Does Contact with the Justice System Deter or Promote Future Delinquency? Results From a Longitudinal Study of British Adolescent Twins
Motz, Barnes, Caspi, Arseneault, Cullen, Houts, Wertz & Moffitt, Criminology, 2020
The authors note there is contradictory research in regards to the effects of the criminal justice system on those introduced to it. A labeling tradition suggests that youths introduced to the criminal justice system’s formal punishments will exhibit the opposite of the intended behavior and engage in future misbehavior, especially if considering claims of a criminogenic effect from the criminal justice system. A deterrence perspective suggests that youths will be deterred from future criminal activity by these formal punishments. In an attempt to address this issue, the authors conducted a longitudinal study of 901 British twins pairs (the use of twins can eliminate any confounding factors like genetics and parental and home differences that may influence study outcomes and allow focus on the environmental factors) to see if different forms of contact with the juvenile justice system earlier in life was associated with an increase of decrease in delinquent behavior. The outcome variable was level of delinquency at age 18, which comprised a number of delinquent acts as self-reported by the participants, with key independent variables including whether they had been in police custody, jail, or prison, whether they had a criminal record or criminal cautions (a legal warning with an admission of guilt that differs from an actual conviction for minor crimes) and whether they had been issued an Antisocial Behavior Order (ASBO). ASBOs, which were introduced in England in 1999, are defined by the Home Office as “civil orders that exist to protect the public from behavior that causes or is likely to cause harassment, alarm or distress. An order contains conditions prohibiting the offender from specific anti-social acts or entering defined areas and is effective for a minimum of two years. The orders are not criminal penalties and are not intended to punish the offender.”
While the twin design rules out many familial sources of confounding, the authors also controlled for other confounding influences that are not shared by twins, such as cognitive and behavioral differences. These included different specific measures of self-reported delinquency at age 12, externalizing problems at age 12, evidence of low self-control up to age 10, cognitive ability at age 12, educational achievement at age 16,and first born twin. The authors used regression modeling utilizing the three independent variables, and in final steps included the family fixed effects.
In analysis, spending the night in jail was shown to significantly increase delinquency at age 18, that effect remained once the covariates like education, cognitive ability and low self-control were included though they did reduce the size of the effect. Similarly, adding the family fixed variable continued to demonstrate a significant but smaller increase in delinquency after spending the night incarcerated. The authors also performed regression on the residuals of the models to ensure that the other non-locked up twin did not experience a deterrent effect from their sibling locked up but the results showed a deterrent effect wasn’t present and that the labeling perspective was supported.
Results followed a similar pattern with ASBOs and criminal record/cautions in that contact in these ways in the juvenile justice system increased delinquency at age 18 even after accounting for the family fixed effects and other criminogenic variables like low self-control, earlier delinquency and cognitive level. However, in regards to criminal records and cautions, the effect sizes were much smaller than observed in the other two models, and the effect on monozygotic (identical) twins in the full fixed effect model did not reach a level of statistical significance. Though the authors didn’t determine if there was a statistically significant difference, interestingly, the effect of increased delinquency was greater in monozygotic twins pairs than in dizygotic (fraternal) twin pairs.
The authors conclude their study supports a labeling theory perspective over a deterrent theory perspective in juvenile justice in that contact with the justice system in these ways promotes misbehavior and results in increased delinquency later in life. The authors note the study suggests future research in this debate could examine whether the “dose” of these sanctions influences level of delinquency, whether the effects are crime and offense dependent, and make attempts to dissect the criminal justice system to determine what mediating mechanisms in these contacts promote delinquency.
Motz, R. T., Barnes, J. C., Caspi, A., Arseneault, L., Cullen, F. T., Houts, R., … & Moffitt, T. E. (2020). Does contact with the justice system deter or promote future delinquency? Results from a longitudinal study of British adolescent twins. Criminology, 58(2), 307-335.
Motz, et al suggest the negative experience of criminal sanctions can generate negative outcomes in the form of increased delinquency. However, Bateson, McManus, and Johnson examine how using the negative experience of childhood trauma to predict negative outcomes in later life can be problematic.
