school and elsewhere.

In neighborhoods where these problems are commonplace, there is also a lack of informal social control by local residents and young people seek refuge in gangs and other criminal organizations that further increase crime, both in their own neighborhood and elsewhere.

According to an Australian study, most criminal acts are committed by persons younger than 20. 29% of them has been involved in criminal activities at least once. However, 70% of all juveniles who appear in a children’s court do not return there. Most crimes are committed by a relatively small group. A study from Sweden, which analyses crime in the period 1973 – 2004 revealed that 3,9 % of the population were convicted for a total of 93.643 violent crimes. However, 25% of this group was responsible for 63% of all violent crime.

Studies of the development of the violent crime numbers in the past decades make it possible to gain a deeper insight into the distal causes of crime. In the second half of 20th century, the crime in the US increases year by year. In the nineties, however, a decline began and, crime is now at historic lows today in many places. In New York City, for example, there were more than 2,200 homicides in 1990. In recent years, there have been fewer than 300 a year.

Levitt investigated every possible reason for this decline and in 2001 he and Donohue he concluded in a scientific publication that the overwhelming explanatory factor was nothing else then the legalization of abortion. At its peak in 1990, there were 1.5 million abortions in the US, compared to 4 million live births. They related this to the extensive literature that gives evidence that unwanted children are disproportionately likely to engage in criminal behaviors, mainly because of the already mentioned weak parent-child attachment. And it is exactly this group of children that is radically diminished due to the legalization of abortion.

In a recent update of their earlier publication Levitt and Donohue confirm the strength of the relationship between abortion and crime for the period between 1997 and 2014: In states with the highest abortion rates, crime rates were 60 percent lower than in states with the lowest percentages.

It is importance to warn against a mono-causal reasoning about the causes of crime. Economic stressors such as poverty, unemployment and debt can lead to crime, for instance because they weaken the parent-child attachment. However, most poor people follow the law and care for their children. Attachment problems in parent-child relations also occur in well-to-do families. It is the frequency and the cumulation of stressors that counts.

Awareness of the economic causes of crime is growing worldwide but it is astonishing that the – albeit complex – relation between economic stressors and criminal behavior does not play a central role in policy. In essence, the ultimate requirements for reducing crime and improving safety are: Sufficient income, adequate housing, fulltime and affordable day-care services, especially for ‘broken families’ and unmarried mothers, and ample opportunities for education of girls.

Care for juveniles who have committed their first crime also is of utmost importance, because a humane approach will prevent most of them making a mistake again. At the same time, society must be protected against the relatively small group of repeat offenders, who are responsible for most crimes, in particular violent crimes.

Social capital

The more intensively residents of a neighborhood interact and have a look at each other’s belongings, the less crime gets a chance. It is even better if they also keep an eye on the activities of the youth. Social control has always been a mighty weapon against crime. To create a safe living environment, contacts do not have to be limited to neighbors.  A study among the 12 Dutch ‘safest’ municipalities – according to the National Safety Monitor – has shown that close cooperation between inhabitants and the police at neighborhood level contributes to safety and to feeling safe. Other conditions are the visual presence of police (preferably at a cycle instead of in a car), detailed knowledge of and communication with youth groups that occasionally cause problems, offering shelter to homeless people to prevent drugs-related crimes and high priority to combatting violence and burglary.

Assessment of security with regard to criminal behavior of 60 world cities: The Economist

A neighborhood-oriented approach does not depend on the size of the city. For this reason, Amsterdam, the most unsafe city in the Netherlands, is one of the safest cities in the world. For cities in Latin America, amid a wave of crime and violence, the price of this disruption is high. A recent study shows that crime and violence in Latin America costs on average of 3% of GDP per year, which is around $261 billion for the region. Cities in Latin America and Asia invest large sums in technological equipment to fight crime. Medellin, the former capital of crime achieved significant improvements by tackling the roots of crime: Poverty, drug trafficking and social contacts. Intervention programs that strengthen social capital take place elsewhere.


APPS Program (Applications for Purpose, Pride and Success) in Columbus

The mission of the APPS program is to reduce crime and violence by creating protective conditions in the lives of young people in Columbus aged 14-23. The initiative’s prevention strategy consists of street-level violence interruption and conflict mediation by trained intervention workers, who are building on relationships with high-risk youth. In addition, the APPS program offers educational, recreational, skill training and arts activities in four community centers.


