Expressive Partner Homicides in Spain: Differences according to Agressors Sex

Introduction: Although the important concert about partner homicides, there are no scientific studies in Spain. The main aim of this study is to analyze differences in Spanish partner homicides between aggressor's sex with a) crime perpetration: attempted or completed and b) criminal process: pre-crime, offence and post-crime behavior. Method: We used Basa Data "Protocolo de Análisis del Crimen Violento en Homicidios Familiares", composed by firmed sentences of family homicides in Spain. Specifically, we analyzed a sample of 323 partner homicide firmed sentences, mostly men. Results: The results show that attempted crimes are more associated with men, and completed homicides with women. Regarding pre-crime behaviors, the existence of previous conflict and previous break up are more associated with men. In criminal context, instrumental behaviors are more related with women while expressive behaviors don't show relation with aggressor's sex; to strangle the victim is more associated with women. Finally, none of post-criminal behaviors show relation with aggressor's sex. Discussion: The results regarding pre-crime and post-crime behaviors are consistent with precedents studies in other countries, also the predominance of expressive behaviors in both, men and women. Concerning to crime perpetration, it hasn't been found previous studies comparing it with offender's sex in partner homicides.


Introduction
Domestic violence is defined as the type violent and no-accidental action that implies the physical and/or psychological damage of a family member. The emotional attachment that exists in the relationship between victim and aggressor makes to consider domestic violence as a unique type of violence [1,2]. Domestic violence covers a wide range of types of violence but domestic homicide is the most serious one and, regarding to intimate partners, homicide is the most common type of domestic homicide [3].
It has been considered that in partner homicides, expressive violence is frequently involved. Terms as expressive and instrumental violence have existed in the criminological literature since seventies and they have been largely defined by authors. Expressive violence is characterized by the lack of control, impulsivity, strong emotional feelings, irrationality and lack of premeditation [4].
The individual, as a result of psychological and external stressors such as anger, negative moods, fury or irritation, would experience states of tension and physical activation and in turn, these emotional and psychological states would be externalized as physical and brutal attacks towards the victim [5][6][7][8][9]. Therefore, in expressive violence, the closer the victim and the offender are, the higher the level of expressiveness will be involved in the offence [5,[9][10][11][12][13]. Because of physical contact and direct contact with the victim are key elements of expressive violence [4,14,15] it has been considered that expressive violence is the most predominant type of violence observed in partner homicides.
On the contrary, instrumental violence is planned, structured, controlled and non-emotional. The objective of the aggressor is not the victim but something that can be gotten from the victim (e.g. sex, properties, money, etc.).
Nonetheless, as some authors have pointed out neither expressive violence, necessary exclude instrumental violence, nor instrumental violence necessary excludes expressive violence. Instead, both types of violence can occur in the same crime and in the same violent aggression, although expressive violence is more prevalent in partner homicides [7,16,17]. This predominance of expressive rather than instrumental violence, and therefore the increase of impulsiveness, irrationality and spontaneity reduce both the rational and the decisionmaking process that offend does when committing the crime [18][19][20][21][22][23][24][25][26][27][28].
Previous investigations have observed that males and females differ in aggressive behaviors being males more predisposed to aggressive behaviors than females. It has been suggested that this is due to sex differences in inhibitory control capacity and in their management of the arousal levels. The arousal levels in which males would express feelings of anger would be lower than the ones under which women would express those feelings. Furthermore, pre-criminal conditions and situations that precipitate the crime are also different between males and females. Whereas in the case of males previous break-ups, separation or feelings of losing someone they used to love can precipitate the crime [29,30], in the case of females, those condition do not usually appear but instead it is frequent that they have a previous history of partner abuse or mental health problems [25,[31][32][33].
Previous investigations have helped to increase the understanding of partner homicide in a scientific way, nonetheless, little is known about this crime in the Spanish context, which, due to different ecological factors results from previous investigations could not be fully extrapolated to the Spanish reality. Official Spanish data exposes that most of the victims of this type of crime are married women who live with their husband, nonetheless little is known of the offender.
This lack of awareness in the Spanish context is not only due to the scarce official data gathered but also because of the lack of studies conducted in Spain examining it. The present investigation aims to palliate this research gap by examining partner homicides in the Spanish context. In order to get this aim, the study is going to examine expressive and instrumental violence as well as the behaviors conducted in all three stages of the crime: pre-criminal behaviors, criminal behaviors and post-criminal behaviors.
Furthermore, the study will examine if any of the behaviors studied are associated to a specific gender. Taking previous research into consideration it is hypothesized that: 1. Expressive behaviors will be related to male aggressors while instrumental behaviors will be related to female aggressors, 2. Regarding pre-criminal behavior previous conflicts, previous breaks-up and previous violence and/or threats will appear in the case of males, 3. Regarding criminal behavior scab wounds and blunt objects will appear in both sexes. Nonetheless, instrumental behaviors are expected to be more predominant in the case of females and expressive behaviors in the case of males.

