Chapter 3 Literature Survey
The first part of this chapter focuses on the incidence rates of H-S in western and non-western countries. Special attention is paid to seven South African studies which investigated the H-S phenomenon on the African continent. The next part of this chapter identifies the common factors involved in H-S killings. These factors include mental illnesses, relationship problems, domestic violence, substance use and intoxication, the weapons used, and financial adversity. This chapter also reports on H-S profiles which were obtained from the various studies involved. This chapter concludes with the rationale and accompanying research questions that directed this current study.
H-S Incidence Rates
Although H-S is considered a rare occurrence, these events are widely publicised in the news and therefore appear to be increasing. In 2009, Eliason mentioned that literature reviews since 1992 indicated that H-S frequency rates had been stable for the past 40 years in the USA and other western countries 2009). After reviewing 17 studies ranging from 1900-1979, involving ten nations, Coid (1983) concluded that that the H-S phenomenon occurs at a relatively constant rate with an incidence rate of 0.2 to 0.3 per 100 000. In a more recent study, Marzuk et al. (1992) also estimated an incidence of H-S between 0.2 and 0.3 per 100 000 per year in the USA.
Other US studies also found incidences of 0.2 to 0.3 per 100 000. Campanelli and Gilson (2002) found similar incidences for the period 1995 to 2000 in New Hampshire, Comstock et al. (2005) reported a similar incidence rate (1994-2001) in Oklahoma, and Bossarte et al. (2006) also found such an incidence rate for the period 2003-2004 in multiple states. Australia showed an incidence rate of 0.16 per 100 000 for a 3 year period (1989-1991), while the European countries of England and Wales showed instances of 0.07 per 100 000 (1980-1990) compared to Scotland’s lower rate of 0.05 per 100 000 for 1986 to 1990 (Byard et al., 1999). In turn, Switzerland reported H-S rates of 0.09 and 0.10 per 100 000 for two different periods, 1991 and 2001 respectively (Panczak et al., 2013). Among non-western countries, researchers showed that the incidence of H-S was 0.22 per 100 000 in Hong Kong (Chan, Beh, & Broadhurst, 2004) over a ten year period (1989-1998), while Gupta and Singh (2008) found a low incidence rate of 0.06 per 100 000 in India (2000-2004).
In the US policing sector, Violanti (2007a) conducted a study involving 29 H-S cases that had occurred between January 2003 and February 2007. This study reported 1 H-S case in 2003, but this increased to 7 events in 2005 and peaked with 15 cases in 2006. The author concluded that police H-Ss were increasing. The methodology involved the use of newspaper articles which were obtained from a website that focused on problems within police households (e.g., domestic violence). Unfortunately, newspaper surveillance studies can be criticised for their subjective nature and unreliable content (Malphurs & Cohen, 2002; Roberts et al., 2010), while Rainey and Runyan (1992, as cited in Violanti, 2007a) argued that newspapers are a feasible and economical way of collecting information on intentional injuries.
A more recent police H-S study conducted by Klinoff et al. (2014) included 43 cases between June 2007 and February 2014. This US study used the same methodology as the previously discussed study undertaken by Violanti (2007a). In contrast to the Violanti (2007a) study, Klinoff and her colleagues (2014) could not determine whether police H-S was increasing annually. However, these killings have increased steadily since 2007 (2 cases) until they peaked during 2010 with 12 incidents. During the period 2011-2012, a decrease was observed (5 & 2 cases respectively), and in the following year, an increase was noted with 8 incidents being reported. Lastly, 2014 included only two months with 2 H-S cases for this period. Other findings from these two law enforcement studies are discussed in the remainder of this chapter.
In the South African context, Jena and her colleagues (2009) studied 46 H-S events in the Pretoria area over a 5 year period (1997-2001). These researchers found estimated annual instances of H-S between 0.8 and 1.3 per 100 000 compared to the total SA population, which is significantly higher than the international rates of 0.2 to 0.3 per 100 000 (Coid 1983; Bossarte et al., 2006). The methods employed to obtain H-S information involved data obtained from the records of the Medico-Legal Laboratory (MLL), mortuary files (autopsy reports), South African Police Service (SAPS) documentation, and lastly, by contacting the SAPS investigating officer. The two municipal areas which recorded the most H-S cases were Mamelodi with 26% (12), and Atteridgeville with 13% (6) of the case load. Most of these deaths occurred in residential areas (78.2% (36)) and the rest of these deaths took place in both public areas (10.9% (5)) and hospitals (10.9% (5)). Lastly, an important limitation of the study conducted by Jena et al. (2009) relates to the low number of cases that had occurred, which makes generalisation of the findings statistically difficult.