The role of the police

The fight against crime in our society is a primarily a battle against criminals with the use of traditional and advanced technological means. In general, the police are considered the main player, almost automatically on the right side. However, there is a growing risk – at least in some countries, including the United States – that the police themselves becomes a disproportionate source of violence and an organization that suppresses parts of the population. For many colored citizens in the US, the police do not represent ‘the good guys’, but it has become part of a hostile state power.  Below, I pay attention to the police in the US, because developments in that country are copied elsewhere. In many countries, including the Netherlands, the police are a much more integrated part of society, albeit its masculine and authoritarian culture prevents smooth integration of representatives of minorities and women.

Historically, police are the protector of state power in most countries and are equipped to suppress resistance. Early police organizations in the US wore the same blue uniforms as the former slave patrols. In most countries, the police are organized on the model of a military hierarchy. To be part of the police, officers must obey orders in the first place, which is detrimental to the development of a personal moral agency.

In 2017, police action in the US led to an unprecedented number of 1100 of deadly victims, mostly colored people. Moreover, the police have been involved in racial profiling for decades. Between 2004-2012, the New York police checked more than 4.4 million residents. The lion’s share did not result in further action. In about 83% of cases, the person stopped was black or Latino, although the two groups together account for just over half the population.

Artwork Nafis White: “It Doesn’t Show Signs of Stopping”

Rules of conduct to reduce the number of fatalities of police operations

Scientist have established eight rules of conduct that can stop the number of deaths by the police and more in general contribute to a more positive image.

  • Require officers to de-escalate situations before resorting to force.
  • Limit the kinds of force that can be used to respond to specific forms of resistance.
  • Restrict chokeholds.
  • Require officers to give verbal warning before using force.
  • Prohibit officers from shooting at moving vehicles.
  • Require officers to exhaust all alternatives to deadly force.
  • Require officers to stop colleagues from exercising excessive force.
  • Require comprehensive reporting on use of force.

The researchers wanted to know in which degree these rules where implemented in the 90 police districts that they studied. It turned out that no department had implemented all. De-escalating situations or exhausting alternatives before resorting to deadly force were required, respectively, in only 34 and 31 departments. Only 30 departments required officers to prevent a colleague from exercising excessive force, and only 15 departments required officers to report on all uses of force, including threatening civilians with a firearm. Researchers calculated on average that the adoption of all eight requirements could lead to a 72% reduction in killings.

Prisons

The US justice system is under great scrutiny and under pressure to reform thanks to initiatives such as the Art for Justice Fund, the Open Society Foundations, and many reports. However, one aspect is largely overlooked: The design of prisons. The American system is harsh and brutal. It is not surprising that most people leave a prison as a criminal for life.

Since 2016, none other than star architect Frank Gehry ( Guggenheim Bilbao!) has focused on this topic. The efforts of Gehry, together with his students, are illustrated in the documentary Frank Gehry: Building Justice. What if we start treating people like human beings—what would prison look like? asks Gehry in the documentary. Part of this documentary can be viewed here.


In the documentary we see how Gehry’s students visit Scandinavian prisons, which offer a better correctional model based on healing instead of hardship.

Impact of the police academy

Susan Rahr’s swapped her role of sheriff of King County, Washington, with that of head of a regional police academy. She had been annoyed for years by the ease and eagerness of the use of violence by police officers and assumed that the police academy could make a difference. When she entered the academy, her first impression was the dominance of a militaristic culture. Compliance was based on the threat rather than on ethical principles.

Her intention, which attracted national attention was to implement the concept of police as guardians: complete police officers with a wide range of skills, who are wise and humane at the same time. To achieve this goal, she added exercises to the program to deal with tense situations without resorting to coercion, striving to solve a problem without making an arrest, and seeing interactions with police through the eyes of a victim.

In an interesting interview with five representatives of groups who feel victim of the US legal system, no one disputes the role of the police as just. But all participants argue in favor of civil oversight, meaning that municipalities decide what a safe society is and with which methods the law is enforced.

The role of technology

Given this background, it is not easy to share the enthusiasm of ‘smart city’ adepts with the benefits of technology to combat crime. Technology will not erase crime and criminal organizations will ultimately surpass the technological expertise of the police. It also does not solve problems caused by poverty, disintegrated families, gender-related and other violence, and addiction that cause crime. However, technology might be an additional tool within a humane approach, given its reliability and the integrity of those who deploy it. Below, is a brief overview of current trends in technology deployment. Most examples come from the Smart City Solution Database, an extensive collection of smart city applications, tools and policy and a report of the McKinsey Global Institute, Smart Cities: Digital solutions for a more livable future

Solving crimes

The speed with which the police arrive at the scene of a crime influences the chance of its resolution.  Sometimes offenders can still be found or their tracks are still fresh. For this reason, cities equip crime-prone streets with gunshot detectors.


