Discussion 5_Week 5_Secondary Prevention_Delinquency and Crime Prevention

Discussion 5

Secondary Prevention

A. Predicting Future Offending

Making predictions about future behavior, whether deviant or conventional,

involves making a number of initial decisions. Perhaps the first issue is the determination

of what is being predicted. In criminal justice, predicting recidivism is perhaps the most

common endeavor. Impact evaluations of interventions, whether punitive or

rehabilitative, typically look at various measures of subsequent offending. Rearrest,

reconviction, reincarceration, seriousness of future activity, and revocation of probation

or parole are common measures of recidivism. Prediction of recidivistic activity,

however, does not address the central concern of secondary prevention, which would be

predictions of initial deviant acts by individuals. This is often expressed in terms of

predicting potential dangerousness. Potential dangerousness is an important consideration

in the activity of criminal justice professionals and nowhere is this more evident than in

the juvenile justice system, which is premised upon potential future involvement in adult

criminal activity. Secondary prevention hopes to keep the potential offender from ever

realizing that potential.

A final important consideration in predicting future behavior is the degree of

accuracy in the predictions. When we make a prediction and it proves to be accurate, we

are not concerned. On the other hand, making wrong predictions can have dire

consequences. Error in prediction takes two forms—false positive and false negative

predictions. Each of these has a different impact on the individual being evaluated and/or

society. In terms of criminal/deviant behavior, false positive predictions are those in

which an individual is predicted to do something in the future (e.g., recidivate, offend, act

dangerously) but is not found to act in that fashion after follow-up. Conversely, false

negative predictions declare that the person is not a future threat but the individual does

engage in the negative behavior at a later time.

Prediction generally falls into one of two categories—clinical prediction and

actuarial prediction. Clinical predictions have predominated in criminal justice,

particularly in terms of sentencing and treating individuals. Actuarial prediction has been

a more recent choice given the availability of large amounts of data and problems with

clinical techniques. Each of these is discussed below. Clinical predictions are based on a

rater’s evaluation of an individual, usually after interviews and direct examination of the

subject and his records. The training and disposition of the individual rater often

determines what variables and factors are important in arriving at a decision. The rater

can use various psychological tests, demographic information about the individual,

family and individual background information, or interviews of the subject in making a

determination. There are no firm rules for which items must be used, when they should

be used, which are the most important, or whether more than one type of information is

used. In most cases, the individual rater has total discretion.

Actuarial prediction refers to making predictions based on known parameters in

the data. The best example of actuarial prediction is the setting of rates by the insurance

industry. The cost of life insurance is based on the known mortality rates for the

population group to which the applicant belongs. Males have shorter life spans, on

average, than females. This results in shorter periods in which to pay into the insurance

account. In turn, the premiums for insurance are slightly higher for males. The prediction

is based solely on known, statistical factors. Similarly, car insurance rates are determined

by past accident levels and claims. Young males are involved in more accidents, which

leads to higher insurance premiums for all young males. The key to actuarial prediction is

the identification of the appropriate predictive items. Factors typically used in criminal

justice include age, race, sex, socioeconomic status, educational status, IQ, criminal

history, the immediate offense, family background, and psychological test results. As in

the clinical studies, actuarial evaluations greatly vary in their choice of items and

techniques.

A wide variety of techniques have been used to construct actuarial prediction

scales. The methods vary from the simple additive procedures of Burgess (1928) to more

sophisticated techniques, including multiple regression, predictive attribute analysis, and

association analysis. The results of these different approaches, however, do not uncover

any single best method (Farrington, 1985; Gottfredson and Gottfredson, 1985; Wilbanks,

1985). Similar results emerge when using each and there is often little predictive power

in any of the methods. Wilbanks (1985) reports between 25 and 33 percent false

predictions for different methods. Similarly, Farrington (1985) finds that an average of 45

to 50 percent of the predictions are false positive predictions, regardless of the method

used, with false negative predictions comprising about 10 to 15 percent of the predictors.

B. Risk Factors and Prediction

A more recent trend in identifying who will commit offenses (i.e. prediction)

involves the identification of risk factors related to deviant behavior. Most discussions of

risk factors do not make the assumption that individuals who exhibit these traits will

inevitably become criminal or act in some inappropriate fashion. Rather, the risk factors

are indicators of who may become deviant in the future. There is usually no attempt to

make predictions about specific individuals. The risk factors are only indicators or flags

that can signal the need for increased attention or possible assistance for individuals.

The identification of potential risk factors is not a new idea. Indeed, the idea of

working with youths at risk of becoming delinquents or later adult criminals is a

cornerstone of the juvenile justice system. The very premise of the juvenile court is to

work with troubled at-risk youths. Implicit in this task is the idea of identifying the

factors underlying the juvenile’s behavior and working to alter those conditions.

Likewise, most criminological theory is based on the idea of identifying the best

predictors of criminal activity in order to develop appropriate interventions. While not

typically referred to as risk factors, the variables that are found to be related to

delinquency and criminality are risk factors. A good deal of research is on identifying

risk factors. Risk factors can be broken down into various categories. Typical groupings

found in the literature are family, peer, community, psychological/personality, and

biological risk factors (see Table 10.3). While different authors may classify individual

risk factors slightly differently, these categories are generally representative of those used

in the literature. The information presented below is not intended as an exhaustive list of

risk factors. Rather, the intent is to offer some insight into some of the more recognized

variables that have been discussed in the literature.

A wide range of family situations and factors influence both the immediate care

of an individual as well as later behavior. As noted in Chapter 8, poor parental

supervision and inconsistent and harsh discipline are key early risk factors for later

deviance. In one early study, McCord (1979) found a strong relationship between

youthful offending and the type of parental discipline and supervision. Others (Capaldi

and Patterson, 1996; Hawkins et al., 1995; Loeber and Stouthamer-Loeber, 1986; Wells

and Rankin, 1988) report that violence and aggression are more prevalent for youths from

homes exercising harsh and inconsistent discipline. One factor often promoted in the

literature is the relation between parental criminality and the behavior of the offspring.

Farrington (1989) claims that having a parent arrested is related to later offending by

male offspring. This relationship appears in a variety of other studies.

The influence of peers is generally viewed as one of the most important factors

involved in adolescent behavior (Elliott, 1994; Elliott and Menard, 1996; Lipsey and

Derzon, 1998; Thornberry et al., 1995). Of particular importance to adolescence is the

presence and/or participation in youth gangs. Participation in gangs is related to higher

levels of offending (Esbensen and Huizinga, 1993; Thornberry et al., 1993; Thornberry,

1998) as well as initiation into deviant activity (Elliott and Menard, 1996). Data from the

Cambridge Youth Study suggest that the antisocial behavior of siblings is also a potential

predictor of delinquent activity.

Another potential source of risk is the community within which an individual is

raised and resides. Studies of community influences are long-standing in criminology

(see Shaw and McKay, 1942). Many analyses look to community problems, such as

economic deprivation, disorder/incivility, poor neighborhood integration, and similar

factors, as contributors to individual deviance (see, for example, Bursik and Grasmick,

1993; Sampson and Lauritsen, 1994; Shannon, 1991; Skogan, 1990). Other community

influences related to levels of delinquency and criminality include the availability of

firearms (Block and Block, 1993; Lizotte et al., 1994; Sheley and Wright, 1995), low

socioeconomic status (Elliott et al., 1989; Farrington, 1989; Lipsey and Derzon, 1998;

Smith and Jarjoura, 1988), and level of gang activity. Each of these community factors

contributes to the level of risk.

Another set of community factors that contribute to risk involves schools. While

schools are a part of the community, they may contribute uniquely to risk. Poor academic

performance and school failure are common factors related to current and later deviance.

Hirschi (1969) points out that schools and academic participation are key factors in

delinquent behavior. Various studies show that dropping out of school and low academic

achievement are strong correlates of delinquency (Farrington, 1989; Gold and Mann,

1972; Maguin and Loeber, 1996; Thornberry et al., 1985; West and Farrington, 1973).

While a relationship between school factors and delinquency clearly exists, the exact

causal mechanism between the various factors is not as easily identified.

An array of psychological and personality variables have been identified as risk

factors for aggressive behavior. Brennan et al. (1993) note that hyperactivity among

preteens is significantly related to violent behavior during the young adult years. Similar

results appear in research comparing hyperactive boys with their non-hyperactive siblings

(Loney et al., 1983). Impulsivity and problems with concentration also are related to

higher levels of adolescent deviance (Farrington, 1989; Farrington and Loeber, 1998;

Loney et al., 1983). Other factors often related to deviant behavior include learning

disabilities, low IQ, and similar issues that may inhibit an individual’s success in school

and elsewhere.