The study carried out by Townsend (2003) served as a pilot investigation for Roberts et al. (2010). Both of these SA studies focused on the Durban region for the same 2 year period (2000-2001). Therefore, only the more recent study undertaken by Roberts et al. (2010) is further discussed. This regional study included 21 cases and found H-S instances of 0.89 per 100 000, which also exceeds the international rates. For unknown reasons, there was a 16% increase in H-S killings from 2000 to the following year; a total of 43 people died in these events. Data originated from coroners or district surgeon reports which were supplemented by police files. The results revealed that 30% of the perpetrators and only one victim (4%) were employed in the police service or security sector. In contrast to the number of perpetrators and victims who were engaged in formal employment, 25% of the perpetrators and 30% of the victims were unemployed at their time of death.
Although this study made a significant contribution towards the understanding of H-S events and creating a first general profile of perpetrators, victims, and the methods employed in such killings, the results cannot be generalised to the rest of South Africa. A national study is of utmost importance in order to obtain a comprehensive profile of the H-S perpetrators and their victims so that the said profile can be generalised for the rest of the country. Secondly, mental illness as a risk factor in H-S that would be statistically representable for this country was not included in the aforementioned study. The international study conducted by Lecomte and Fornes (1998) clearly illustrates that a substantial number of people who committed H-S suffered from a psychiatric disorder. By including the psychiatric history, a more detailed profile of H-S offenders could be generated.
In the third SA study, researchers used a cross sectional design which included 1349 perpetrators of intimate femicide on a national level (Mathews et al., 2008). According to these researchers, the estimated H-S rate in South-Africa amounts to 1.7 per 100 000, indicating an H-S rate which greatly exceeds the reported incidences in developed countries. Intimate femicide was further sub divided into intimate femicide-non-suicide in 80.6%
(1088) cases and intimate femicide-suicide in 19.4% (261) of the cases. The latter group (19.4%) of perpetrators of intimate femicide committed suicide within one week of the original murder. The methods used for data collection in this study included sourcing information from mortuary files, autopsy reports, and police interviews. This study can be criticised for the choice of the comparison groups that were used. The two groups consisted of an intimate femicide group and intimate femicide-suicide group in which the former group refers to the killing of a female by her intimate partner without committing suicide, and the latter, the intimate partner kills himself after committing the homicide. By comparing these two groups, the common risk factors which contribute to each different subtype can become clouded or even invisible.
The following two SA studies used press reports as the method to gather information on H-S cases (Osborne, 2001; Skead, 2010). The first study identified 83 cases of H-S between 1997 and 2001, while the second study identified 328 H-S cases that had occurred during an eight-year period (2002-2009). Both studies not only included femicide-suicide, but all types of H-S killings (e.g., filicide and extra-familial suicide). The Osborne (2001) study did not report on incidence rates while Skead (2010) found an average incidence rate of 0.09 per 100 000 for the eight-year period. This incidence rate is significantly lower than most international (0.2 to 0.3 per 100 000 (Coid 1983; Bossarte et al., 2006)) and national (between 0.8 and 1.3 per 100 000 (Jena et al., 2009); 1.7 per 100 000 (Mathews et al., 2008)) H-S rates. This finding could be attributed to the fact that H-S tragedies often occur in poorer communities, which renders these events less newsworthy than interesting events (Osborne, 2001).
The last noteworthy South African study was conducted before the first democratic elections of 1994 (Roos et al., 1992). This apartheid era study included 11 H-S cases which spanned over an unspecified 2 year period. The data collection method included interviews with extended family members, neighbours, the investigating officers of the police, general practitioners of the families, church ministers, and even school teachers. The H-S cases were identified by utilising police referrals and the said study focused on family murders. The study highlighted the psychiatric and psychological factors involved in H-S killings. Roos and his colleagues concluded that certain personality traits, suicidal ideation, and depression precipitated by marital discord, and financial problems created a slippery pathway to H-S. Unfortunately, this dated study did not report incident rates and can be criticised regarding the small sample that relied on only police referrals.
Although it is clear that South Africa has an unusually high H-S rate compared to other countries, an important question still remains: What are the factors that could lead to H-S events?
