Public Safety Assessment


 

Predictors of Crime

INDIVIDUAL RISK FACTORS

COMMUNITY RISK FACTORS

Prior Delinquency

Socioeconomic Status

Academic Attitude & Performance

Community Disorder

Substance Abuse

Community Cohesiveness

Psychological Condition

Social Ties

FAMILY RISK FACTORS

PROTECTIVE FACTORS

Household Composition

Parent-Child Relationship

Parent Characteristics

Family Stress

In order to determine how to prevent crime, it is important to know what causes crime. Research has shown a consistent set of factors that predict the likelihood that individuals will commit crimes (Loeber & Farrington, 1998). These predictors, or risk factors, are broad and include individual, family, and community factors. Studies show that the strength of any given predictor is dependent upon the age of the person exposed to the predictor. (See Table 2.).

Some of the strongest predictors of future offending for youth ages 6 to 11 include prior criminal behavior, substance use, low socioeconomic status, antisocial parents, and aggressive behavior. In contrast, many of the strongest predictors of future offending for youth ages 12 to 14 revolve around peers; these include a lack of strong social ties, antisocial peers, poor school attitude and performance, psychological conditions such as impulsiveness, and prior delinquency. Individuals with these strong risk factors are 5 to 20 times more likely to engage in serious offending than those without these risk factors (Loeber & Farrington, 1998).

Table 2.
Strongest Predictors of Crime by Age of Individual Exposed to Factor and Strength (Ranking) of Factor

Ages 6 to 11

Ages 12 to 14

Rank 1 (Strongest Predictors)

Criminal history
Substance use

Social ties
Antisocial peers

Rank 2

Family poverty
Antisocial parents

Criminal history

Rank 3

Aggression

Aggression
School attitude/performance
Psychological condition
Parent-child relationship
Physical violence

Rank 4

Psychological condition
Parent-child relationship
Social ties
Problem behavior
School attitude/performance
Medical/physical
IQ
Other family characteristics

Antisocial parents
History of crimes against persons
Problem behavior
IQ

Rank 5

Broken home
Abusive parents
Antisocial peers

Broken home
Family poverty
Abusive parents
Other family characteristics
Substance use

Note: Ranking 1 includes the strongest predictors.
Source: Loeber & Farrington, 1998

In addition to the age of the child, the number of risk factors to which an individual is exposed influences the probability that (s)he will engage in criminal behavior. As Table 3 shows, the risk of offending is greatest among individuals with multiple risk factors. In a 1990 study of Denver youth ages 11 to 17, 14% of children with one risk factor committed serious or violent offenses, while 68% of children with four risk factors committed serious or violent offenses.

Table 3.
Percentage of Juveniles with Risk Factors Becoming Type of Offender

Number of Risk Factors

Serious/ Violent Offender

Minor Delinquent

Non-Delinquent

Percent of Juveniles with Risk Factors

0

5%

13%

83%

22%

1

14%

25%

62%

39%

2

26%

32%

43%

27%

3

44%

32%

25%

10%

4

68%

20%

13%

3%

Note: Due to rounding errors, some rows or columns total more than 100%
Source: Loeber & Farrington, 1998

This research suggests that, when determining how to prevent crime, it is important to provide services to youth with multiple risk factors that affect the strongest predictors at the right ages. Because many of the factors that predict criminal behavior are the same factors that predict substance abuse, dropping out of school, early sexual involvement, and teen pregnancy (Loeber & Farrington, 1998), reducing the prevalence of these factors in Travis County would have benefits beyond crime reduction.

RETURN TO TOP

INDIVIDUAL RISK FACTORS

Prior Delinquency

At any age, often the best predictor of future criminal behavior is past criminal behavior (Loeber & Dishion, 1983). In 1999, two of every three juveniles referred to Travis County Juvenile Probation (TCJP) had at least one prior referral to Juvenile Probation.4

For first-time juvenile offenders, the likelihood of re-offending depends upon the type of offense committed. Figure 1 shows that Travis County juveniles who committed the most serious first offenses were the most likely to re-offend. For example, while only 12% of first-time offenders with status offenses re-offended, 23% with felonies re-offended.