Biological risk factors are identified in a number of studies, although they do not

receive the same degree of attention as other factors. The major reason for this state of

affairs is the fact that criminology is dominated by social scientists who focus their

attention on other variables. Among the possible biologically based risk factors are

prenatal and perinatal complications (Brewer et al., 1995; Farrington, 1996; Kandel and

Mednick, 1991; Reiss and Roth, 1993). Included here are low birth weights,

complications with pregnancy, drug use while pregnant, and poor nutrition.

Neurotransmitters, such as serotonin, are other biological risk factors possibly related to

deviant behavior (Moffitt et al., 1997). Neurotransmitters are bodily chemicals that

transmit messages in the brain.

Many researchers attempt to use risk factors as predictors of later deviance.

Lipsey and Derzon (1998) conducted a meta-analysis to specify risk factors from two

different age groups—ages six to 11 and ages 12 to 14—on violent and serious behavior

among individuals aged 15 to 25. They report that the key predictors from the six-to-11

age group are general offending and substance use, socioeconomic status of the family,

and having antisocial parents, and being male. From the 12-to-14 age group, the best

predictors are general offending, violence and aggression, and having antisocial peers.

Overall, offending, substance use, and antisocial peers are the strongest risk factors for

later deviant behavior (Lipsey and Derzon, 1998). A very important qualifier in their

study, however, is the fact that there is a high level of false positive predictions using

these risk factors. That is, use of the risk factors to predict behavior will err by predicting

many individuals will be deviant when, in fact, they will not.

C. Predicting Places and Events

Prediction for secondary prevention does not have to be limited to predicting

which individuals in which situations will turn to delinquency or criminality. It is also

possible to consider predicting the where and when of offending/victimization. This

activity is not a new or unique idea. Indeed, it is common for police agencies to distribute

their resources differentially across their jurisdiction and at different times of the day.

Today, researchers are employing new and developing technologies and data sources for

identifying the “where” and “when” of offending. The following discussion will look at

prediction in terms of hot spot analysis and hot products. The following section will

examine repeat victimization and related topics.

It has long been common practice for the police to identify locations and times

that are more prone to criminal activity. Neighborhood bars, for example, experience

more aggression and violence than lounges in nice restaurants. Similarly, assault is more

prevalent in the evening than during the mid-morning. Knowing even these two basic

facts shows that many problems cluster in both time and place. The challenge, therefore,

is to identify these clusters and use that information as a starting point for implementing

appropriate interventions. Perhaps the most recognizable example of this activity

involves “hot spot” research. Sherman (1995, p. 36) defines hot spots as “small places in

which the occurrence of crime is so frequent that it is highly predictable, at least over a

one-year period.” Analyzing calls for police service in Minneapolis, Sherman et al.

(1989) find that 50 percent of all calls for service came from only 3 percent of the

locations. All domestic disturbance calls appear at the same 9 percent of the places, all

assaults are at 7 percent of the locations, all burglaries occur at 11 percent of the places,

and all robbery, sexual misconduct, and auto theft calls appear at 5 percent of the possible

locations.

Attempts to identify hot spots are also useful in pointing out what types of crimes

and locations coincide. Block and Block (1995), using mapping techniques to examine

crime data for three Chicago communities, report that hot spots often surround elevated

transit stops and major intersections. These are locations where potential victims can be

located and offenders have options for escape. Looking at auto theft in Philadelphia,

Rengert (1997) identifies hot spots but notes that the locations of the hot spots change

according to different times of the day and night. Tourist attractions and educational

institutions may be hot spots for auto theft during the day, while entertainment venues,

bars, and other adult night spots become greater target areas in the evenings and at night

(Rengert, 1997). Clearly, hot spots can be anywhere—businesses, schools, abandoned

buildings, vacant lots, housing complexes or intersections—or anytime—evenings, late

night, weekends, holidays, or vacation months.

An important qualifier when considering hot spots should be stability over time.

Is the identified crime concentration a temporary situation, or does the hot spot persist

over a period of time? Townsley and Pease (2002) argue that relying on hot spots

identified with limited temporal data may lead to targeting anomalous crime

concentrations that will disappear as the crime settles back to its normal level (i.e., a

regression artifact). Perhaps more importantly, Johnson et al. (2005, 2008) claim that the

movement of crime, even over short periods, limits the value of identifying hot spots

using traditional methods. The authors suggest the use of prospective mapping, or the

creation of maps that predict future crime locations based on knowledge of recent events.

This is based on findings that show a burglary at one location results in heightened

chances of victimization at nearby locations (Johnson and Bowers, 2004). Prospective

mapping alleviates the problem of targeting hot spots when they move around.

A finding that crime concentrates in certain locations or at certain times suggests

that the targeting of hot spots may be an effective starting point for crime prevention. The

identification of a hot spot should prompt analyses to uncover what factors make a

location a good spot for crime (Spelman, 1995) and offer insight into preventive

responses. One set of tools that is becoming a central component to police planning is

computer mapping programs. Software, such as ArcGIS, MapInfo, the Spatial and

Temporal Analysis of Crime program (STAC), and Drug Market Analysis Program

(DMAP), are commonplace in policing (Rich, 1995). Mapping and hot spot research can

supply information not only on crime but also on information about the neighborhood,

site, or time at which an activity is taking place.

An interesting new approach to hot spots is the idea of hot products. Clarke

(1999) discusses hot products as items that attract attention and are targeted by thieves.

Further, such items may help explain the existence and distribution of hot spots. Hot

products are those that fit Felson and Clarke’s (1998) idea of VIVA: Value, Inertia,

Visibility, and Accessibility. Value is determined by potential offenders and not

necessarily the monetary cost of the item. What has value today may not be of interest

tomorrow. This could be due to the maturation of the offender, the saturation of the item

in society, changes in taste, or other factors. Inertia deals with the weight and portability

of the item. Further, a target can only be at risk if it is Visible to potential offenders.

Finally, the target must be Accessible to offenders (Felson and Clarke, 1998). The extent

to which a target meets these criteria will have an impact on the chances of an offense

occurring. Clarke (1999) expands on the idea of VIVA by proposing CRAVED

(Concealable, Removable, Available, Enjoyable, and Disposable), which further explains

the existence of hot spots. Clarke argues that identifying and acknowledging the

influence of “hot products” leads to a number of potential prevention measures. Physical

design ideas, such as electronic tagging, location transmitters, barcoding, and similar

methods of identifying property are prime examples of ways to address hot products. It is

also possible to develop other actions specifically aimed at hot products.

One problem with identifying repeat victimization involves the impact of short

time frames within which repeats can occur. Ellingworth et al. (1995) note that most

levels of repeat victimization are probably underreports because they rely on repeats only

within a limited time frame, which minimizes the potential for repeats before or after the

survey boundaries. The problem of short time frames for repeat victimization is very

evident when considering the NCVS, which uses a six-month time frame. Compared to

the 12-month time frame of the ICVS, the NCVS reveals significantly lower repeat

victimization compared to the ICVS for every category of victimization, including sexual

offenses (51 percent repeats in the ICVS; 23 percent in the NCVS), assaults and threats

(46 percent ICVS; 26 percent NCVS), and burglary (40 percent ICVS; 18 percent NCVS)

(Farrell et al., 2005). Kleemans (2001) notes that 9 percent of repeat burglaries occur

within one month, 30 percent occur within six months, and almost half occur within one

year. Thus, the time frame under consideration makes a difference for the finding of

repeat victimization.

Explanations for repeat victimization can generally be divided into two categories

—risk heterogeneity and state dependence (Farrell et al., 1995). Risk heterogeneity, or a

flag explanation (Gill and Pease, 1998), suggests that the prior victimization or some

other factor identifies the victim or location as an appropriate target for further

victimization. As such, subsequent victimizations may be committed by different

offenders who are attracted to the target by its apparent vulnerability or some other

characteristic. Farrell et al. (1995) use the example of repeated fights at a bar as an

indication of risk heterogeneity, where people looking for fights or interested in risky

situations are attracted to establishments with a reputation for conflict. Those locations

and/or the employees of those bars are then at a higher risk for repeat victimization.

Event dependency, or boost explanations (Gill and Pease, 1998), refers to situations in

which (usually) the same offender commits another offense based on the past experiences

with that victim or location. Successful past offending leads to another attempt against

the same target. It is possible under this situation that a new offender commits a follow-

up offense as a result of information shared between offenders. In this case, specific

information about the target based on a past offense is the key to subsequent actions.