Common Factors in H-S
H-S can be described as a complex human phenomenon which involves biological (e.g., psychopathology) and social factors (e.g., availability of firearms, conflict in relationships and financial difficulties) (Fox & Levin, 1998; Kennedy-Kollar & Charles, 2010), which are intertwined. The literature identifies several variables that are presumably responsible for triggering an H-S incident. These factors, with their accompanying studies, are discussed in the following section in greater detail
In most of the international studies, a mental disorder was identified in the perpetrators of H-Ss. In most cases, depression was diagnosed (Cohen et al., 1998; Bourget, Gagne &
Moamai, 2000; Campanelli & Gilson, 2002; Malphurs & Cohen, 2002; Chan et al., 2004; 64
Malphurs & Cohen, 2005; Moskowitz et al., 2006; Bourget et al., 2010; Dogan et al., 2010) followed by other psychopathology. Other psychological disorders included substance abuse, for instance, alcohol problems (Koziol-McLain et al., 2006), drug abuse (Bossarte et al., 2006), schizophrenia (Virkunnen, 1974; Campanelli & Gilson, 2002), psychosis (Gudjonsson & Petursson, 1982; Hatters-Friedman et al., 2005), antisocial personality disorder (Dogan et al., 2010), and schizoaffective disorder (Hatters-Friedman et al., 2005). In SA, the Jena et al. (2009) study found a psychiatric history in 15% (7 perpetrators) of the H-S cases and of these cases, 42.9% (3 offenders) displayed a positive psychiatric history. The authors concluded that this sort of tragedy usually peaked during autumn and winter, which possibly indicates mood disorders. The apartheid era study conducted by Roos et al. (1992) found that 83% of the perpetrators had experienced mood disorders prior to the killings. More specifically, 59% of these offenders suffered from a major depressive disorder compared to the 6% who suffered from a bipolar mood disorder.
The press report study by, Skead (2010) found psychiatric illnesses in some of the cases (1.22%), but unfortunately the media did not specify the type of psychiatric disorders that played a role in these killings. The international study undertaken by Cohen et al. (1998) used both newspaper surveillance and medical examiner files as data sources. This study reported a distinct difference between older (over 55) and younger H-S offenders. Higher levels of depression were found amongst older perpetrators from both West Florida and South-Eastern Florida (37% and 19% respectively) compared with younger offenders (10.8% and 11.5%) from these locations. Other mental illnesses were also indicated in this study, but unfortunately, these disorders were not clearly identified.
Another international study conducted by Malphurs and Cohen (2002) also made use of newspaper surveillance as the research methodology. Histories of psychiatric illnesses were found in only 3.8% of the cases without specifying the type of psychological disorder involved. Unfortunately, press reports do not usually investigate psychiatric motives (Roma et al., 2012) and as mentioned previously, these studies are often criticised for their subjective reporting and unreliable content (Malphurs & Cohen, 2002; Roberts et al., 2010).
This lack of reliability could exert a significant influence on the understanding of the H-S phenomenon.
Most of the international studies employed various definitions of mental illness and failed to indicate the degree of overlap between the various psychological disorders (Roma et al., 2012). For instance, the Campanelli and Gilson (2002) study identified three different psychiatric disorders, namely depression (38%), alcoholism (25%), and schizophrenia (6%). Unfortunately, a possible overlap (comorbidity) between these mental disorders remains unclear. The same trend was observed in the Rosenbaum study (1990) which identified depressive disorders in 75% of the cases involved. These mood disorders included major depression, bipolar, and dysthymia. Research carried out by Klein and colleagues (2000) reported that certain individuals can suffer from both major depression and dysthymia simultaneously and that this ‘double depression’ was not considered in the aforementioned study
Most H-S cases involved a male killing a female, usually a wife, girlfriend, or ex-wife/girlfriend (Marzuk et al., 1992; Bossarte et al., 2006; Banks et al., 2008; Eliason, 2009). All five of the SA studies reached the same conclusion (Osborne, 2001; Mathews et al., 2008; Jena et al., 2009; Roberts et al., 2010, Skead, 2010). More specifically, the two SA press report studies conducted by Osborne (2001) and Skead (2010) found that men were the most common perpetrators in 88% and 90.8% of the cases respectively. Skead (2010) also concluded that an intimate relationship was the most common victim-perpetrator relationship (45.5%) in H-S.
Jena et al. (2009) identified 45 male perpetrators in their Pretoria regional study and found that 36% of the victims were married, 4% divorced, and 54% single. In the Durban regional study conducted by Roberts et al. (2010), they also reported that 95% of the perpetrators were male and that 75% of the couples were either married or had current or past intimate relationships. In the Mathews et al. (2008) cross sectional study, 40.4% of H-S killings were committed by the cohabitating partner, secondly by the husband (30.8%), and thirdly by the boyfriend of the victim (28.8%).
In a limited number of cases, females were the perpetrators and males were the victims in H-S killings. The North American study undertaken by Gartner and McCarthy (2009), which spanned over a 90 year period (1900-1990), found that 20% of the female offenders committed suicide after killing their intimate male victims. The Swiss national cohort study of Panczak et al. (2013), which included 73 H-S cases, reported that only three female perpetrators (4.1%) killed their husbands compared to the 55 events (75.3%) in which male spouses murdered their wives. In the SA context, Jena et al. (2009) identified only one female perpetrator, while Skead (2010) reported that female offenders mainly targeted their male children rather than their male consorts or spouses. The apartheid study carried out by Roos et al. (1992) reported that 59% of the perpetrators were male compared to their female (41%) counterparts.