Understanding the Use, and Misuse, of Adverse Childhood Experiences (ACEs) in Trauma-Informed Policing
Bateson, McManus, & Johnson, The Police Journal, 2019
The authors discuss the use of Adverse Childhood Experience (ACE) scores in policing as a component of a trauma informed approach but the authors caution the scores may be misused in determining when it’s appropriate to intervene. ACE scores are determined by positive responses to 10 adverse experiences that include physical, emotional and physical abuse, physical and emotional neglect, and household dysfunction that includes parental substance abuse, mother subject to physical abuse, parental mental illness, parental incarceration, parental separation/divorce. Research has demonstrated a strong and almost linear, as well as additive, effect of ACEs on negative life course outcomes, including a host of health problems, mental and emotional issues, relationship problems including domestic abuse, and substance abuse, as well as poor educational, employment, and life satisfaction outcomes. Recent research has indicated that about 50% of the population have experienced at least one ACE and about 10-20% have experienced four or more.
The authors note this information can assist in developing and presenting a trauma informed approach, as they stated “there is potential to develop a common language and understanding about trauma informed practice across different workforces. For professionals, it encourages a shift in thinking from “what’s wrong with you?” to “what’s happened to you?” and for service users a shift from “there’s something wrong about me” to “I’m not a bad person, I’m like this because bad things happened to me”.” ACE scores can also assist in policing by identifying to law enforcement individuals who may be vulnerable to negative outcomes, and can assist in the decision in how to relate to and intervene for that individual. However, the simple process of producing an ACE score may contribute to an oversimplification in the decision of intervention. As the ACE score is a list of different kind of diverse experiences an individual has encountered, it doesn’t measure the duration, severity, or magnitude. A low score could mask the level of trauma an individual has actually experienced from the event. The authors caution ACE scores, while suggestive. cannot assess risk for committing an offense or experiencing other negative life course outcomes, nor can it be used to assess current needs, as the ACE is retrospective and should not be used in place of careful assessment of the individual. Research has shown that a number of other factors not included in the ACE score like poverty, housing, social isolation and discrimination, can significantly influence adult outcomes. While research has demonstrated that early trauma can result in negative psychological adult outcomes, and that different kinds of trauma can have different psychological effect, the exact relationship between different kinds of trauma on different physical and behavioral outcomes remains unclear. Complicating this is research that demonstrates that children can vary in their resiliency, with some being more susceptible to negatives outcomes after suffering ACEs compared to other children.
The authors believe there is a potential for misuse in using ACE scores a screening device. Practitioners should understand that individual ACEs should be weighed differently as some may have more of a negative impact than others, as well as that impact differing dependent on the developmental stage of the child. While the subscales may need to be weighed differently in different cases, determining whether ACE scores are valid and reliable means of assessing future needs is lacking evidence. One main drawback, because they rely solely on self-reports, is the possibility of faulty recollection or assessment and it may be an unreliable predictive tool of criminal justice outcomes. and while the use of data available to the police such physical abuse or parental separation and substance abuse may provide confirmatory evidence, other ACEs like emotional neglect may be more difficult to determine and may require follow-up investigation
The authors conclude that “ACEs may provide an easily understandable framework to identify vulnerable adults and children, which could help to develop trauma informed practice and responses, ultimately safeguarding children from harm. In addition, it has the potential to enable a common language and understanding across different workforces nationally and internationally. The advantages and enthusiasm around ACEs offer great opportunities to drive the prevention, early intervention and trauma-informed agendas. However, there are valid concerns about the limited research base being misunderstood and yet translated into practice…hence, use of ACEs questionnaires as a checklist, using ACE scores or thresholds in practice are not yet supported by evidence.” The authors encourage further inquiry into issues of validity and reliability, consent and information sharing of the data and appropriate training and supervision of practitioners, while holding the view that knowledge of ACEs does not need fuel a fatalistic or deterministic view because childhood adversity does not always result in negative outcomes.
Bateson, K., McManus, M., & Johnson, G. (2019). Understanding the use, and misuse, of Adverse Childhood Experiences (ACEs) in trauma-informed policing. The Police Journal, 0032258X19841409.
Understanding the factors that signal negative life outcomes is important as the greater our knowledge of the influences that push young people and adults toward criminal behavior, the more that can be done to address it. Examining the circumstances around juvenile homicides can also lead to a better understanding of the influences of race, region, and different types of social conflict on who might perpetrate these murders.