Method Sample
Data was collected from the Spanish Judicial Base Data (CENDOJ). The Base Data gathers a total number of 517 family homicides firmed sentences in Spain from 2000 to 2010. We considered as partner homicides those cases in which victim and offender have or have had a sentimental relationship. Following this criterion, three hundred and twenty-three (N=323) cases out of the total sample were considered as partner homicides.

Instrument
In order to conduct the study it was used the "Protocolo de Analisis del Crimen Violento en Homicidios Familiares" (PACVHF) [34]. This is a 96-item questionnaire that gathers information from family homicides committed in Spain. It covers seven thematic groups: 1. Criminal justice information; 2. Social, familiar, personal and medical information about both victim and aggressor; 3. Criminal event variables (i.e., chronology, geography, methods of approximation); 4. Criminal scene variables (i.e., victim's body position, victim's personal objects, cause of death, injuries, sexual activity and weapon); 5. Witnesses' information; 6. Forensic evidences and 7. Penal consequences.

Procedure
In order to analyze the expressiveness of partner homicides we selected those variables that according to previous research refer to expressive and instrumental violence [11,13,15,25,[35][36][37][38]. According to literature, those variables considered as expressive (previous threats/ violence, previous break-up, previous conflict, asphyfiation, scab wounds, gunshot, blunt object, suicide, turn himself to the police) were group together and codified as "Expressive violence", alike those variables considered by previous research as instrumental (strangulation, burns and intoxication) which were group together and codified as "Instrumental violence".
Furthermore, in order to examine the sex difference in all three stages of the crime the following variables were selected and classified in three main groups in order to be studied independently. The three main groups were: 1. Pre-criminal behavior, 2. Criminal behavior and 3. Post-criminal behavior. Pre-criminal behavior gathers those behaviors conducted before the crime, three of them were considered for the study: previous threats/violence, previous break-up and previous conflict.
Criminal behavior gathers the criminal actions that were conducted along the crime and resulted in the death or the injury of the victim. The behaviors examined are: asphyxiation, strangulation, scab wounds, burns, intoxication, gunshot and blunt object. Finally, post-criminal behavior gathers the actions carried out by the offender after committing the crime, they were the act of turning him/herself in to the police or the commission, or suicide attempt.
Based on Spanish firmed sentences (from 2000 to 2010) considered as a family homicides from Database CENDOJ (N=323), two trained investigators in PACVHF analyzed them separately in order to collect data about variables. A third investigator reviewed individually data collected in order to solve discrepancies between the two first investigators, obtaining impartial dates.
Our study is based on results obtained from the third investigator (reviewer). The collected information was systematize in a database and it was conducted a descriptive analysis in order to observe percentage differences between both genders for each of the behavior considered for the present investigation.
Furthermore, it was conducted a one-way analysis of variance (ANOVA) in order to examine if any of the actions were associated with any of the sexes.