Despite the increased interest in repeat victimization and evidence of the

effectiveness of targeting repeat victimization, there are issues that require more

attention. First, while evidence shows that there is a good deal of repeat victimization, not

all criminal acts are followed by another one against the same location or individual.

Identifying which acts will result in a repeat victimization prior to the subsequent act is

an elusive task. The existing research offers an after-the-fact analysis of the extent of

repeat victimization. It is possible, therefore, that targeting prevention activities at past

victims may result in a great deal of unnecessary effort. On the other hand, such targeting

should be more effective than interventions aimed at the general public, many of whom

would never become a victim in the first place.

A second issue deals with virtual repeats (Pease, 1998). A virtual repeat involves

a follow-up victimization of a similar person, place, or item. For example, a series of

robberies at different locations of a single company (such as a fast-food store) or theft of

the same brand of car could be considered repeat victimization if the subsequent offenses

are committed due to the similarity in the situations (such as similar store layout or

similar auto amenities). Pease (1998) suggests that these should be considered repeat

victimizations. As such, they offer different issues for directing crime prevention

activities. Johnson and Bowers (2002, 2004) illustrate this issue when they consider

burglaries that take place at neighboring homes as a type of repeat victimization, which

they call a near repeat. Their argument is that a local burglary elevates the risk of

burglary, at least in the near term, for other proximate homes.

Prediction is an important part of crime prevention—particularly secondary

prevention. This is true especially for attempting to intervene with individuals and

situations where a high propensity for criminal and deviant behavior exists. The

identification of persons who are headed for future juvenile or adult criminality would

allow the introduction of appropriate crime prevention techniques prior to the deviant

activity. Unfortunately, the prediction of future behavior of individuals typically results

in large numbers of false predictions. More recent research on risk factors uncovers a

number of variables related to later deviant behavior. It may be possible to use identified

risk factors as a basis for focused, preventive interventions.

D. The Theoretical Basis

The root ideas for situational crime prevention can be traced largely to the crime

prevention work of the British Home Office in the 1970s (Clarke, 1983). In pursuit of

interventions that could successfully address different crime problems, the Home Office

undertook a wide array of projects aimed at reducing factors specific to different crimes,

places, and situations. Clarke (1995) points out that much of this work grew out of the

recognition that crime often reflects the risk, effort, and payoff as assessed by the

offender. In essence, offenders make choices about which opportunities are the most

profitable, and act in accordance with that assessment. Rational choice, routine activities,

the lifestyles perspectives, and crime pattern theory are important considerations for

situational prevention. Each set of ideas provides insight into the ability of offenders to

respond to crime opportunities. While it is common for discussions of these perspectives

to focus on potential offenders, the potential victim is also an important part of the

equation.

The cornerstone of situational crime prevention is the belief that offenders

respond to opportunities and make choices in offending. Rational choice theory posits

that individuals make decisions on whether to commit an offense based on an array of

inputs, including the effort involved, the potential payoff, the degree of peer support for

the action, the risk of apprehension and punishment, and the needs of the individual

(Clarke and Cornish, 1985; Cornish and Clarke, 1986a, 1986b). This does not mean that

individuals commit an offense every time an opportunity presents itself. Rather, potential

offenders make a calculated decision about crime based on the available choices and the

risk, effort, and payoff involved. Research on burglars serves as a good example of the

issues involved in rational choice. Bennett (1986) and Wright and Decker (1994) point

out that burglars often commit their crimes in order to fulfill other needs and desires.

These may be immediate needs, such as cash for drug purchases or to meet expectations

of one’s peers, or longer-term desires for property or status. In attempting to satisfy these

drives, the offender may consider a wide range of targets and methods. Among the

physical and social factors that affect a burglar’s decision may be the level of

concealment, the amount of light, the presence of locks, evidence of valuable property,

surveillability from other places, and the presence of other people (Bennett, 1986;

Bennett and Wright, 1984; Cromwell et al., 1991; Nee and Taylor, 1988; Rengert and

Wasilchick, 1985; Wright and Decker, 1994). It is important to note that throughout these

discussions there is at least a tacit recognition of the limited nature of the offender’s

choices. Indeed, many of the choices may be made with little or no conscious decision

making by the individual at the time of the offense (Wright and Decker, 1994). While it

may appear that rational choice is not taking place, the decisions may have been

fashioned through a variety of past experiences, activities, and inputs. This same process

should not only apply to burglary, but also to other offenses.

The appearance that offenders’ behavior is based on little conscious thought or

choice may be due to the fact that individuals rely on crime scripts that drive their

actions. Cornish (1994, p. 159) notes that scripts are “a useful analytic tool for looking at

behavioral routines in the service of rational, purposive, goal-oriented action.” A crime

script outlines the steps and actions required to commit a crime, including the responses

that are needed to complete the act (Cornish and Clarke, 2008). A simple example of

script steps and actions for auto theft is presented in Table 11.1. The script scene

represents the steps and the script actions are the actual behaviors at each point in the

scene. Different crimes will have different scripts. While scripts are typically developed

over time through experience and practice, they can become second nature, requiring

little, if any, conscious thought. A potential offender, therefore, happens across a

situation offering all the elements outlined in a successful script and decides to act with

little apparent decision making.

One source of the information upon which an offender builds scripts and bases

decisions, whether consciously or subconsciously, is his daily routines. Routine activities

theory argues that the daily activity of individuals results in the convergence of motivated

offenders with suitable targets in the absence of guardians (Cohen and Felson, 1979).

This convergence provides opportunities for crime to occur. Cohen and Felson (1979)

demonstrate the importance of routine activities by showing that increases in the number

of unoccupied homes during the day and the greater availability of portable valuables

during the 1960s help to explain increases in residential burglary. Increasing mobility in

society serves to bring targets and offenders together with greater frequency than ever

before. Both the opportunity and choices for offending are enhanced.

Where routine activities deals with both the offender and the victim, the lifestyle

perspective specifically focuses on the activity of the victim as a contributing factor in

criminal acts. Hindelang et al. (1978) suggest that an individual’s lifestyle and behavioral

choices help determine whether he will be victimized. For example, frequenting a bar in

which violent fights are common increases the risk you will be involved in such a

confrontation. Similarly, working in a convenience store located in a high-crime

neighborhood enhances the possibility of being a robbery victim. In both situations the

individual’s lifestyle has an impact on the potential of becoming a victim or a repeat

victim.

The rational choice, routine activities, crime scripts, and lifestyles

theories/perspectives fit nicely with crime pattern theory. This theory argues that criminal

behavior fits patterns that can be understood in terms of when and where crime occurs

(Brantingham and Brantingham, 1993b). It is through daily activities that individuals

develop templates about the social and physical environment within which they operate.

This information is important for identifying both targets as well as threats for potential

offenders. Likewise, studying the behavior of offenders and analysis of past offending

provides insight into the crimes and potential prevention mechanisms.

E. Situational Typologies

Surveillance included many of the ideas discussed in earlier chapters, including

the concepts of natural surveillance, formal surveillance, and surveillance by employees

(see Clarke and Mayhew, 1980). Target hardening included interventions such as locks,

unbreakable glass, safes, and other security devices. Environmental management referred

to making changes that reduce the opportunity for crime. As one would expect, the

original three categories quickly became too simplistic and confining as the number of

situational crime prevention interventions grew. In 1992, Clarke offered an expanded

classification of situational techniques that reflected three very general orientations:

“increasing the effort,” “increasing the risk,” and “reducing the rewards.”

Clarke and Homel (1997) responded to concerns and limitations in the original

12-cell typology by proposing an expanded list of 16 situational techniques. The

expansion sought to address two key issues (Clarke and Homel, 1997). First, several of

the original categories could be divided to enhance internal consistency of the ideas

(Clarke and Homel, 1997). For example, “controlling facilitators” was divided into

“controlling facilitators” (such as guns and other items that make crime easier) and

“controlling disinhibitors” (factors that reduce the social and psychological barriers to

crime commission, such as the use of alcohol and other drugs). Second, the original 12

categories failed to include techniques that focused on the social and psychological

contexts of offending. Clarke and Homel (1997) added categories that addressed guilt,

shame, and embarrassment. Guilt and similar feelings may emerge because of the

incongruence individuals see between their actions and the moral code they hold or the

view of significant others in their lives.

The situational typology serves various purposes. First, the typology places the

great array of situational crime prevention activities and programs into a theoretical

framework. Many specific interventions, such as the installation of locks and lights, take

place with little understanding of the underlying rationale for why they should work.

While there are implicit theoretical arguments in many of the programs, understanding

why a program does or does not work requires more explicit recognition of the

mechanisms at work. The situational typology helps to organize those discussions.