Racial Differences Among Juvenile Homicide Offenders: An Empirical Analysis of 37 Years of U.S. Arrest Data
Heide, Michel, Cochran, & Khachaturian, Journal of Interpersonal Violence, 2017
The authors state that while initially concerning as youth murder offenders rose to prominence during a couple of periods in the twentieth century, youth and adult murder offenses are declining but interest in youthful offenders continues, especially in potential differences by offender race. However, most research on juvenile offender race had been limited to White and Black offenders as the number of Asian/Pacific Islander and American Indian/Alaskan Native offenders have been very small, For example in 2012, of Juvenile Homicide Offenders (JHOs), 98% were either White (47.2%) or Black (58.8%), numbering in the hundreds, while there were only 10 American Indians and 1 Asian JHOs. To enhance the ability to examine racial differences, the author use 37 years (1976-2012), divided into three distinct time periods including a pre and post period of an epidemic period of juvenile homicides, worth of JHO (aged 6 to 17) arrests from the UCR’s Supplemental Homicide Reports and asked three research questions: 1) Did the offender and offense characteristics of all JHOs arrested during the study period vary by race and 2) did the characteristics of victims, weapons used, crime circumstances, and offender count in incidents in which JHOs killed single victims differ across racial groups, and 3) are offender, victim, and offense characteristics predictive of racial classification?
Chi square was used in the bivariate analyses to determine significant relationships. The authors found a number of relationships that were both statistically significant and had a large enough effect size to be meaningful (only the percentages for the significant different racial categories are shown). Results demonstrated that there were significant differences, and medium effect sizes, in the race more likely to be arrested by region. Black youth (43.7%) were significantly more likely to be arrested for homicide in the South than the other racial groups. AI/AN juveniles (50.6%) were significantly more likely than White and Black youth to be arrested for murder in the West. A/PI youth (63.7%) were significantly more likely than all three other racial groups to be arrested in the West. AI/AN juveniles (2.6%) were significantly less likely than Black, White, or A/PI youth to be arrested for murder in the Northeast.
There was also a small but significant effect size on arrest location by race. In general, nearly 80% of JHOs were arrested in large cities (59.5%) or suburban areas (19.9%) while the remaining 20% were arrested in small cities (12.6%) or rural areas (8.1%). By race Black and A/PI youth (68.7% and 66.9%) were significantly more likely to be arrested in large cities than their White and AI/AN counterparts (47.8% and 25.9%). White youth (26.7%) were significantly more likely than the other three racial groups to be arrested in suburban areas. AI/AN juveniles were significantly more likely to be arrested in rural areas (42.7%) compared with White youth (12.6%), and Black and A/PI youth (4.2% and 3.4%).
Offenders varied significantly by race within the different time period within the study frame albeit to a small effect. Black youth became significantly more involved in homicide arrest during the 37-year time frame than White youth, although both increased across the three time periods. A/PI juveniles stood apart from the other three racial groups in significant way, notably, while their involvement in the pre-epidemic period (6.5%) was far lower than the three other groups, their arrests increased dramatically during the post-epidemic period compared to the other races. Nearly 64% of A/PI JHOs were arrested between 1994 and 2012, compared with less than 50% of JHOs in the three other racial groups.
There were also large significant racial differences found in comparing JHO race to the victim race. While in general, almost 97% of JHOs were arrested for killing White or Black victims, each of the four racial groups was significantly more likely to kill members from their own racial group. However, there were significant differences found for each racial group with respect to killing White victims. While 90.4% of the victims of White JHOs were also White, 48.3% of American Indian/Alaskan Natives JHOs killed White victims, while for Asian and Pacific Islanders, 32.8% of their victims were White, and for Black JHOs, 22.6% of their victims were white.
The relationship between JHOs and the offense characteristics also varied significantly, although to a small effect, in three areas, victim-offender relationship, weapon used, and circumstances of the offense. While the typical victims of JHOs in general were acquaintances (46.0%) or strangers (35.9%), the authors’ data showed White juveniles (8.1%) were significantly more likely than the other three racial groups to be arrested for killing parents while A/PI and Black JHOs were significantly more likely to be arrested for killing strangers (39.5% and 39.3%) than White and AI/AN JHOs (31.6% and 27.3%).