As with juveniles, the number of prior offenses is a predictor for adult offending. Most adult offenders have a history of offending that extends back into their youth. Almost all (94%) of the adult offenders incarcerated in the Community Justice Center (Travis County's state jail) from March 1998 to February 1999, for example, had been arrested previously, and 80% had convictions prior to their state jail felony conviction (Travis County Justice and Public Safety, January 2000).

Figure 1.
Re-Offense Rates for First-Time TCJP Offenders by First Offense, CY 1998

Source: Travis County Juvenile Probation Department

RETURN TO TOP

Academic Attitude and Performance

Academic failure, a low commitment to schooling, truancy, and dropping out of school have consistently been shown to predict later criminal behavior. A study by Farrington, for example, showed that 20% of boys with poor performance in elementary school were convicted of violent offenses as adults (Loeber & Farrington, 1998).

Figure 2.
Percent of Students in Attendance, by School District, School Year (SY) 1998


Source: Texas Education Agency, AEIS reports

One measure of a child's commitment to schooling, and a precursor to dropping out of school, is school attendance. Attendance rates at all Travis County school districts improved from SY 1994 to 1997. From SY 1997 to SY 1998, however, attendance rates at Eanes, Lago Vista, and Lake Travis Independent School Districts decreased slightly. In SY 1998, Austin, Del Valle, and Lago Vista Independent School Districts (ISDs) were below the State average on attendance rates. (See Figure 2.) Districts with decreases may benefit from implementing truancy prevention programs in their schools.

In addition to attendance rates, academic success can be measured by scores on standardized tests. The percentage of students passing all parts of the TAAS test increased from SY 1994 to SY 1998 at all Travis County school districts. From SY 1998 to SY 1999, however, the percentage of students passing the TAAS test decreased at Austin, Eanes, Lake Travis, and Manor ISDs. In SY 1999, Austin and Manor ISDs were below the State average in the percentage of students passing the TAAS.

RETURN TO TOP

Substance Use

Studies show that substance use, especially at a young age, is a predictor of future criminal behavior. After years of steady increases, the use of illicit drugs among American teenagers is decreasing (Yahoo! News, November 22, 1999). This national trend is evident in Travis County, as well. From SY 1998 to SY 1999, AISD 4th through 12th grade students reported decreases in recent use of tobacco, alcohol, inhalants, and marijuana. Despite these decreases, nearly one in two AISD 12th graders report recent use of alcohol; over one in three report recent use of tobacco; and over one in four report recent use of marijuana. (See Figure 3.)

80% of the adults in U.S. prisons are locked up because of criminal activity linked to drug and alcohol abuse.

Superville, January 8, 1998

Figure 3.
Percent of AISD Students Reporting Recent Use of Substances by Grade Level and Substance, SY 1999
Source: Austin Independent School District, Title IV Evaluation Report

Because similar data is not available for other school districts, it is difficult to know whether rural areas of Travis County have seen similar decreases in substance use. A study by the National Center on Addiction and Substance Abuse at Columbia University suggests, however, that young teens in rural areas are significantly more likely to smoke, drink, and use illegal drugs than teens in urban areas (Yahoo! News, January 26, 2000). The study found that 8th graders in rural areas were 83% more likely to use crack cocaine, 50% likelier to use cocaine, 34% likelier to smoke marijuana, 29% likelier to drink alcohol, 70% likelier to get drunk, over 200% likelier to smoke cigarettes, and nearly 500% likelier to use smokeless tobacco than were students in urban areas. This study suggests that rural areas of Travis County may especially benefit from increased substance abuse prevention programming.

RETURN TO TOP

Psychological Condition

Research shows that psychological characteristics, such as hyperactivity, attention deficit, impulsivity, risk taking, aggressiveness, antiestablishment and antisocial attitudes and behavior, learning problems, and developmental disabilities are related to criminal behavior (Loeber & Farrington, 1998).

Figure 4.
Percentage of Juveniles Referred to TCJP with Mental Health Assessment Scores, 1999


Source: Travis County Juvenile Probation, SUS scores

The Texas Department of Mental Health and Mental Retardation estimates the annual prevalence of mental illness among adults in Texas to be about 20% (Heikes, February 2000). According to a state study, Texas prisons had about 15,700 inmates and county jails had about 5,900 inmates receiving treatment and/or drugs for mental illness. In addition, 40% of the juveniles in the Texas Youth Commission's residential population in 1998 were assessed as emotionally disturbed. In Travis County, about 15% of county jail inmates have mental health issues (Travis County Sheriff's Office), and about one-half of juveniles referred to Juvenile Probation showed high to moderate need on a mental health assessment (Travis County Juvenile Probation Department).