Second, the cataloging of the diverse prevention efforts into a classification system helps

to identify the potential causal factors at work. That is, the underlying theories gain

support when it is possible to demonstrate their applicability and usefulness. The rational

choice, routine activities, and lifestyle perspectives all contribute to the development of

the situational techniques and benefit from the alignment of different studies into a

coherent typology. The recent addition of techniques for addressing precipitating factors

broadens the theoretical traditions and causal mechanisms under consideration in

prevention initiatives. Finally, on a very practical note, a classification scheme such as

this serves as a simple reference tool for those attempting to implement prevention

programs.

As with any topic, there are a number of concerns and unresolved issues. First, the

general categorizations are not mutually exclusive. That is, various interventions may

influence more than one factor (both objectively and subjectively). A technique that

increases risk also may increase the effort. For example, the presence of security guards

or CCTV increases the risk of being caught. These actions, however, also increase the

effort needed to successfully complete an offense. At the same time, successfully

completing a crime despite high risk and effort could lead to greater (psychic) rewards (or

monetary gain if the presence of protection reflected greater value). The fact that the

categories are not mutually exclusive does not negate the usefulness of the classification

scheme. Rather, it suggests that the underlying mechanisms are more complex than they

first appear in the typology. Second, the typologies are incomplete and in need of further

explication. Clarke and his colleagues view the typology as a dynamic undertaking that

will require modification as research and theory emerge. The very fact that the typology

evolved from 12 to 25 general techniques over roughly a 10-year period attests to the

dynamic nature of the undertaking. Indeed, an attempt to finalize a typology could be

viewed as limiting its usefulness.

F. Implementing Situational Prevention

Besides offering an extensive typology of potential interventions, situational

prevention distinguishes itself by approaching problems in a very systematic fashion.

Situational prevention has borrowed the SARA process from problem-oriented policing

as the model for problem-solving. Proposed by Eck and Spelman (1989), SARA stands

for Scanning, Analysis, Response, and Assessment. Under scanning, situational

prevention starts with a specific, identifiable crime problem. That problem is then

subjected to analysis, drawing on as wide an array of information and perspectives as

possible. This stage offers an important distinction from many other crime control efforts

that rely on traditional police and criminal justice system responses to solve problems.

Under situational prevention, formal social control agents are only one source of input.

Based on the findings of the analysis, a response (intervention) is identified and

implemented. At this point the prevention process continues with an assessment of the

program’s impact, with the intent of making changes in the response, if necessary.

Situational prevention, therefore, is a dynamic process of problem identification,

response identification, program implementation, and evaluation and adjustment.

While SARA is the most recognized approach, Ekblom (2002) proposed the 5Is

as an alternative model for problem solving that is receiving increased attention in

situational prevention. The 5Is are Intelligence, Intervention, Implementation,

Involvement, and Impact. The relationship between the 5Is and the more recognized

SARA process is depicted in Figure 11.1. The greatest distinction to note is the greater

detail introduced by breaking down the Response into Intervention, Implementation, and

Involvement. Intervention represents the identification of actions and methods to “block,

divert or weaken the causes, and attend to risk and protective factors, of future criminal

events and careers or of wider community safety problems” (Ekblom, 2011a, p. 85).

Once the crime prevention action is identified, Implementation represents the actual tasks

of putting the plan into action. The third aspect, Involvement, focuses on marshalling the

participation of other people and organizations in the intervention. Ekblom (2011a)

argues that the greater emphasis on knowledge generation in the 5Is makes them useful

for not only situational crime prevention, but also policy making and beyond.

The 11 D principles fall within three basic, overlapping modes of action (depicted

in Figure 11.2). These modes are: Practical (where the environment is altered to limit

perpetrator action), Psychological (actions that impact the thought processes of the

perpetrator), and Personal (involving identifying, tracking, and catching perpetrators)

(Ekblom and Hirschfield, 2014). The overlap in the modes indicates that specific

Ds/interventions may impact the perpetrator/event in more than one way. The D

principles reflect a need to provide prevention and security professionals with steps for

turning situational prevention techniques into action. These principles are midrange

between theory (such as rational choice and routine activities) and extended lists of

potential interventions (such as the 25 situational techniques). Ekblom and Hirschfield

(2014) suggest that this approach offers professionals a useful framework for identifying

appropriate interventions for different situations. The 11 Ds can be viewed as a

refinement of the broader situational crime prevention ideas.

The balance of this chapter offers a brief overview of studies examining the

implementation of various situational crime prevention techniques. For most of the

discussion, the emphasis is on the impact of various prevention techniques. In some cases

the examples highlight suggestions for situational prevention for specific crime problems.

It is not possible to provide example studies for each of the 25 techniques offered by

Cornish and Clarke (2003). Instead, the studies reviewed here should provide the reader

with some insight into the types of initiatives/approaches that have been undertaken and

the breadth of the problems to which situational techniques have been and could be

applied. A broader array of studies can be found in other forums.

The failure to pay fares is a common problem on public transportation and a

single response is not possible due to the diversity in types of transit system (bus, train,

etc.) and the means by which offenders evade payments. An assessment of fare evasion

on the British Columbia transit system identified various methods of fare evasion, which

formed the basis for a redesign of ticket machines and passes, promotions encouraging

the purchase of transit passes, and focused investigations of counterfeit fare media

(DesChamps et al., 1991). An evaluation of the intervention revealed significant drops in

the level of fare evasion. In an evaluation of the Dutch transit systems, van Andel (1989)

found that fare evasion on buses often occurred due to the ability of riders to enter and

exit the buses through a rear door, thus avoiding the driver. In response, the system

introduced monitors on the buses and a change in procedure that required passengers to

enter the bus near the driver and prove payment (van Andel, 1989). Fare evasion

significantly fell from the pre-intervention levels.

Motor vehicle theft is a continuing problem that has generated a variety of

different proposed solutions. Mayhew et al. (1976) and Webb (1994) examined the

impact of steering column locks on the level of motor vehicle theft in England, Wales,

and West Germany. Both analyses show that increasing the difficulty of motor vehicle

theft led to reductions in the level of theft. Since the mid-1990s, increased use of

automobile immobilizers has been undertaken to avert auto theft. Immobilizers are

electronic devices that, in the absence of the key, prevent the car from operating (Brown,

2004). These devices have been required in all new cars in the European Union since

1995. Analysis of immobilizers in the United Kingdom reveals clear reductions in auto

theft (Brown, 2004). Efforts to address motorcycle theft have often involved legislating

the use of helmets (Mayhew et al., 1989). The need for helmets should reduce

opportunistic theft because most offenders do not have a helmet available to them, thus

making them easily recognizable to others. Evaluations in England, The Netherlands, and

(the former) West Germany all find reduced motorcycle theft after the initiation of helmet

law.

Situational techniques have been used in relation to a wide array of theft offenses.

The impact of electronic tagging of property has been demonstrated in a number of

studies. In one study, Scherdin (1986) found that the tagging of books in a university

library resulted in both a significant decrease in lost books and an increase in the number

of items processed through circulation. Perhaps of greater impact was the simultaneous

reduction in lost audiovisual materials that could not be tagged (Scherdin, 1986). This

suggests a diffusion of benefits from the books to other library holdings. Farrington et al.

(1993) report that electronic tagging significantly reduced the level of retail shoplifting.

Similarly, Handford (1994) indicates that computer stores using electronic tagging

experience low theft levels relative to national data for comparable stores.

One important aspect of situational crime prevention includes designing or

redesigning at-risk products for crime prevention. It is possible to consider physically

designing products in ways that protect them from theft or being used in other offenses.

As Clarke and Newman (2005) point out, the targets of crime are everyday objects. It is

possible to alter the design of objects to make them less amenable to crime. Many

products lend themselves to crime. These so-called hot products which are highly

targeted by thieves may be characterized by being CRAVED—Concealable, Removable,

Available, Valuable, Enjoyable and Disposable (Clarke, 1999). In essence, these are

products desired by people because of their construction or makeup and are easily

targeted by offenders. For example, electronic devices such as MP3 players and PDAs

are easily concealed, expensive, small enough to carry away, and easily disposable. At

the same time, automobiles are not easily concealable or even disposable, but they are a

common theft target. It is also possible with modern technology to take seemingly low-

risk items and make them more susceptible to theft. This can include making fraudulent

credit cards, drugs, designer bags, and food (to name a few).

There are a number of other examples of product design for prevention purposes

(Clarke and Newman, 2005). Theft of small items from stores can be made more difficult

by putting the item in large packages that are more difficult to conceal. Purse snatching

can be prevented by designing them with stronger straps that cannot be cut or broken.