For JHOs, firearms (69.7%) or knives (15.9%) were most commonly used to kill their victims but the four racial groups differed significantly from each other on the use of guns with Black JHOs being most likely to use guns (76.1%) followed by A/PI JHOs (71.5%), then by White JHOs (61.5%), and finally by AI/AN (42.9%). In contrast, AI/AN (28.0%) were significantly more likely to use knives than White JHOs (21.3%), A/PI (13.6%), and Black JHOs (11.7%). AI/AN were also significantly more than twice as likely to use personal weapons (16.7%) than the other racial groups (all 7% or less).
A large majority (85%) of JHOs were arrested in crime-related (33.8%), conflict-related (35.7%), or gang-related killings (14.5%) but different races predominated in these three homicide circumstances. Black JHOs (39.7%) were significantly more likely to be involved in crime-related homicides relative to the other three racial groups. In contrast, AI/AN JHOs (45.6%) were significantly more likely to be involved in conflict-related killings than the other racial groups, while A/PI JHOs (40.6%) were significantly more likely to be arrested for gang-related killings than their White (19.9%), Black (9.6%), or AI/AN (6.0%) counterparts.
The authors put the significant variables of region, location, victim race, victim-offender relationship, weapon and homicide circumstances, from the bivariate analyses into three homicide circumstances logistic regression models. Each of the four racial groups had significant difference from the others in the different homicide variables and only those significant relationships are described below.
Whites
Regarding differences involving White offenders, JHOs who killed a White victim were 36 times more likely to be White than Black, 5% more likely to be White than Asian/Pacific Islander and 7% more likely to be White than AI/AN. JHOs who killed a family member were twice as likely to be White as Black and 39% more likely to be White compared to A/PI. White JHOs were also 33% more likely than A/PI JHOs and 62% more likely than AI/AN JHOs to be involved in a homicide in the South. White offenders were also significantly more likely (24%) than AI/AN to be involved in a homicide in a large city. White offenders were also 52% more likely to use a gun in a homicide compared to AI/AN. Gang-related juvenile homicides were almost 3.5 times as likely to involve a White offender than a Black offender and 25% more likely than an AI/AN offender. For conflict-related juvenile homicides, White offenders were about 1.5 times more likely to be involved than a Black offender and 63% more likely than A/PI offenders.
Blacks
Significant differences also existed between Blacks and other races in regard to their homicides. Juvenile homicides committed in the South were 57% more likely with Blacks than Whites, 19% more likely than Asian/Pacific Islanders, and 18% more likely than AI/AN and those committed in large cities were 67% , 65%, and 35% more likely to involve a Black offender than White, Asian/Pacific Islander, and American Indian/Native Alaskan offenders, respectively. Also statistically significant, juvenile murders involving a gun were 67% more likely to be committed by a Black offender than by a White offender and 35% more likely than an AI/AN offender. Finally crime-related juvenile homicide incidents were 30% more likely to involve a Black offender than a White offender, twice as likely to involve a Black offender than an AI/AN offender, and 35% more likely than A/PI JHOs.
Asian
White victims were 1.9 times more likely to be killed by Asian/Pacific Islander JHOs than by Black JHOs. For juvenile homicides committed in large cities, Asian/Pacific Island offenders were 21% more likely involved than American Indian/Alaskan Native JHOs. Gang-related juvenile homicides were 15% more likely to involve A/PI than American Indian/Alaskan Native offenders and they were also more likely to be involved in those homicide circumstance than White offenders (1.6x) and Black offenders (6x). Asian/PI offenders were also 1.2 times more likely than White JHOs to be involved in crime related homicides. Juvenile homicides involving guns were 1.3 times more likely to involve Asian/Pacific Islander than White JHOs and 40% more likely than with AI/AN.
American Indian
White victims of juvenile homicide offenders were 2.5 times more likely to involve American Indian/Alaskan Native juvenile offenders than to involve Black juvenile offenders and 1.3 times more likely than involvement from A/PI JHOs. In addition, juvenile homicides involving family members were 1.7 times more likely to involve American Indian/Alaskan Native offenders relative to Black offenders and 2.3 times more likely than Asian/Pacific Island JHOs . For conflict-related juvenile homicides, American Indian/Alaskan Native JHOs were 1.7 times more likely to be involved than both Black and A/PI JHOs. Crime-related juvenile homicides were more likely to involve American Indian/Alaskan Native compared with White JHOs (1.7x) and Asian/Pacific Islanders JHOs (1.4x).