Shot spotter

To detect gunshot activity, approximately 30 – 40 acoustic sensors per square kilometer are required. These acoustic sensors record noise, location and time. When a weapon is fired, audio triangulation locates the correct location and machine learning algorithms analyze the sound. There are only 45 seconds between the moment the shots are fired and the alarm goes off at the nearest police station. The short video below shows how the shot spotter works.


There are other tools that support criminal investigation. Each crime leaves a footprint, sometimes literally, and technology like the Criminal Finder helps to find the owners of that footprint.


Criminal Finder

The Criminal Finder is an artificial Intelligent solution capable of finding criminals by looking at an alarm or a first information report. It takes data related to crimes done by different criminals and organized gangs as input, looks for patterns with machine learning. When a new crime is reported, it can be fed into this trained model to find out if the crime might have been done by a known criminal.


Crime prevention

Crime prevention is the ultimate goal of police activity, albeit this goal is far beyond reach of the police alone and without dealing with the roots of crime. Nevertheless, nobody will deny that preventing crime opportunities is worth pursuing. A rather elementary example of crime prevention is mapping data on the times, locations and nature of past crimes to provide insight where, and at what times, police patrols should patrol, in order to optimize the use of resources. Sometimes, these data are communicated with the public. The mobile app CrimeRadar discloses real-time danger zones in Rio de Janeiro that the public can better avoid.


CrimeRadar

CrimeRadar uses algorithms to predict crime patterns based on crime records. The application estimates the likelihood of future crime, based on a mathematical assessment of the location, timing, and characteristics of millions of specific crime events (i.e. homicide, assault, kidnapping, robbery, rape) from January 2010 to April 2016. After this date the underlying dataset is not being updated, for unknown reasons. Citizens who use this app know when they better avoid certain places.


Predictive policing

Predictive policing includes mathematical methods and other analytical techniques to identify potential criminal activities.It includes methods for predicting crimes, offenders, and victims of crime.

The New York police deploys an advanced specimen of predictive policing that operates a so-called criminal group database. One percent(!) of the people in the database are white, 66% are black, and 31.7% are Latino. With this system the police can immediately obtain detailed information being followed or arrested. The system is connected to 9,000 video cameras in New York City.It also has access to data from 2 billion license plate readings, 54 million emergency calls, 12 million detective reports, 11 million arrests and 2 million warrants.

A machine learning algorithm known as Patternizr connects potential criminal suspects with unsolved crimes to speed arrests and close old cases.  The algorithm is trained on a decade of historic police data from manually identified crime patterns. Patternizr was commissioned in 2017 with the help of one hundred civilian analysts. The initial development costs were $350 million. Many other cities implement similar systems, for instance York and Glasgow. A fully operational system, albeit technologically less advanced can be found in Moscow.


The Moscow

A machine learning algorithm known as Patternizr connects potential criminal suspects with unsolved crimes to speed arrests and close old cases.  The algorithm is trained on a decade of historic police data from manually identified crime patterns. Patternizr was commissioned in 2017 with the help of one hundred civilian analysts. The initial development costs were $350 million.

Video surveillance system

The system – which costs $ 250 million- uses 128,000 (!) video surveillance cameras in public areas, courtyards, entrances, schools and a further 9,000 cameras in transport hubs, the metro, and cultural and sport and social facilities. It can identify faces and supports a wide array of policy duties, such as automatically producing fines, recognizing whether seat belts are fastened or not and if a person is talking on a mobile phone. Currently, the cameras are used to check if the trash is collected and they register and red-light running and driving too fast.  Roughly 75,000 fines are imposed per day.


These advanced systems are operating too short to evaluate their efficacy. Police officers themselves report a significant drop in criminal activities. However, a study from the Max Planck Institute for Foreign and International Criminal Law after three years concludes not to be able to make definite statements about the efficacy of the software. The 3-year pilot project has entered a second phase in 2018.

The need of civilian oversight

In the US sepsis reigns over the integrity of the technological system, and a number of cities, like Seattle, Oakland, Berkeley and Davis have mandated civilian oversight of new forms of police surveillance. They all have barred municipal police from deploying new surveillance technology without approval from the local government.

In New York, a bill to regulate the use of machine learning and artificial intelligence, the Public Oversight of Surveillance Technology Act (POST) was first proposed in 2017. Unlike the ordinances in San Francisco and Oakland, the New York City bill requires only that the police disclose basic information about the technology that is being deployed. It does not condition the use of new surveillance technology on the approval of the City Council. Until now the bill still is under discussion.

The Legal Defense and Educational Fund, a prominent civil rights organization in the US, urged the city administration to ban the use of data derived from discriminatory or biased enforcement policies.