Computer hacking has been made more difficult with the introduction of special

software. Caller ID has helped to stem telephone harassment (Clarke and Newman,

2005). The design of smart guns recognize the owner and will only discharge if used by

that person (Lester, 2001). These and many other examples of product design changes are

made for the purpose of preventing different types of crimes.

Cozens (2014) notes there is a crime life cycle for products that is tied to the

market for each type of product. The first stage is Introduction of a new product. The

product is relatively rare and has little market share or recognition; thus, it is not a high

target for theft. This moves to the Growth stage where sales are increasing and

knowledge and demand is on the up slope. The third stage is the Mature stage. Here the

product is well known and desired. There are a lot of the items available for theft and the

demand leads to theft for both use and sale to others. This is the peak period for crime.

The final stage is Decline, where the product has either saturated the market or is being

replaced with new products. Theft falls off in this stage. Some products go through this

cycle quickly, some very slowly, and others may not fit the cycle at all (Cozens, 2014). It

is not unusual that public pressure or government intervention is need to help interrupt

this cycle and force project changes.

G. Community Policing

Understanding community policing requires some knowledge of the more

traditional view of what the police are and what they should do. For most people, the

formal police role is to answer calls about crime, undertake investigations, make arrests,

and assist in the prosecution of offenders. Fighting crime (and crime prevention) is

primarily the responsibility of the police. Consequently, the police are often judged by

the level of crime in the community. Arrests are the benchmark by which the public

judges police effectiveness. The police also are called on to provide order maintenance,

that is, functions that do not deal with an immediate criminal action. Order maintenance

includes responding to disabled autos, escorting funerals and parades, dealing with

barking dogs, responding to false alarms and noise complaints, and delivering messages.

Various studies show that these activities consume the majority of police time, with the

police spending roughly 20 percent of their time on actual law enforcement efforts

(Kelling, 1978; Lab, 1984; Walker, 1983; Wilson, 1968). While critics argue that the

police should not be involved in order-maintenance activities, Hoover (1992) argues that

order maintenance is a key part of traditional policing and serves to keep society

functioning in an orderly fashion. As such, order maintenance enhances the law

enforcement efforts of the police.

Trying to respond to all of the desires of the citizenry is not an easy task and is

one possible reason for the inability of the police to control serious crime. Criticism of

the police and the apparent failure of past police practices to stem the crime problem have

led to the introduction of different strategies. One approach was the revival of foot patrol.

Research shows that foot patrol has had a mixed impact on crime (Bowers and Hirsch,

1987; Esbensen, 1987; Police Foundation, 1981; Trojanowicz, 1983), although it appears

to reduce the level of citizens’ fear and improve attitudes toward the police (Brown and

Wycoff, 1987; Police Foundation, 1981; Trojanowicz, 1983). The police are also a key

ingredient of both the establishment and maintenance of neighborhood watch and citizen

crime prevention initiatives. Underlying these activities has been the recognition that the

police cannot solve crime on their own or address the increased calls for assistance by the

public. The police need to build better relations with citizens and increase the

involvement of citizens in crime prevention and crime policy.

The fact that elements of community policing appear in past practices

demonstrates that community policing is not a totally new idea, although the term

“community policing” is a relatively new one. In addition, there has been a shift in the

basic orientation of everyday police activity emerging under the rubric of “community

policing.” This shift is from the traditional view of “crime fighting” through arrests to a

view that fighting crime involves a broader set of interventions. Arriving at a single

definition of community policing has proved to be an elusive goal. While a single

definition has not emerged, the various definitions found in Table 12.1 tap the essential

elements offered by most writers. Most of these definitions reflect the fact that

community policing is more of a philosophy of policing, rather than a clearly definable

method (see, for example, Greene and Mastrofski, 1988; Trojanowicz and Bucqueroux,

1989; Walker, 1999). Beyond the fact of this philosophical shift, the various definitions

of community policing generally include several essential features. These are community

involvement, problem solving, a community base, and redefined goals for the police.

The emphasis on problem solving is perhaps the most important element of

community policing. Rather than simply dealing with the crime that occurs through

investigation and arrest, community policing challenges officers to identify the

underlying causes and contributors to the crime, and seek out solutions to those problems.

Community policing, therefore, sees crime as a symptom of more basic concerns. The

police can either deal with the symptom or try to address the ultimate cause. While law

enforcement should do both, the community policing orientation shifts the primary

attention to the underlying problems.

This orientation also appears under the name problem-oriented policing. Problem-

oriented policing means approaching issues and problems differently based on the

uniqueness of each situation. This is a drastic shift from the traditional view that the

police should use the criminal code to respond to calls for service. If the code prohibits

the activity, the police can (and should) make an arrest and set the criminal justice system

in motion. At best, this response will eliminate further criminal behavior through either

its deterrent or incapacitative effect. Events that are not proscribed in the criminal code

can be ignored. The problem-oriented approach argues that invoking the criminal code is

only one avenue for dealing with societal issues. Instead, different problems require

alternative solutions or interventions. The police, therefore, need to identify and pursue

solutions to the root problem.

Critical to community policing is the decentralization of the police operation.

Community policing typically means assigning officers to a specific neighborhood. This

may be done in a variety of ways, including the establishment of neighborhood stations,

storefront offices, foot patrol, and others. The assignments are long term, with the

expectation that the officers will intimately get to know the community, its problems, and

its citizens. Community policing assumes that the failure of the police to identify

problems and relate to citizens, among others, is due to the distance that central stations

and patrol cars place between officers and citizens. The daily interaction with the

community should alert the officer to the problems and needs of the residents. The

expected result is increased goodwill on the part of residents and an increased desire by

citizens to assist and involve themselves with the police.

The fourth major element of community policing involves altering the goals of

policing. In one sense, this is closely allied to the idea of identifying and attacking the

root causes of problems. At the same time, however, this could mean a great shift in how

the police are judged. Most departments are judged, both by themselves and the citizenry,

in terms of the number of arrests made. Community policing initiatives, to the extent that

they deal with underlying causes and involve officers in non-arrest activities, require that

the department and officers be held to different standards. In addition to arrests,

community policing programs can be judged by reduced crime, the elimination of

problem properties, increased feelings of safety, less neighborhood disorder, community

cohesion, and many other outcomes. It is important to note that community policing

should emphasize the ends, rather than the means to the ends. That is, instead of focusing

on how things get done, the primary concern is the elimination of the root problem.

H. Problem Identification

Perhaps the central task of any partnership is the identification of problems and

their solutions. Eck and Spelman (1989) offer a four-step process for problem solving.

These steps are referred to as SARA: Scanning, Analysis, Response, and Assessment.

Scanning involves the identification of the problems, issues, and concerns in the

community. This information may arise from the observations officers make as they work

in the community, from residents or businesses who bring problems to the officers, from

other agencies (such as schools or hospitals) in the community, or from the systematic

study of data and information on the area. One primary method used in scanning involves

computer-generated analyses of when and where crime occurs. The analysis of crime

“hot spots” (Sherman, 1995) and the generation of crime maps represent two such

attempts. Other efforts may involve the analysis of calls for service, systematic

observation of the community, or surveying citizens or community groups. In every case,

the police should be working in partnership with others to identify problems that need to

be addressed.

The second stage is the analysis of the problem. It is particularly important that

more than just the police are involved in this activity. An array of individuals and

agencies should participate in the analysis. For example, if drug dealing is centered in a

house, apartment, or public housing building, the police, landlords, housing authority

personnel, the health department, and/or the city attorney need to be involved in the

problem analysis. If the drug activity centers on youths, it may be advisable to include the

schools, probation office, or youth groups in the process. The intent is to bring together a

diversity of expertise and insight. This diversity will bring different information and

viewpoints and assist in understanding what factors are involved in the problem.

It is from this cooperative interaction that different responses will emerge. Who is

involved in implementing the response will vary greatly. In some cases the police may

have little day-to-day involvement in the intervention because the identified response

requires expertise and abilities that the police do not have. An example of this would be

the use of civil litigation against owners of property where drug use is allowed to

continue. While the police can deliver summonses, much of the work will be conducted

by lawyers for the jurisdiction and other departments that can help to shut down the

building (such as a health department).