The authors conclude that in some regards such as offender age and sex, number of offenders, number of victims, and age and sex of victims there were no significant differences by race of the offender however in other aspects like homicide circumstances, type of victim, weapons used, region, and location there were significant differences between the races. The authors noted this was likely the first nation-wide study on race and offending that had a significant focus on Asian and Native American offending and revealed some interesting results. Regression analysis found that American Indian/Alaskan Native and Asian/Pacific Islander juveniles involved in murder could be distinguished from White and Black JHOs. Gang-related juvenile homicides were much more likely to involve A/PIs than Blacks, Whites, or AI/ANs. Juvenile killers of White victims were also more likely to be A/PI JHOs than to be Black JHOs. In addition, gun-involved juvenile homicides were much more likely to involve A/PI JHOs than either White or AI/AN JHOs. In contrast, juvenile homicides of White victims, family members, and those arising from conflict-related circumstances were much more likely to involve American Indian/Alaskan Native JHOs than either Black or Asian JHOs. American Indian/Alaskan Natives JHOs examined in this study were significantly more likely to kill with a knife and during a conflict-related situation than JHOs in the other three racial groups The proliferation of knife and conflict-related homicides among American Indians may be attributed to frustration that stems from the structural disadvantage faced by many American Indians, as well as the strong culture of honor in Native American communities.
The authors explain that other distinctions between the races exist because of social, cultural influences and geographic locations. Southern murders tend to involve more Blacks and Whites, suggesting their historical exposure to the honor culture and subculture of violence in the south, while rural murders were more associated with AI/ANs. The authors also suggest the data can be useful for making investigative decisions in that by focusing on the homicide circumstances and locations, law enforcement may be better able to determine the race of the offender for example focusing on Black offenders if the homicide was crime related or on Asian offenders if the murder is gang related.
The authors state that future research in this area could focus on four areas: (a) multiple victim homicides committed by juveniles; (b) correlates of juvenile homicide within the individual periods that include the pre-epidemic, epidemic, and post-epidemic periods; (c) broadening the racial analyses from simply White and Black JHOs, if possible, to include Hispanic and White and Black non-Hispanic JHOs; and (d) more in-depth analysis of JHOs’ social histories, levels of functioning, motivational pathways, and their crime scene behavior.
Heide, K. M., Michel, C., Cochran, J., & Khachatryan, N. (2017). Racial differences among juvenile homicide offenders: an empirical analysis of 37 years of US arrest data. Journal of interpersonal violence, 0886260517721173.
Heide, et al find distinct differences between juvenile killers of different races in the circumstances around their homicides, including geographic region with the suggestion that Black homicide offenders prevalence in the South may be influenced by its traditional honor culture. Dogan examines a possible theoretical underpinning to honor killings in a qualitative study of Turkish offenders.
Can Honor Killings Be Explained With the Concept of Social Death? Reinterpreting Social Psychological Evidence
Doğan, Homicide Studies, 2020
Dogan explains that while often associated with Muslim countries, honor killings can occur in a society or culture that has an inordinate focus on honor and respect. Honor killings are described as homicides where typically the victim has engaged in some sort of behavior that is viewed as unacceptable and thus disrespectful in the eyes of the culture and their family. This damages the honor of the victim’s family and the way that honor is restored is by killing the person responsible for bringing dishonor on the family. Victims are typically female and the perpetrators are typically male family members. Behaviors that cause a loss of familial honor might include being accused of prostitution, being an unaccompanied female which generates rumors of immoral behavior, being suspected of engaging in an extramarital affair, or having, or claiming, a family member was raped. As Dogan explains, “the concept of honor that inspires violence has a collective aspect, shaped and constructed by a gender-specific formula. To a large extent it is shaped by the perception that a man’s honor not only depends on his own conduct but is also dependent on the proper behavior of his female relatives and the members of his family or group. In this gender-specific conceptualization, the honor of a man obliges a man to defend his honor and the honor of his family, and the honor of a woman obliges a woman to maintain and protect her purity. In other words, the collective honor of the group is dependent on the control of female sexuality, and controlling female sexuality is a prerequisite for a man’s honor. The establishment, protection, and restoration of honor are paramount in honor societies and killing the individual who brought the victim or victim’s family dishonor is seen as necessary and acceptable. Individuals in these cultures, indoctrinated into accepting that loss of honor must be answered with violence, come to accept that while criminal, killing to restore honor is the correct response.”