RETURN TO TOP

Social Ties

As was mentioned previously, social ties are important predictors, especially for young adolescents. Low popularity and being engaged in few positive social activities are predictors of criminal behavior (Loeber & Farrington, 1998). In addition, a relationship exists between having delinquent or antisocial friends or siblings during adolescence and later violence. Not surprisingly, gang membership is also a predictor of criminal behavior (Loeber & Farrington, 1998).

Table 4.
Number of Gang Members Referred to TCJP by Zip Code of Residence, CY 1999

Zip Code

Gang Members Referred

Percentage of Total

78702

35

11%

78744

34

11%

78723

32

10%

78704

27

9%

78741

25

8%

Others

163

51%

Total

316

100%

Source: Travis County Juvenile Probation

Data from Juvenile Probation shows that the percentage of juveniles referred that were affiliated with gangs decreased from 14% in CY 1998 to 9% in CY 1999. As Table 4 shows, nearly one in every three gang members referred to Juvenile Probation in 1999 lived in the 78702, 78723, and 78744 zip codes. This data suggests that if gang reduction programs are implemented, they should be directed to these areas of the County.

RETURN TO TOP

FAMILY RISK FACTORS

Family functioning factors are among the strongest predictors of crime, especially for younger children. Household composition, poor parental monitoring, poor discipline, a family history of criminality, parental conflict, a lack of family cohesion, and low socioeconomic status are general risk variables for criminal behavior.

Household Composition

Several studies have concluded that children from single-parent families or stepfamilies are more likely to engage in criminal behavior than are those from two-parent families. Nationally, the percentage of children living in two-parent homes declined from 85% in 1970 to 68% in 1997 (Snyder & Sickmund, September 1999). In the general Travis County population, 22% of children live in single-parent families (Texas Association of Child Care Resource and Referral Agencies, 1999). However, 55% of juveniles referred to Juvenile Probation in 1999 lived in single-parent households. (See Figure 5.)

Some research suggests that low socioeconomic status due to a single income, rather than single parenthood itself, may be the primary reason for the increased risk of offending for children from single parent families. In 1997, for example, almost one-half of American children living in female-headed households lived in poverty (Snyder & Sickmund, September 1999).

In addition to being a risk factor for committing crime, household composition is related to child victimization. The National Incidence Study of Child Abuse and Neglect reported that children of single parents were twice as likely to be neglected and marginally more likely to be abused as children living with both parents. Again, low socioeconomic status may be the primary reason for the increased risk.

Figure 5.
Referrals to Travis County Juvenile Probation by Household Type, CY 1999


Source: Travis County Juvenile Probation

The study also reported that children living in large families with four or more children were more than twice as likely to be neglected than those living in smaller families with two or three children (Snyder & Sickmund, September 1999). Again, this increased risk may be due to socioeconomic issues. Currently, the average family size in Travis County is 3.1 people (Texas Association of Child Care Resource and Referral Agencies, 1999). National data shows that, in the last 30 years, the percentage of small families (one or two people) has been increasing, while the percentage of large families (five or more people) has been decreasing (United States Census Bureau).

In addition to family type and size, teen parenting is a predictor of crime, not only for the child, but also for the parents, particularly the father. According to the Office of Juvenile Justice and Delinquency Prevention (OJJDP), teenage fathers are more likely than other youth to commit delinquent acts, be involved in drug dealing, use alcohol, and drop out of school (Office of Justice Programs, February 3, 2000).

The number of Travis County teens that became pregnant decreased by 11.6%, from 942 in 1997 to 833 in 1998. Despite the decrease, Travis County's teen pregnancy rate is among the highest in the state (Texas Department of Health). (See Figure 6.)