Results
The main aim of the present investigation was to examine the sex differences between aggressors in partner homicides. Three hundred and twenty-three cases of partner homicides composed the final sample. In 61.3% of the cases, the aggressor was a male (n=198) and in the 38.7% of the cases the offender was a female (n=125). Table 1 shows that four of the behaviors examined are associated with one of the genders: instrumentality, strangulation, previous breakups and previous conflicts. Results reveal that instrumental behaviors are associated with females. Considering all the behaviors examined in the group of crime perpetration, Table 1 shows that scab wounds and blunt objects are the most prevalent ones in both sexes.
Nonetheless, none of them is associated with a particular gender. In fact, only one behavior of this category is associated with one of the sexes, that is, strangulation is associated with female offenders. Regarding pre-criminal behaviors, two out of the three behaviors are associated with males, they are, previous break-ups and previous conflicts. On the contrary, none of the post-criminal behaviors are associated with any gender (Table 1).  Table 1: Descriptive analysis and One-way analysis of variance (ANOVA). * p<0.05

Discussion
The results of the present investigation confirm the hypothesis of the study by showing that instrumental violence is associated with female aggressors. This result is in accordance with previous studies examining expressive and instrumental violence that also concluded that whereas expressive behavior is more observed in male offender, instrumental violence is frequently observed in female offenders. As some authors have observed, women tend to evaluate more negatively than male the possible consequences of losing control of the situations. Furthermore, women loss control and show expressive violence under situations that tend to be more stressful if compared with the situations under which men generally react in an expressive manner [13,20,29,30]. A hypothetical explanation for that could be at situations when a self-defense woman forms an attack. Considering all, it makes sense that, the only behavior, of the crime perpetration group, that was associated with females was an instrumental behavior, that is, strangulation. It must be noticed that strangulation do not only refer to manual strangulation but also to the strangulation carried out with the help of an object such as a rope.
As it was hypothesized, scab wounds and blunt objects were the most common behaviors observed in both sexes, being both conducts instrumental. One possible explanation of these results is that, as previous research suggests, expressive aggression is the most basic form of aggression in both sexes [15,18,25,37,[39][40][41]. In addition, when expressive violence is showed in a criminal behavior as homicide, the unplanned situation and impulsivity make the offender committed the crime with something around him, as a knife or a blunt object as a lamp or a hammer for example.
The hypothesis that pre-criminal behaviors would be associated with males is also verified. Previous studies have observed that partner homicides commonly occur in contexts of interpersonal conflicts (i.e., arguments, threats, maltreatment). Both previous conflicts and previous break-ups are generally present in gender violence in the Spanish context [7,11,15,18,25,[35][36][37][38][39]41,42], what could explain why the present investigation has found an association between males and these two variables. Interestingly, contrary to previous research, it has not been found an association with previous violence. A possible explanation could be the fact that the number of violent reports is reduced because women are afraid of doing this and suffer reprisal from her partner. It doesn't mean a lack of previous violence, but we could not prove it.
Finally, the last hypothesis related to post criminal behavior wouldn't be associated with any sex. In other researches about intimate partner homicide, results show an association between expressive violence and post-criminal behaviors related to offender's regret, especially in women offenders. But we could not prove it in Spanish intimate partner homicides.

Limitation and Future Directions
The present investigation has some methodological limitations that must be considered when analyzing the results. First, the data was obtained from firm sentence that, although being a reliable source of information, they fail, in many cases, to provide all the relevant information to fulfill the "Protocolo de Ánalisis del Crimen Violento en Homicidios Familiares" (PACVHF). Second, the quantitative nature of the study enables to control variables and study them independently, nonetheless, it does not allow to study other variables that, although could have an impact in the study, are not gathered in the used instrument. Those limitation, although do not underestimate the results obtained must be considered when extrapolating the results.
This investigation helps to provide a more scientific view of the reality of partner homicides in Spain. Nevertheless, further research is needed examining both, other factors and behaviors that can be observed in this type of crime as well as the reasons of the results obtained in the present study.