Dogan considers that the social psychology concept of “social death” might be applied to cases of honor killings to offer an explanation of the psychological mechanism that drives these killings. The social death concept is “generally used when a person/group has experienced extreme and profound loss, such as loss of social identity, role, networks, and connections.” Social death occurs when an individual experiences isolation, ostracization, maltreatment, and stigma This social death, and its separation from culture, society, family, and friends subjects the individual to a state of nonexistence, both in others’ eyes, and in their own. A person who has lost honor no longer exists
Dogan applied this concept by examining 39 cases of 34 male perpetrators and 5 females perpetrators. For the scope of the research article, Dogan focused on 16 interviews where they made reference to having experienced explicit or implicit pressure from the community or extended family members to restore the family honor. He describes some of the scenarios that led to the killings and examined whether indications of social death were present in the interviewees accounts. Some interviewees indicated the exclusion and pressure to act in these cases:
“But, as the villagers knew what was going on in my home, they did not visit me. Even my wife kept telling me that we did not have any right to go out and look at people’s faces.” (Interviewee 30, male, age 43)
“As he (victim) continued to sexually harass me, people started to treat me differently and they started to treat me as if I was a tart and infidel to my husband; and it was me who should be blamed for what had happened. People did not believe that I was raped. I felt as if I was excluded.”
(Interviewee 36, female, age 23)
“Ten months passed between my sisters’ elopements and the killing. During this period, many things happened to me and my family . . . People began to stop greeting me. Then, I started to go out secretly. I used to check the corner of every street whether there was anybody that I knew. What I felt most was shame. I still feel shame . . . I was having difficulty to find a permanent job as a concrete worker because of the shame brought by my sisters… If they feel that they can question your honor, or your honor is in question, they try to take advantage of this point to find support among people. They say “Look! He did not do anything to his sisters and he thinks that he is a man! No way. Go save your honor first.” (Interviewee 34, male, age 35)
“There were rumors about my mother. Everybody was telling me different things about my mother . . . Especially my cousins and half-brothers were doing that. I talked with them and said “What do you want me to do? Do you want me to kill her or kill myself?” This time they said, “No, just ask your father to divorce her.” But, then they said “Those days were in the past. You cannot find any more such brave men who kill their relatives if they behave like that.” (Interviewee 7, male, age 23).
The evidence suggests that exclusion and loss of identity play a role in the scenarios as members of the community disassociate with the dishonored and deny their existence. Dogan contends that “if a person has previously been subjected to a powerful behavior pattern and mind-set suggesting how to act in specific situations, conditions, or circumstances, he or she will act or think in accordance with this pattern in such situations without judging the propriety or justifiability of the suggestion. However, this does not necessarily mean that all individuals who are engrossed in the way of thinking that the right to claim honor requires killing, and who experience the same or similar pressure are bound to commit honor killings. As I illustrated before in detail, as long as there is a way to escape from the publicity of dishonorable conduct, there may still be an alternative to restore honor without seeking violence.”
Regarding community members’ attitudes toward the perpetrators in their community after the murder, 16 perpetrators mentioned in the interview that they experienced an affirmative attitude from the community or extended family members, and their behavior either in the form of words or actions was supported by them. Some perpetrators said that people from their hometown or the community supported them in prison by sending money and visiting them, or asking how they were. Some perpetrators said that their visitors mentioned that they had done what was required, and had cleansed their honor.
“Everybody in my village was expecting the murder; they knew that it was going to happen. After the offence, they started to write me letters and send money and visit me. Later, when I was granted permission to visit my village, even people that I do not know approached me and said, “Well done, you did right thing.” (Interviewee 7, male, age 23)
Dogan did note that in three of the cases perpetrators were told by people they did not approve of the murders and that it was not right but for many of the perpetrators (11 male and three female), there is simply no other way to deal with the issue of dishonor other than the killing suggested or designed by the community, by making reference to such social determinants by using expressions such as “I had no choice,” “it was not in my hands,” “there was no other remedy,” and “this problem could only be solved like that.” Dogan states that by using these words in their interviews, they both tried to reflect the pressure that they had experienced as well as neutralize the killing they have committed by appealing to their loyalty toward cultural norms.