Figure 6.
Number of Pregnant Teens (Ages 13 to 17) per 1,000 by County, 1998


Source: Texas Department of Health, Vital Statistics Bureau

RETURN TO TOP

Parent-Child Relationship

According to social control theory, children that bond to their families are less likely to engage in delinquent behavior. Not surprisingly then, disruptions of the parent-child relationship are often related to criminal behavior. The degree and quality to which parents interact with their children and are involved in their lives have also been shown to predict criminal behavior.

Table 5.
Likelihood of Arrest by Type of Child Abuse/Neglect Experienced

Type of Abuses

Juvenile Arrest

Adult Arrest

Any Sexual Abuse

22.2%

20.3%

Any Physical Abuse

19.9%

27.4%

Any Neglect

28.4%

30.7%

All Cases of Abuse & Neglect

26.0%

28.6%

Control Group

16.8%

21.0%

Source: National Institute of Justice, March 1995

Along these lines, research indicates that abuse or neglect during childhood increases the risk of later criminal behavior. The National Institute of Justice (March, 1995) conducted a study on victims of childhood sexual abuse and later criminal consequences and found that children who were abused or neglected were more likely than people who were not maltreated to be arrested later in life. As Table 5 shows, the likelihood of arrest was especially high in cases of neglect.

Travis County child abuse and neglect rates remained relatively stable from 1996 to 1998, with about 1% of Travis County children being confirmed victims of child abuse in 1998. However, according to the Texas Department of Protective and Regulatory Services (TDPRS), in 1998, an estimated 27,830 children in Travis County, or one in every six or seven children, were considered by TDPRS to be at risk for child abuse and/or neglect. In addition, as Figure 7 shows, of the seven highest child populated areas in the state, Travis County has the highest rate of confirmed victims per 1,000 children.

Figure 7.
Confirmed Victims of Child Abuse/Neglect Per 1,000 Children, by County, FY 1998


Source: Texas Department of Protective and Regulatory Services

Abuse or neglect is sometimes so severe that it results in the death of a child. Homes in which children die from abuse or neglect often include (Texas Department of Protective and Regulatory Services):

  • parents who are overwhelmed by care-giving responsibilities, and exhibit a low tolerance for stress and hold unrealistic expectations for their child's behavior,
  • emotionally immature parents, who are indifferent toward their child and put adult needs first and repeatedly make choices that jeopardize their child's safety,
  • single mothers who make poor choices about their male partners who consequently engage in behavior leading to the child's death,
  • parents and other household members with a history of violence and drug and alcohol abuse,
  • very young children or children who are otherwise vulnerable due to physical or mental impairments, and
  • a chaotic home invironment with minimal resources and poor support systems.

In addition to abuse or neglect, a low level of supervision, inconsistent discipline, and a low level of warmth or a negative attitude towards a child increase the risk of engaging in criminal behavior. Because these predictors are difficult to measure, it is not currently possible to determine to what extent these predictors are present in Travis County. However, research suggests that parenting classes are an effective tool for addressing these family risk factors.

RETURN TO TOP

Parent Characteristics

In addition to the way a parent interacts with or disciplines a child, the attitudes and behaviors of parents can affect the likelihood of criminal behavior. Criminal activity, negative attitudes, and psychopathology of parents contribute to criminal behavior of their children.

Parental attitudes favorable to violence, for example, increase a child's risk of engaging in criminal behavior. A study of 8,000 middle school students in a large urban area found the strongest predictor of aggressive behavior was the perception of parents' feelings about fighting. Middle school students who say that their parents are against fighting are significantly less likely to fight or behave aggressively (Yahoo! News, November 19, 1999).

Similarly, parents who engage in criminal behaviors are more likely to have children who are delinquent (Yahoo! News, December 20, 1999). Between January 1998 to July 1999, approximately 60% of the female and 46% of the male intakes into the Travis County Community Justice Center had children (Wilkinson, et. al., January 2000).

National statistics also show that children who learn about the risks of drug use from their parents are 36% less likely to smoke marijuana, 50% less likely to use inhalants, 56% less likely to use cocaine, and 65% less likely to use LSD than children whose parents do not teach them about the dangers of drugs (Office of National Drug Control Policy).

RETURN TO TOP

Family Stress

High family stress and marital discord are also predictors of criminal behavior. Divorce is often a result of years of marital discord. Despite the increase in the Travis County population, the number of divorces decreased by 4% from 1997 to 1998.