While the deliberations in New York are continuing, San Francisco went a few steps further, and became the first city to ban the use of facial recognition technology by the police and other public agencies. The city council agreed that facial recognition once might become a valuable tool. However, at this time it cannot be ensured that facial recognition will be used responsibly and without discriminatory effects. Simply for the fact that the technology is not ready for it: Experts acknowledge that the artificial intelligence underlying facial recognition systems still struggles with accuracy, particularly when it comes to identifying the faces of people of color — who are the people most likely to be affected by it.


Rekognition: Amazon’s face surveillance technology

In a test last year by the ACLU, Rekognition, Amazon’s facial recognition software falsely matched the faces of 28 members of U.S. Congress to the mug shots of people who had been arrested. The mismatches disproportionately affected representatives of color. The Georgetown’s Center on Privacy and Technology concludes that since 2016, the world has seen such a dramatic range of abuse and bias in facial recognition that local, state, and federal governments should place a moratorium on the technology.


Private domain

While resistance to facial recognition based on algorithms and machine learning is growing rapidly in the public domain, the commercial applications of the technology are booming. This technology, for example, promises an alternative to failing police intervention against shoplifters.


Facewatch

Facewatch is a facial recognition company in the United Kingdom. The cloud-based security system for face recognition protects businesses against crime. The system sends immediate alerts when persons involved in shoplifting and other criminal activities enter the premises.


Aside from reliability, these systems put the clerk in an impossible position. it is asking too much that a store clerk refuses a customer based on a warning from the system. The police do not intervene in time, and physical resistance is a bad idea. In a department store, safety officers can stalk suspects in an obtrusive way until they leave the premises to find opportunities elsewhere.

Safety in the humane city

The decrease of crime is related to improving the quality of life of the poorest part of society instead of allowing inequality to grow, as is happening in the United States. This is necessarily accompanied by the improvement of educational opportunities for the youngest children, the creation of a respectful and challenging environment in schools and the prevention of gender inequality.

It would be naïve to believe that less inequality and improved income, jobs and housing for the poorest groups will completely eliminate crime. Greed, sensation seeking, boredom, wrong connections, imitation, drugs and mental disorder are not reserved for poor people. Besides, rich people involved in activities on the edge of the law and beyond are surrounded by lawyers and powerful relations to offer effective protection.

Not only education at schools, especially if it starts at a very early age, can partially compensate for problems at home, but the quality of social life in neighborhoods, including a certain level of social control does the same. In all these cases, the visible presence of district police officers without unnecessary display of power supports the co-operation between police and civilians. This is necessary to prevent young people from making mistakes again after a first criminal activity.

The necessity of a highly professional and technological equipped police is beyond question. No technology, if proven effective and non-discriminatory, has to be excluded in advance. The path outlined above is long, but it deviates radically from what is currently happening around the world. I am afraid that the use of technology in a militaristic and authoritarian police organization estranges the police from citizens.

Finally, I summarize the characteristics of a humane approach to safety in our cities, considering the relationship between crime and poverty, living conditions, lack of education and disrupted family circumstances.


Actions for a humane approach to safe cities

  • Availability of jobs, sufficient income and good living conditions for all adult members of society, considering individual preferences.
  • The provision of high-quality compulsory education for all children aged two years and over, thereby strengthening their intellectual, social and creative development and resilience. Opportunities for pre-school care for children of three months and older.
  • Developing of a versatile living environment in neighborhoods, offering ample opportunities for play and social activities for all in residents, thereby creating a certain level of social control.
  • Non-aggressive visibility of police in neighborhoods, to support the quality of life and with a trained view of emergent deviant behavior of children (and adults) in collaboration with relevant institutions.
  • Visibility of police in the whole city supported by proven technological and non-discriminatory tools.
  • The use of technological tools by the police and the judiciary, including the use of algorithms and machine-learning, is supervised by community representatives. They are supported by experts who advise on the efficacy of deployed and future equipment and the (unwanted) side-effects.
  • White-collar crime is being tackled vigorously, partly to prevent it from becoming an alibi for other people to break the law.
  • Prisons offer humane living conditions, even if protecting society for recidivism necessitates a long period of detention.
  • Society invests in in the prevention of (young) criminals becoming repeat offenders, which goes far beyond than just providing community service orders.
  • Those who cause damage to public and private property are always obliged to compensate it. Reimbursement takes place in affordable monthly installments and the debt can be reduced in case of persistent good behavior.
  • The organization of the police changes from a military to a modern public organization that takes into account the need for personal moral agency and responsibility and the need to coordinate actions effectively.






  • Header: Photo: The Ascent: A community of storytellers documenting the journey to happiness and fulfilling
  • This article was brought to you by Professor Herman van den Bosch, Professor at Open University of The Netherlands.

Leave a Reply

Your email address will not be published. Required fields are marked *