I. Partnership Efforts and Assessment

Partnership programs have grown throughout the United States, Canada, the

United Kingdom, and many other countries. Under the heading of community policing,

Wycoff (1995) reports that approximately 800 law enforcement agencies in 1993 had

implemented community policing in the United States. As of 2001, more than 10,000 law

enforcement agencies report more than 113,000 community police officers on duty in the

United States (Zhao et al., 2002). Partnerships, however, appear under a variety of

headings and involve police in a wide range of capacities. These partnerships vary

greatly, largely due to the emphasis on identifying interventions that address a specific

problem. The following pages discuss various partnership initiatives.

The Chicago Alternative Police Strategy (CAPS) is perhaps the best example of

successfully implementing a community-oriented policing approach. Chicago opted to

move its entire police force into community policing. CAPS began in five of the city’s 25

police beats in 1993 with the support of the police administration and the mayor (Hartnett

and Skogan, 1999). Key aspects of the program include assigning officers to permanent

neighborhood beats, the involvement of residents in the identification of problems and

potential solutions, and reliance on other agencies (both public and private) to address

identified issues. Citizen interaction is the cornerstone of the program and the police meet

with neighborhood residents to engender meaningful interaction. These meetings identify

a wide range of local problems, including gangs, drugs, graffiti, burglary, and physical

and social disorder.

The recognition that crimes often cluster in place and time has led to specific

efforts to address that crime convergence, an approach called hot spots policing.

Attacking hot spots requires analysis of the locale and the generation of interactions

appropriate for the problem. This can include everything from traditional policing to

problem-oriented policing to situational crime prevention techniques.

An interesting partnership for dealing with problems, particularly drug issues,

involves the use of civil abatement and injunction procedures. Civil abatement uses

mainly civil and administrative law and codes to control locations and behavior at

locations. These activities may involve landlords, citizens, health departments, zoning

boards, the police, and city/county attorneys. Injunctions involve court orders placing

rules and/or restrictions on the behavior of individuals. Civil gang injunctions may

prohibit gang members from associating in public, marking territory, trespassing,

loitering, or other similar activities (LA City Attorney’s Office, 2009). The advent of

these efforts can be traced to work in Portland, Oregon, in 1987.

Oakland’s Specialized Multi-Agency Response Teams (SMART) is a prime

example of civil abatement. SMART relies on the cooperation of police, citizens, and

other groups to solve neighborhood problems. Civil court remedies are a cornerstone of

the project, although police enforcement and patrol are important components. In a series

of reports, Mazerolle and her colleagues demonstrate the effectiveness of the program.

One hundred target sites, divided evenly into experimental and control groups, formed

the basis of the evaluation. At experimental sites, landlords were contacted by police and

received assistance (from the police and other agencies) in dealing with problem tenants,

including evicting intransigent individuals. Civil proceedings could be brought against

both the landlords and the tenants. Data from calls for police service, interviews, and

observations over more than three years reveal significant declines in signs of physical

and social disorder, decreased drug sellers, and increased levels of civil behavior (Green,

1995b; Mazerolle and Roehl, 1999; Mazerolle et al., 1998). Control areas showed

increased problems or little change in key outcomes.

Two other examples of civil abatement are San Diego’s Drug Abatement

Response Team (DART) and Project TOUGH (Taking Out Urban Gang Headquarters) in

Los Angeles. The Drug Abatement Response Team (DART) works to compel landlords

to take action against properties and tenants involved in drug offending (Eck, 1998; Eck

and Wartell, 1998, 1999). Project TOUGH (Taking Out Urban Gang Headquarters) does

the same in relation to gang headquarters and other gang hangouts. Both programs seek

to force owners to take responsibility for their property. Cristall and Forman-Echols

(2009) note that the use of civil suits is in some ways easier than invoking the criminal

code because the burden of proof is “preponderance of evidence” rather than “beyond a

reasonable doubt” and there is no right to jury trial. An evaluation of the DART program

using a randomized control trial in problem neighborhoods showed greatest improvement

at full DART intervention sites and the results persisted across a 30-month follow-up

period. Evictions were more evident in the treatment group, as were lowered levels of

crime and drug problems.

Three major U.S. initiatives for partnership building are the Comprehensive

Communities Program (CCP), the Strategic Approaches to Community Safety Initiative

(SACSI), and Project Safe Neighborhoods (PSN). Each of these projects has similar

features. The Comprehensive Communities Program (CCP) was initiated in 1994 as a

partnership-building initiative to fight crime and improve the quality of life in

communities (Bureau of Justice Assistance, 2001). Fifteen communities have participated

in the project. The major key to CCP is to use a problem-solving approach that includes a

wide array of community individuals, agencies, and groups. Of particular interest is to

bring the individuals and groups most affected by crime into the project. Materials

available on the CCP spend a great deal of time describing the activities of the

partnerships in each site, with special attention paid to the coalition-building activities

(see Bureau of Justice Assistance, 2001; Kelling, 1998). Few of the sites provide outcome

measures on changes in crime, and those that do generally fail to provide data for

comparable control areas (Bureau of Justice Assistance, 2001). It is not possible,

therefore, to know whether the results are due to the CCP initiative or some other cause.

The Strategic Approaches to Community Safety Initiative (SACSI) project was

initiated in five sites in 1998 to fight primarily violent personal crimes. The lead agency

in the SACSI sites is the local U.S. Attorney’s Office, which attempts to build a

partnership consisting mainly of other criminal justice agencies. In addition, each

partnership includes a local research team whose task is to analyze the problem in the

community, participate in selecting the appropriate response, and evaluate the operations

of the SACSI team. Local, state, and federal law enforcement agencies, probation and

parole offices, local, state, and federal attorneys, and social service agencies form the

primary core of participants. Most material available on the SACSI project is restricted to

process evaluations, which show that partnerships can be successfully established,

although a concerted effort is often needed to sustain them. One analysis of SACSI in St.

Louis (Decker et al., 2005) did look at reduction in the target behavior (gun crime) and

notes that homicides decreased after the initiation of the project. An evaluation of the

TimeZup SACSI program in New Haven uncovers decreased levels of violent gun

crimes, calls for service, and numbers of guns seized. Other positive results included

reduced fear of crime and increased confidence in the police (Hartstone and Richetelli,

2005). An impact evaluation of Chicago’s PSN project presents significant reductions in

homicide, gun-related homicides, and aggravated assaults, although no impact on gang-

related homicides (Papachristos et al., 2007). Finally, Rosenbaum and Roehl’s (2010)

national evaluation of SACSI reveals decreases in homicide and violent crime in 10

program cities compared to all cities of 100,000 population or more.

Several partnerships have addressed gun violence, particularly among juveniles

and gang members. Perhaps the most well-known of these is Boston’s Operation

Ceasefire. Begun in 1996, the project had several key features. Perhaps the most

important feature was the creation of an interagency working partnership. The task for the

group was to assess the nature of the gun problem and the dynamics of youth violence in

Boston, and to identify and implement an effective intervention (Kennedy et al., 2001).

The outcome of the planning was a set of interventions aimed primarily at gangs and

gang members. Utilizing the manpower and resources of the police, probation, parole,

and district attorney’s office, the project used a strict enforcement policy for all

individuals and groups involved directly or indirectly in gun violence. Dubbed “pulling

levers,” the project would take any and all actions possible against violators. That meant

that any gun violence would result in the immediate arrest and full prosecution of

violators. Probation and parole violators were vigorously prosecuted for any violation of

the conditions of their release. Gangs and gang members were notified of the project and

the potential consequences of their actions (Kennedy et al., 2001). Social service agencies

and federal agencies (such as Immigration and Naturalization, which was used to deport

non-citizen offenders) were also included in the project.

The passage of the U.K. Crime and Disorder Act in 1998 mandated the

establishment of community partnerships to combat crime and related problems. These

partnerships are meant to include the local police and a variety of community

constituencies, including housing authorities, victims, health professionals, probation

officers, and others. Newburn (2002) notes that the goal is to address “multi-dimensional

problems with multi-dimensional responses.” Crime is not the only problem to be

tackled, thus the need for wide participation by varied groups other than law

enforcement. More than £925 million were spent funding partnerships between 1999 and

2005.

In the mid-1990s, the Office of Juvenile Justice and Delinquency Prevention

(OJJDP) initiated the Community-Wide Approach to Gang Prevention, Intervention, and

Suppression Program. The program aimed to initiate a comprehensive set of strategies

including suppression, opportunities provision, and social interventions. OJJDP funded

program implementation and evaluation in five cities (Bloomington-Normal, IL; Mesa,

AZ; Riverside, CA; Tucson, AZ; and San Antonio, TX). Each of the programs drew on

the expertise of law enforcement, schools, employment organizations, and other social

service agencies. Several factors emerge across the evaluations. First, several cities

struggled with building programs that included grassroots community organizations.