“A person lives for his honor and his dignity. Honor is something that holds the family and people together. It enables people to have a decent life and you live for your honor.” (Interviewee 26, male, age 26)
“Honor is a person’s pride and praise. It means everything for a person. Without it nothing can happen, nothing has a meaning. It would be better for a person to die rather than being dishonored. Without honor death would be better than life.” (Interviewee 21, male, age 47)
This social death also generates a psychological pain on the level of physiological pain that helps drive the cognitions and behavior of those afflicted.
“In my childhood, I grew up with the idea of honor, respect and reputation. My family used to say, “if somebody tries to steal your bread, or points a finger at your honor, kill him.” I was brought up like that. Therefore, for me, suffering from a wound caused by being labeled as dishonorable is more painful than a gunshot wound.” (Interviewee 13, male, age 43)
“Honor is not something light and easy. It is like a heavy burden. It is too heavy to carry. But, it has to be carried.” (Interviewee 12, male, age 43)
Dogan does declare that it’s implausible to state that all honor killings in Turkey invoke a social death dynamic nor that all those who do experience a social death are bound to commit honor killings but the concept of honor that equates loss of honor with loss of life, and thus suggestive of violence to restore honor and life, can “endure and exist mainly in such societies or communities where the individual constantly uses the concept of honor and shame to assess his own conduct and that of his fellows, and face to face personal, as opposed to anonymous, relations are the main type of relationships among members of the society.” Dogan indicates his findings suggest that whether or not an individual seeks approval through violence is dependent on the frequency, duration, and intensity of his association with the perceived norms and discourses of honor killing. It is also dependent on whether there is a way to escape from the publicity of dishonorable conduct but the lack of an alternative course, i.e. being trapped, means violence is more likely to occur.
Dogan, R (2020). Can honor killings be explained with the concept of social death? Reinterpreting social psychological evidence. Homicide studies, 24(2), 127-150.
Dogan suggests utilizing a social concept to understand the psychological mechanism behind the murder of family members. However, psychological and criminological theories can go beyond understanding what drives human behavior in a current event to being able to assist in predicting future criminal events as Lee, et al examine the nature of hotspot offending and predictions.
A Theory-Driven Algorithm for Real-Time Crime Hot Spot Forecasting
Lee, SooHyun, & Eck, Police Quarterly, 2020
The authors claim that real time crime hotspots forecasting algorithms currently in use have some drawbacks. They note a high percentage of hotspot misidentification as well as a lack of transparency in the methodology used in constructing the algorithms because they typically contain proprietary information. They consider this can be important when stop and frisk is practiced in hotspots as recently New York City, Chicago, and Los Angeles police departments have been sued over not releasing information about the algorithms used by their predictive programs.
The authors contend that the many different forecasting models lack agreement on the spatial unit to be analyzed as well as the appropriate temporal period for use in predicting future crime events based on previous events. While new models continue to be developed to assist in both short term and long term hotspot predicting, they invariably have limitations. The authors note one such limitation is that, outside of the field of criminal justice and criminology, these forecasting models rely heavily on mathematical constraints and statistical assumptions but lack theoretical foundations. The authors state their forecasting model includes the theoretical underpinnings, common in social science, of population heterogeneity and state dependence.
In the context of criminal justice, population heterogeneity would suggest that, amongst the “population of places”, certain places, i.e. targets, have features that signal desirability and vulnerability to individual offenders, making them want to target that place. For their study, the authors explain “We classify places where the hot spot forecasting is consistently successful over several months versus places where forecasting is not successful. First, for each place in the study area, we calculate the probability of crime occurring in the target month based on the distribution of crime in prior months. The more true-positive cases over several months, the more a place is consistently experiencing crime. Thus, our algorithm selects places with high amounts of true-positives over the entire study period, and screens out places with no crime, randomly occurring crime, and low probabilities of crime.”
A state dependence perspective explain repeat victimization as experiencing a crime elevates the victim’s chance of being revictimized in the short term. In the context of place, state dependence suggests that once offenders have learned about the suitability of a place for crime they will continue to engage in crime at that location. The authors note that the two theories can work in concert to understand crime patterns and predict future offenses and utilized crime data from Portland and Cincinnati to test the model using Excel statistical software and a grid cell geograph.
