The Unpredictability of Murder

• Author: Matt DeLisi,Stephanie M. Cardwell,Alex R. Piquero
• Published date: 01 January 2016
• Date: 01 January 2016
• DOI: http://doi.org/10.1177/1541204014551805
• Subject Matter: Articles

Article

The Unpredictability of

Murder: Juvenile Homicide

in the Pathways to

Desistance Study

Matt DeLisi

1

, Alex R. Piquero

2,3

, and Stephanie M. Cardwell

2

Abstract

There is minimal research that has investigated the characteristics distinguishing youth who commit

murder to otherjuvenile offenders. Ofthe research that has been done,scholars have identifieda wide

variety of factors that distinguish these offenders, including poor family environments, emotional and

socialproblems, poor mental health,and behavioral disorders.Using data fromPathways to Desistance,

a study of1,354 serious youthful offenders,we examined how8 demographic characteristicsand 35 risk

factorsdistinguish betweenthose youth who werecharged with some type ofhomicide and those youth

who werenot charged with any type of homicide.We find that only 18 (1.33%)youth were charged with

a homicide offense. Among the predictors, age, intelligence quotient (IQ), exposure to violence, per-

ceptionsof community disorder,and prevalence ofgun carrying are significantlydifferent acrossthe two

groups. Results from a rare-events logistic regression that simultaneously examined the relationship

betweenthese five risk factors and their abilityto distinguish between thetwo groups indicate that only

lower IQ and a greater exposure to violence were significant. Finally, a higher number of risk factors

were associatedwith a higher likelihood that youth would be charged with homicide.

Keywords

juvenile homicide, murder, criminal career; Pathways to Desistance, crime prediction

Introduction

In the wake of a spate of homicides involving juvenile victims and, often, juvenile offenders, First

Lady Michelle Obama has urged policy makers, law enforcement personnel, and community leaders

1

Department of Sociology, and Faculty Affiliate with the Center for the Study of Violence at Iowa State University,

Ames, IA, USA

2

University of Texas at Dallas, Dallas, TX, USA

3

Key Centre for Ethics, Law, Justice, and Governance, Griffith University, Queensland, Australia

Corresponding Author:

Matt DeLisi, Department of Sociology, and Faculty Affiliate with the Center for the Study of Violence at Iowa State

University, 203A East Hall, Ames, IA 50011, USA.

Email: delisi@iastate.edu

Youth Violence and JuvenileJustice

2016, Vol. 14(1) 26-42

ªThe Author(s) 2014

Reprints and permission:

sagepub.com/journalsPermissions.nav

DOI: 10.1177/1541204014551805

yvj.sagepub.com

to allocate resources to better understand the predictors of juvenile homicide (Lee, 2013). Although

recent years have seen sizeable reductions in homicide generally and juvenile homicide specifically,

it is also true that young age is one of the most robust predictors of homicide offending and victi-

mization (Farrington, Loeber, Stallings, & Homish, 2012; Fox & Piquero, 2003; Heide, 2003;

Loeber & Farrington, 2011) especially among delinquent youth (Teplin, Abram, McClelland, &

Mileusnic, 2005; Teplin et al., 2014). For instance, between 1992 and 2011, the U.S. homicide rate

declined by almost 50%, from 9.3 homicides per 100,000 U.S. residents in 1992 to 4.7 in 2011

(Smith & Cooper, 2013). The homicide rate is highest among males, Blacks, and young adults and

firearms account for the majority of them. With respect to age, Smith and Cooper (2013, p. 5)

reported that ‘‘From 2002 to 2011, young adults ages 18 to 24 had the highest homicide rate of any

age group. Young adults also experienced the largest homicide rate decline (down 22%), from 15.2

per 100,000 in 2002 to 11.9 in 2011. Persons age 65 or older (down 17%) and youth ages 12 to 17

(down 19%) also experienced large declines in their homicide rates from 2002 to 2011.’’

These trends are important but equally important is identifying the risk factors that are associated

with homicide perpetration. Although popular press and public discourse are replete with anecdotes

and commentary about the sole reason why some persons kill (see Zimring, 2013), the research evi-

dence does not exist to back up such claims. Two principal reasons underlie this unfortunate happen-

stance. First, empirical research into the correlates and causes of homicide perpetration are very rare,

largely because of the lack of such data even in large data sets of offenders (Cornell, Benedek, &

Benedek, 1987; DeLisi, 2001; DeLisi & Scherer, 2006; Woodworth, Agar, & Coupland, 2013).

Second, many studies of criminal activity collect both self-report and official data regarding vio-

lence, but most of those investigations reveal that violence, especially serious violence (and most

certainly homicide perpetration), is also very rare (Elliott, Huizinga, & Morse, 1986; Farrington,

1989; Piquero, 2000; Piquero, Farrington, & Blumstein, 2003, 2007; Reiss & Roth, 1993, p. 374).

In fact, there have only been a handful of investigations looking into the characteristics that may

distinguish persons who committed murder compared to persons who may have committed other

crimes but not murder. These are reviewed next.

1

Predictors of Juvenile Homicide Offending

A diminutive literature has investigated predictors of juvenile homicide with comparisons to other

antisocial youth or violent delinquents. In their case control study comparing 21 male juveniles con-

victed of homicide to 21 convicted burglars incarcerated in the United Kingdom, Hill-Smith, Hugo,

Hughes, Fonagy, and Hartman (2002) identified several risk factors that differentiated murderers

from their antisocial peers. Youth convicted of murder were more likely to have been reared by emo-

tionally and physically harsh parents, to have mothers and fathers who had been arrested, and to have

had generally greater and more extensive educational and school problems. Like their antisocial

peers, juvenile homicide offenders led childhoods of abuse and deprivation and faced considerable

obstacles in school.

Several studies using data from the high-risk Pittsburgh Youth Study (PYS) (Farrington, Loeber,

& Berg, 2012; Farrington, Loeber, Stallings, & Homish, 2012; Loeber & Farrington, 2011; Loeber

et al., 2005) have shown that several background factors are associated with homicide offending sta-

tus independently of their association with delinquency and violent delinquency. For instance,

Loeber and his colleagues (2005) compared 33 juvenile homicide offenders, 193 youth who were

convicted of other serious violent crimes, and nearly 500 youth who self-reported their violent delin-

quency. They found that youth with four or more risk factors for homicide were over 14 times more

likely to later commit homicide than their antisocial peers with fewer than four risk factors. Across

their studies, a me´lange of risk factors have been shown to predict later homicide. These include

weapon carrying, living in a bad neighborhood, having a young mother, having a positive attitude

DeLisi et al. 27