Most of the participants remained official criminal justice agencies and other social

service providers (Spergel et al., 2001, 2002, 2004a, 2004b). Second, suppression

remains the primary response in at least three cities (Spergel et al., 2001, 2002, 2003,

2004b). Finally, the more successful programs offered a wider array of activities that

could be considered opportunities provision and social interventions, such as counseling,

referrals, and job training.

Another major gang suppression initiative is the Comprehensive Anti-Gang

Initiative (CAGI). This program was developed by the U.S. Department of Justice in

response to the earlier “community-wide” program and PSN. As with the earlier

initiative, CAGI seeks to build a coalition of agencies, including law enforcement,

prosecutors, social services, governments, community groups, schools, and others. CAGI

addresses enforcement, prevention, and reentry (McGarrell et al., 2012). A total of 12

cities received funding for program implementation.

J. The Scope of Drug Use

The extent of drug use is somewhat difficult to gauge due to the difficulties of

measuring private behavior. Unlike most other crimes, drug use has no victim

independent of the offender—they are one and the same. Available data relies on self-

reports of drug use in the general population or on information on known offenders. A

discussion of both of these sources is presented below in order to gain some

understanding about the drug problem.

Perhaps the most well-known survey of drug use is the Monitoring the Future

(MTF) project carried out by Johnston and associates. This project consists of surveying

representative high school students (eighth, tenth, and twelfth graders), college students,

and young adults (Johnston et al., 2014). “Young adults” refers to high school graduates

within 10 years of leaving school. While the survey probes a variety of factors, the most

important set of information deals with the level and type of drug use. Table 13.1

presents information on lifetime, annual, and past-month drug use for high school

seniors, college students, and young adults in 2013. It can be seen that drug use varies by

type of drug.

The self-report figures must be considered cautiously due to some inherent

deficiencies. With the MTF data there is a question of generalizability. The data are not

representative of the entire population. The fact that the high school and college subjects

are (or were) attending school ignores the fact that many youths drop out of high school

or do not go to college. Dropping out is especially great among inner-city youths, where

the drug trade appears to be most concentrated. Johnston et al. (1989) point out that

roughly 15 to 20 percent of high school students drop out and are not included in the

survey each year, and that dropouts tend to use drugs more often than non-dropouts. In a

study of runaway/homeless youths, Fors and Rojek (1991) find that these individuals

report two to seven times the level of substance use/abuse of school youths. These facts

suggest that the MTF data underreport the level of drug use.

It is commonly assumed that many offenders are regular users of illicit drugs and

that drug use is intricately related to the commission of crime. Support for such a

contention comes from various studies of offenders. The ADAM II (Arrestee Drug Abuse

Monitoring) program provides in-depth information on drug use by offenders in 10 cities.

The original ADAM program was an expansion of the Drug Use Forecasting (DUF)

program sponsored by the National Institute of Justice, which began in 1987. ADAM II is

a scaled-down version of the original ADAM program, which gathered data in 35 cities.

Drug use figures are also available for inmates of correctional institutions and

those in court. Mumola and Karberg (2006) note that over one-quarter of federal inmates

and almost one-third of state prisoners report committing their offenses while using

drugs. Additionally, over half of the prisoners report drug use in the month before

offending Sickmund et al. (2013) report that roughly 7 percent of the youths in residential

facilities are there due to drug-related offenses, while juvenile court statistics show

roughly 165,000 youths entered court for drug law violations in 2010 (Puzzanchera and

Hockenberry, 2013). These figures, based on offending populations, provide a more

serious picture of drug use than those from surveys of the general population. As with the

other data, there are potential problems to keep in mind. First, this information reflects

only the individuals who are caught by the system. It is possible that the use of drugs

increases the risk of apprehension for these offenders. Those not apprehended may not

use drugs, or at least not at the same level. Second, the ADAM II data reflect drug use in

10 major urban areas. Consequently, the results are only generalizable to other

comparable large cities. Drug use probably differs between large and small communities,

just as crime differs.

K. The Drugs–Crime Connection

The connection between drug use and crime has received a great deal of attention.

A simple inspection of data reveals a strong correlation between drugs and crime.

Situations in which many of the offenders test positive for drug use provide a strong basis

for claiming a causal relationship. Simple correlations, however, are not enough to

establish clear evidence of causation. The relationship between drug use and crime has

been hypothesized to take a variety of forms. White (1990) outlines four possible models

for the relationship (see Figure 13.1). First, drug use causes criminal activity. Second,

criminal activity causes drug use. Third, there is a reciprocal relationship in which both

drug use and criminal activity cause one another. Finally, the relationship between the

two is spurious with other factors (possibly the same ones) causing drug use and crime.

Various studies have attempted to untangle which causal sequence is correct.

The first model attempts to show that drugs cause the user to commit other

crimes. Crime may be the result of a psychopharmacological reaction, economic need, or

simple participation in the drug trade (Goldstein, 1989). The psychopharmacological

explanation suggests that various drugs have a direct impact on the user, both physically

and psychologically, which impels the individual to act in such a way that society deems

unacceptable. The intent of the individual may not be to commit a crime. The drug

simply determines the action, which may or may not be criminal. The economics of drug

use can also lead to deviance. The increasing need for money to secure drugs can lead to

property crimes. Various studies show that drug users are often involved in property

offenses (Anglin and Speckart, 1988; Chaiken and Chaiken, 1982; Collins et al., 1985;

Harrison and Gfroerer, 1992; Johnson et al., 1985, 1988; National Institute of Justice,

1990). Participation in drug use also may lead to systemic violence (Goldstein, 1989).

Systemic violence refers to violence resulting from competition among drug dealers,

retaliation for poor drug quality or high prices, robbery of drug dealers or users, and other

factors related to the drug trade. Using New York City data, Goldstein et al. (1992) report

that 74 percent of all drug-related homicides are due to systemic factors of the drug trade.

To the extent that crime fits one of these categories, the first model finds qualified

support. Of the three potential arguments, the psychopharmacological explanation is the

most difficult to prove and has been criticized for ignoring evidence that much drug use

actually reduces physical action and violent tendencies, and only appears in interaction

with specific dispositions and social/cultural settings (McBride and Schwartz, 1990). The

economic and systemic arguments require attention to the temporal order in the drugs–

crime relationship.

Studies claiming support for drug use causing crime typically rely on studies of

drug addicts or high-rate users of drugs. Ball et al. (1983), studying 354 heroin addicts,

note that their crime rate is four to six times higher when they are actively using drugs.

Similarly, Collins et al. (1985) report that daily heroin/cocaine users tend to commit

property offenses at a substantially higher rate than weekly users or non-users. Drug use,

especially involving expensive drugs and drug habits, necessitates the commission of

“income-generating crimes” in order to maintain the pattern of use (Collins et al., 1985).

Anglin and Hser (1987) and Anglin and Speckart (1988) note that arrests and self-

reported crimes increase at the onset of first narcotics use and first daily use, and decrease

at last daily use. Use also declines during treatment and increases when subjects leave

treatment (Anglin and Hser, 1987; Anglin and Speckart, 1988). Huizinga et al. (1994)

note that changes in substance abuse precede changes in the levels of other delinquent

activity.

The second possibility is that involvement in crime causes drug use. Numerous

studies suggest that involvement in criminal behavior precedes drug use. Using

longitudinal data on almost 2,000 high school graduates in the Youth in Transition

project, Johnston and associates (1978) report that general delinquency predates most

drug use. The authors argue that youths turn to drug use as an extension of other deviant

behavior. Data from the National Youth Survey (NYS), an ongoing longitudinal panel

study, reveal essentially the same result. Information on delinquency, drug use, and

demographic factors for the first six waves of data collection (1976 to 1983) shows a

general progression in behavior starting with minor delinquency and leading to alcohol

use, index offenses, marijuana use, and polydrug use, in that order (Huizinga et al.,

1989). Except for the early appearance of alcohol, illicit drug use temporally follows

delinquent/criminal behavior. Indeed, the authors note that minor delinquency precedes

polydrug use 99 to 100 percent of the time (Huizinga et al., 1989). Chaiken and Chaiken

(1990), examining data from the NYS, a survey of prison inmates, and a sample of New

York City drug addicts, note that delinquency predates drug use at least 50 percent of the

time. The same pattern of drug use following delinquency is reported by Inciardi et al.

(1993) in a study of serious inner-city delinquents. Other authors (Anglin and Hser, 1987;

Anglin and Speckart, 1988; Elliott and Ageton, 1981; Elliott et al., 1989; Hunt, 1990;

Johnson et al., 1985) also note that drug use follows crime. The typical explanation for

this finding is that drug use is simply another form of deviant behavior, and involvement

with delinquency/criminality provides resources and contacts necessary for entering into

drug use.