The authors overlaid the cities with grid cells 500′ square and for the population heterogeneity component they calculated the Poisson probability of a crime occurring in each grid cell, in each month, based on the previous 12 month distribution of crime in that cell. If the probability is greater than a .5 threshold and the target month experiences crime, the forecast has produced a true-positive case and the distribution of true positives over the study period indicted some areas are more predictable than others.
For each grid cell, the Excel formula they utilized returned either a 1 or 0 for each month. Averaging these binary values over the entire study period, they obtained the average true-positive value for each grid cell. The closer this value is to 1, the more predictable the grid cell is. Using these average true-positive values, they sorted the grid cells from the most predictable to the least predictable, analogous to a situation where some places are more vulnerable while other places are less vulnerable to crime.
For the state dependence component of the model, the number of crimes recorded for the current month were assessed to calculate the elevated risk on the grid cell toward the forecasted month. The grid cells were sorted by their average true-positive values (from the most to the least predictable ones), then sorted by the number of crimes in the most recent month from the highest to the lowest.
In a three step process the authors explain that “Step 1 inputs average true-positive values for each grid cell based on the distributions of crime in the past 12 months. In Step 2, grid cells were sorted by their average true positive values in descending order from highly predictable ones (e.g., 100% and 90%) to less predictable ones (e.g., 10%). (In Step 2) based on the population heterogeneity, we select highly predictable grid cells ..but discard less predictable ones. Then we look at the number of crimes in the most recent month for each grid cell to apply the state dependence process in the final step. In Step 3, we reorder the preselected grid cells in Step 2 by their number of crimes in the current month.”
The authors used measures of accuracy and efficiency to test the model. To calculate the accuracy, or how well the model correctly forecasted what areas would actually become hotspots, they used the ratio of the number of forecasted hot spots that became true hot spots in the forecasted month compared to the total number of true hot spots that developed that forecasted month. For a measure of efficiency of how well the forecasting algorithm predicted the number of crimes in the forecasted hotspots they developed a Prediction Efficiency Index, (PEI) which is the ratio of the number of crimes in forecasted hot spots compared to the number of crimes in actual hot spots.
Comparing their Portland calls for service data forecasting results to the winners of the NIJ Crime Place Forecasting Challenge they found that the PEI scores for the winning model only performed slightly better than the authors’ model in the one, two and three month forecasting periods for the CFS groupings but still generated efficiency scores ranging from over 90% for all CFS, over 80% for streets crimes, around 20% for Burglary and almost 60% for auto theft in the first month, which, similar to the challenge winners, dropped to approximately 35% for the three month forecast. Accuracy ranged from over 70% for all CFSs to over 60% for Street Crime CFS but down to an average of around 15 % for Burglary across the forecasting periods. While accuracy for Auto Theft was over 40% at 1 and 2 months, accuracy dropped to slightly more than 20% at the 3 month mark.
The authors also analyzed Cincinnati incident report data and found that for all crime incidents the PEI increased from 69% to 82%, as did accuracy from 43% to 54%. PEI (41%-59%) also increased across the forecasting periods for Part One Violent Crimes but while the accuracy was approximately 40% for the one and three month forecasts, its highest accuracy (62%) was in the 2 month forecast period. Auto theft PEI and accuracy were both low but did increase with future forecasting, ranging from 7 to 21 percent, and 7 to 19 percent respectively. For Part One property crimes, PEI increased from 70 to 82% across the forecasting periods. Accuracy ranged between 53 and 58% but the lowest accuracy for auto theft was in the 2 month forecasting period. In Portland, while PEI scores decreased with more distant forecasting, for Cincinnati, the PEI tended to increase as researchers forecast further into the future
The authors found, as has past research, that property crimes are not as easy to forecast accurately as violent crime, and that further model development should be done to enhance the forecasting of specific crime types. The authors conclude that while the NIJ Challenge winner’s algorithm performed slightly better in its PEI score, their model is an improvement over other algorithms in use which are extremely poor at capturing the number of crimes relative to crimes in actual hot spots. The authors also note that compared to other forecasting models theirs involves a lower fiscal investment and is easier to utilize as it can be done as a simple formula in Microsoft Excel and provides transparency for methodology and data inquiry purposes.
Lee, Y., SooHyun, O., & Eck, J. E. (2020). A theory-driven algorithm for real-time crime hot spot forecasting. Police Quarterly, 23(2), 174-201.