Figure 8.
Reported Incidents of Family Violence in Travis County, 1996 to 1998


Source: Texas Department of Public Safety

When marital discord escalates into violence, children, as well as spouses, suffer. Studies show that children who witness violence are more likely to display future aggressive or violent behavior than are children who do not (Loeber & Farrington, 1998). Similarly, children who witness domestic violence are more likely to become violent offenders than are children who do not. In studies of men's batterer treatment programs, 63%-75% of men had either experienced physical abuse or had witnessed physical abuse involving their parents (Maryland Family Violence Coalition, 1991).

Figure 8 shows the reported incidents of family violence in Travis County. After a 7% increase from 1996 to 1997, reports of domestic violence decreased by 16% from 1997 to 1998.

RETURN TO TOP

COMMUNITY RISK FACTORS

Socioeconomic Status

Community-level analyses have linked neighborhood poverty to violent crime (Loeber & Farrington, 1998). Farrington found that low socioeconomic status as a child was predictive of teen violence. While 8.8% of the boys in his study that did not suffer from poverty were convicted of violent offenses, 23% of the boys living in poverty were convicted of violent offenses.

Travis County, like the rest of the nation, is currently experiencing an economic boom. The percentage of Travis County residents living in households with incomes below the federal poverty threshold decreased from 15.4% in 1996 to 13.1% in 1999. In addition, the Travis County unemployment rate has been steadily decreasing for the last decade and is now a low 1.9%, the second lowest rate in Texas (Texas Workforce Commission, April 2000).

Figure 9.
Percent of Travis County Students Who Are Economically Disadvantaged, by School District, SY 1999


Source: Texas Education Agency, AEIS reports

Despite the good economy, not all Travis County residents are experiencing prosperity. It is estimated that over 87,000 residents live in poverty (United States Census Bureau, February 1999). In addition, in SY 1999, approximately two in three Del Valle Independent School District (ISD) students and one in two Austin ISD and Manor ISD students were economically disadvantaged. (See Figure 9.)

Likewise, while the unemployment rate of the general Travis County population is around 2%, the unemployment rate of parents of juveniles referred to Travis County Juvenile Probation (TCJP) is 14%. (See Figure 10.)

Figure 10.
Juveniles Referred to TCJP by Parents' Employment Status, 1999


Source: Travis County Juvenile Probation

RETURN TO TOP

Community Disorder

For decades, research has confirmed what most people intuitively know: there is a relationship between crime, fear of crime, and disorderly conditions in communities. As far back as 1967, research presented to the President's Commission on Law Enforcement clearly stated this conclusion (Biderman, et. al, 1967). This relationship between disorder and crime is known as the "broken windows" theory and is described in Figure 11.

Following the theory, "broken windows" can refer to both physical decay and social incivilities. Examples of physical decay include abandoned buildings, graffiti, litter, vacant and trash-filled lots, unkempt yards and housing exteriors, and abandoned cars. Examples of social incivilities include public drinking or drunkenness, rowdy and unsupervised teen groups, sexual harassment on the street, and fighting among neighbors.

Figure 11.
"Broken Windows" Theory



Source: Taylor & Harrell, January 1996

One of the most readily identifiable and quantifiable indicators of physical decay and neglect is graffiti. Since 1995, the City of Austin has operated a Graffiti Abatement Program that has maintained both a hotline for citizen reports of graffiti and patrols by clean-up crews to remove graffiti. From a peak of close to 5,000 in 1997-98, reported graffiti has declined the past two years as the backlog of existing graffiti was finally eliminated and the rate of new "tagging" has declined. Over the life of the program, response time (from report to cleanup) has improved from over a month (38 days in 1996-97) to under a week (5.5 days in 1999-00) (City of Austin, February 2000).

In general, Austin residents surveyed in 1999 indicated a high level of satisfaction with the physical appearance of their neighborhoods. Of all respondents, 81% had a favorable opinion of their neighborhood appearance. Even the least positive response, from Southeast Austin, was 65% positive (City of Austin, 1999).