Given the studies reporting a different sequencing in the drugs–crime relationship,

it is plausible to argue that the actual relationship is reciprocal. That is, criminal activity

leads to drug use and drug use leads to criminal activity. Support for a reciprocal

relationship can be found in many studies. Nurco et al. (1988) note that addiction

increases crime by “previous offenders.” Studies of drug addicts typically relate that

arrests “increase” after drug use or intensified use, leading to the conclusion that the

subjects committed offenses prior to that point in time (Anglin and Hser, 1987; Anglin

and Speckart, 1988; Collins et al., 1985). Hunt (1990) points out that, while prostitution

increases with drug use, a substantial number of female drug users participated in

prostitution and other offenses prior to drug use. Finally, van Kammen and Loeber’s

(1994) analysis of data for Pittsburgh youths shows that property offending predicts the

onset of drug use, although drug use escalates the commission of personal crime. It would

appear that, regardless of which came first, drug use and crime may contribute to each

other. Drug use leads to crime and crime leads to drug use.

The argument that seems to be gaining the strongest support is the claim that the

relationship between drug use and delinquency is spurious. This simply means that, while

use and crime exist at the same time and vary in a similar fashion, neither is the ultimate

cause of the other. Rather, they are caused by either the same common factors or by

different factors. Huba and Bentler (1983) and Kandel et al. (1986) claim that there is no

causal relationship between drug use and crime. Rather, these two sets of behavior are

caused by other similar factors. White et al. (1987), using self-report data for almost 900

youths, point out that there are common causes of delinquency and drug use. Foremost

among these factors are peer and school influences. The same analysis of the NYS data,

which points out a sequence of behavior beginning with minor offending and ending with

polydrug use (Huizinga et al., 1989), concludes that the actual cause of the behaviors

probably lies with a common set of spurious influences. Other research leads to the same

conclusion.

L. Drugs and Crime Prevention

Interventions aimed at limiting drug use and related crime take a variety of forms.

Most of the approaches fall under three general areas—law enforcement, treatment, and

prevention. The great number and diversity of possible responses to the drug–crime

problem cannot be adequately discussed in the space available. Indeed, some of the

proposed solutions do not fit in the present discussion. The following discussion,

therefore, is somewhat selective. Among the topics not covered here (or covered in

abbreviated fashion) are efforts at drug interdiction and foreign policy issues, targeting

organized crime, drug testing, decriminalization, and legalization. While each of these

topics is worthy of discussion, they are beyond the scope of the present work.

The current “drug war” is primarily an effort that uses law enforcement

techniques as the primary weapon against drug use. This is evident in the federal

government’s budgetary priority favoring law enforcement over treatment or prevention

programs. Underlying law enforcement actions is the assumption that drug use and

related crime can be limited or eliminated by supply reduction. Taking drugs off the

street will make it more difficult for users to locate the drugs and, if the drugs can be

found, the price will be driven so high that many potential users will simply abstain. In

essence, these approaches assume that the demand for drugs is a factor of the supply.

Police crackdowns on drug availability are common responses. New York City’s

Operation Pressure Point involved saturating the Alphabet City area on the Lower East

Side of Manhattan with police officers and resources. Drug buyers could shop around

different dealers, often in plain view, in order to secure the best price and quality in the

area (Zimmer, 1987). The police operation was credited with thousands of arrests, the

elimination of the drug supermarket, and significant reductions in robbery, burglary, and

homicide (Kleiman, 1988). Unfortunately, the program led some dealers and buyers to

find new means for doing business (i.e., displacement) and the program lasted only as

long as the police maintained their heightened presence.

Drug treatment can take a wide variety of forms and may involve greatly

divergent approaches. Research on treatment programs has provided some insights

regarding what is most effective. The National Institute on Drug Abuse (NIDA) offers a

list of 13 principles of effective treatment (see Table 13.4). These principles address the

general issues that can be applied across different programs in ways that respond to the

unique needs of the clients. The wide array of factors contributing to drug use/abuse

means that treatment needs to comprehensively address the problems. Figure 13.2

provides a comprehensive model for drug abuse treatment.

Maintenance programs, a common intervention for addicted individuals, seek to

establish a steady state in which the individual does not experience withdrawal symptoms

when the drug begins to wear off. Consequently, the user will be able to function more

normally and participate in everyday activities without the constant need for the drug

(Stephens, 1987). The most common form of maintenance program is methadone

maintenance. This involves the provision of methadone to heroin/opiate addicts.

Methadone is an oral substitute for heroin, which needs to be taken only once in a 24-

hour period. Over time, those on maintenance will no longer experience the highs and

lows of addiction to other drugs. A primary assumption underlying these programs is that

the patient is unable to function without some form of drug use and that methadone is an

acceptable substitute for other, more damaging drugs.

Closely related to maintenance programs is detoxification. Detoxification uses

drugs in a short-term program of controlled withdrawal. The basic idea is to wean the

client from the addiction with the minimal amount of discomfort and pain. These types of

programs can be found in many hospitals and facilities throughout the country, and target

a wide range of drugs from alcohol to heroin. Detoxification needs to be accompanied

with counseling, referral, or other services to address the social, psychological, and

related issues to the abuse (NIDA, 2012). Anglin and Hser (1990) point out that while

short-term follow-up shows that detoxification is successful at eliminating drug use,

detoxification has not been adequately evaluated over the long-term. Bellis (1981) notes

that detoxification is used by some addicts to reduce the need for massive amounts of a

drug in order to get high. The detoxification simply allows the addict to start the cycle of

addiction over again by achieving a high with smaller amounts of the drug.

Therapeutic communities emphasize providing a supportive, highly structured

atmosphere within which individuals can be helped to alter their personality and develop

social relationships conducive to conforming behavior (Anglin and Hser, 1990). These

residential programs operate as surrogate families for clients. In many cases, therapeutic

communities are run by current or past clients. The daily routine is often very structured

and includes intensive group sessions. Programs may also include education, vocational

training, or mandatory employment (NIDA, 2012). Examples of therapeutic communities

are Synanon, Daytop Village, and Phoenix House.

Outpatient programs often resemble therapeutic communities in most respects

except for the residential component. Individual and group counseling is the cornerstone

of these programs and may involve professionals or simply other group members. Social

skills training, vocational programming, social interaction, referral to other sources of

assistance, and possibly short-term drug maintenance are also common components

(Anglin and Hser, 1990). Alcoholics Anonymous and Narcotics Anonymous are well-

known examples of this type of program. The impact of these programs is highly

questionable. The primary concern with most evaluations is the fact that clients can drop

out at any time. These outpatient programs are more likely to suffer from client mortality

than are other interventions (Anglin and Hser, 1990). Hubbard et al. (1984) note that the

programs have a marginal impact on crime due to their open nature. Evaluations based

only on those individuals who remain in treatment may result in artificially high success

rates. The actual success based on all clients who enter the program, whether they

complete or not, would be less impressive.

Prevention programs that aim to keep individuals from initially using drugs

(primary prevention) usually target juveniles. It is during adolescence that most people

experiment with and enter into patterns of drug use. Prevention modalities cover a range

of issues and approaches, including the dissemination of factual information about drugs

and their consequences, the building of self-esteem, taking responsibility for making

choices, and learning how to handle peer pressure. Most often, prevention programs

incorporate more than one approach. In a similar fashion to treatment programs, NIDA

outlines 16 principles for effective prevention programs (see Table 13.5). The following

discussion is divided into the two most prominently promoted and evaluated prevention

techniques. These are the information/education/knowledge programs and the resistance

skills techniques.

Resistance skills training comes under a variety of names with the most well-

known being the “Just Say No” campaign and the D.A.R.E. (Drug Abuse Resistance

Education) program. While many individuals view this as too simplistic an approach,

resistance skills training involves a set of ideas dealing with recognizing problematic

situations and issues, dealing with peer pressure, recognizing pressure from media

presentations, knowing proper responses to temptations, building self-esteem and

assertiveness, and knowing how and when to take a stand. The implicit assumption in this

type of prevention is that drug use is largely a function of situation and peer involvement.

Youths need to learn how to recognize peer pressure and how to make proper decisions in

the face of that pressure. These programs may also provide factual information about

drugs within the larger discussion of resisting temptations to participate in drug use. A

key part of any information presented deals with the actual levels of use in society,

emphasizing the fact that most individuals do not use illicit drugs. This provides youths

with data that say they are in the majority if they resist drugs.

Students also viewed