According to Bossarte et al. (2006), H-S incidents seldom occur between strangers (less than 5%) when compared to current or former intimate partners (58%). Gartner and McCarthy (2009) reported that 21% (83 of 395) of H-Ss occurred between offenders and victims who cannot be classified as either intimate partners or kin. This ‘sub category’ of victims included landladies, neighbours, and co- workers with whom the perpetrators were infatuated or had even pursued in the past, but had been rejected by them. The Northern Australian study of Richards and Weaver (2009) also included nurses and prostitutes in this sub category, which highlighted morbid obsession amongst offenders. In these cases, males turned violent against females who rejected their sexual advances, whom Mathews et al. (2008, p.553) referred to as ‘rejected would-be lover(s)’. There were never romantic relationships between the killers and their victims and these ‘relationships’ were based on the fantasies and mental derangement of the men.
In some of these studies, estrangement was an important factor in H-S cases alongside impending divorce and separation (Hannah, Turf, & Fierro, 1998; Palermo et al., 1997). In SA, amorous jealousy was identified in 13.9% of the 46 cases in the study conducted by Jena et al. (2009). This jealousy coupled with possessiveness by the partner of the female could be an important causal factor in H-S events (Cooper & Eaves, 1996; Bourget et al., 2000; Dawson & Gartner, 1998 as cited in Mathews et al., 2008). Even perceived or real infidelity could be an important precipitating factor in H-S (Aderibigbe, 1997; Chan et al., 2004), especially when the relationship is terminated. The ending of a relationship has been described by Richards and Weaver (2009, p.310) as ‘an especially dangerous moment’ for the female.
The SA newspaper study of Skead (2010) found that the most common precipitating factor in H-S was a quarrel or conflict between a male and a non-spousal female in 17.68% of these cases, followed by marital conflict (14.94%). A similar conclusion was reached by Mathews et al. (2008), who found that H-S took place after an argument (53.5%) or the female ending the relationship (29.0%). Cohen et al. (1998) discovered a distinct difference between older (over 55) and younger couples in H-S. Older couples were more likely to suffer from medical ailments whereas the younger couples were more inclined to have a history of verbal arguments. Lastly, separation also played a more dominant role in the younger couples compared with the older group. In turn, Panczak et al. (2013, p.4) reported that H-S peaked in middle and older age groups. This study found that “amorous jealousy” was an important factor in younger couples between 30-49 years, while “mercy killing suicide” in older couples (70-80+ years) was the impetus for H-S killings.
In most of the previous studies, males killed their spouses or consorts and then committed suicide. It could possibly be deducted that the female victims did not have any desire to be killed by their intimate male partners. In contrast, during lover’s suicide pacts, both parties decide to kill themselves using the same method. This ‘variation’ in H-S can be attributed to different precipitating factors, including opposition of parents to marry, family feuds, social isolation, and chronic illness (Murthy et al., 2001). Chronic illnesses can include gastric cancer, lymphoma, HIV/AIDS, diabetes, chronic back pain (Jensen, Gilbert, & Byard, 2009), and thalasemia (Murthy et al., 2001).
In the police literature, Violanti (1997) suggested that relationship problems are one of the major causes of police suicides and that aggression which results from dysfunctional relationships can also create a slippery pathway to H-S (Violanti, 2007a). A similar conclusion was reached by Klinoff et al. (2014) that divorce and estrangement were important motives in the majority of H-S cases, especially if the relationship had been characterised by a history of previous separations.
In certain instances, H-S killings can also be extended to the offspring of the couple. The descriptive study conducted by Hatters-Friedman et al. (2005) included 30 filicide-suicide cases which spanned 44 years (1958-2002). Results showed that 65% of the husbands attempted to kill their spouses along with their children compared with 0% of the female spouses who killed only their children, but not their husbands. Byard et al. (1999), who identified 13 cases during 1969-1998, reached a similar conclusion. The latter study reported that mothers killed only their children and not their spouses, while the male spouses killed their wives as well as their children. Other important findings of the Hatters-Friedman study (2005) relate to the sex of the children. Most of the victims were girls (65%) compared with the 35% who were boys. Also, seven additional children were attacked, but survived the event. More girls survived the attack (57%) compared with a lower percentage of boys (43%). Older children with a mean age of 7.1 years were the main the target in H-S killings, rather than infants. Byard and his colleagues (1999) reported an average age of 6 years for the 22 children included in their study, with an equal sex distribution of 11 girls and 11 boys.
GET THE COMPLETE PROJECT