RETURN TO TOP

Community Cohesiveness

While the discussion of community disorder focuses largely on the tangible and visible indicators of community, another closely related line of research focuses on the intangible indicators of community cohesion. These intangibles capture the feelings of safety and security that result from "collective efficacy", such as mutual trust, a willingness to intervene in the supervision of children, a willingness to help neighbors, and the general maintenance of public order. When the level of collective efficacy is low, feelings of danger and fear are likely to rise. This approach has been used to explain variation in violence and criminal victimization across otherwise comparable neighborhoods (Sampson, et. al., 1997).

According to the Central Texas Indicators 2000 report, 72% of Travis County residents reported being "very" or "somewhat comfortable" asking a nearby neighbor for help or a small favor. Analyses showed a positive relationship between the number of years that the individual lived in the same neighborhood and comfort levels (City of Austin, March 2000).

Figure 12.
Percent of Austin Residents Who Felt Safe Walking Alone, 1995 to 1999


Source: City of Austin, November 1999

The citizen survey conducted by the City of Austin in early 1999 shows generally high feelings of safety across the community. (See Figure 12.) The data does, however, indicate some notable exceptions to this finding. First, there are marked differences in feelings of security during the day compared to after dark, especially in downtown Austin.

Figure 13.
Percent of Residents Who Rated Neighborhood Safety as Favorable, 1999


Note: Rating options for neighborhood safety included "favorable", "unfavorable", and "neutral".
Source: City of Austin, November 1999

In addition, perceptions of general neighborhood safety vary depending upon the location of residence. While almost all residents living in western areas of Austin reported favorable levels of neighborhood safety, this was not true for residents of eastern areas of the city (City of Austin, November 1999). (See Figure 13.)

RETURN TO TOP

PROTECTIVE FACTORS

In addition to negative factors that put individuals at risk for criminal behavior, there are protective factors, or assets, that, if in place, act as buffers and help reduce the likelihood of criminal behavior. Research by the Search Institute, among others, identified protective factors that have a positive affect on the lives of youth. These factors include characteristics of the child, their family, and their community and are listed in Table 6.

Table 6.
Internal and External Protective Factors

Category

Asset

Description

Internal Assets

Commitment to Learning

Motivation to achieve; Engagement in school; Involvement in homework, Bonding to school, Reading for pleasure

Positive Values

Caring, honest, & responsible; Believes in equality & social justice, Shows integrity & restraint

Social Competencies

Planning & decision making skills, Interpersonal & cultural competence, Resistance skills, Peaceful conflict resolution

Positive Identity

Personal power, Self-esteem, Sense of purpose, Positive view of personal future

External Assets

Support

Support from family & other adults, Caring neighborhood & school, Parent involvement in schooling, Positive family communications

Empowerment

Youth as resources to community, Service to others

Boundaries & Expectations

Family, school, & neighborhood boundaries, Adult role models, High expectations from others, Positive peer influence

Constructive Use of Time

Creative activities, Youth programs, Spiritual community, Time at home

Source: Search Institute, 1997

Similarly, the Travis County Community Justice Council and the University of Texas School of Social Work helped local neighborhoods identify protective factors in their communities. They held focus groups in various neighborhoods in Travis County to determine what makes neighborhoods safe and healthy. As Table 7 shows, community members listed community cohesiveness, positive relationships, and community order as the top indicators of safe and healthy neighborhoods.

Table 7.
Top Indicators of a Safe and Healthy Neighborhood by Neighborhood

Indicators

Northeast Austin

Southeast Austin

Dawson/ S. Congress

Del Valle Community

Knowing your neighbors

X

X

X

X

Little street crime

X

X

X

Diversity of residents

X

X

X

Businesses to support neighborhood

X

X

X

Activities for youth and elderly; prevention programs

X

X

X

Good schools in the neighborhood

X

X

Opportunities for learning

X

X

Community pride; clean neighborhoods

X

X

Relationship with police

X

X

Government investment in community: access to services, library, transportation

X

X

Most residents employed or retired

X

Parents active in schools

X

Kids and people visible in yards/outdoors

X

Natural areas preserved

X

Source: Travis County Community Justice Council

Assessment Home


4. While one in three juveniles had at least one prior referrals, on average, only 30% of juveniles referred to Juvenile Probation offend more than once in any given calendar year (Travis County Juvenile Probation)