How are the homicidal women in Spain? A case series

Abstract

Introduction

Homicides committed by women are of special scientific and social interest. In forensic literature, common characteristics differentiating women’s homicide of those committed by men have been observed. These characteristics concert the act of homicide itself, and also the victims and aggressors characteristics. Our objective is to find if in Spain, homicides committed by women present such characteristics or substantial differences exist.

Material and method

We analysed a case series (n = 18) of homicides committed by women. We evaluated their biography, homicide characteristics, previous mental illness history and drug use, and data of interest concerning the legal procedure.

Results

In our sample, as highlighted differences between our study and others, we found less traumatic events in the biography of women who have a homicide conviction. We also found less history of previous aggresions and threats to the women by their victims.

Final Considerations

More studies are necessary to confirm these findings.

Resumen

Introducción

Los homicidios cometidos por mujeres presentan un interés especial a nivel científico y social. En la literatura forense, se han observado unas características comunes a los homicidios con autoras femeninas que los diferencian de los cometidos por hombres, tanto a nivel acto homicida en si mismo, como de las características de agresora y víctima. Nuestro objetivo es averiguar si en España los homicidios cometidos por mujeres presentan dichas características o hay diferencias sustanciales.

Material y métodos

Analizamos una serie de casos (n = 18) de homicidios cometidos por mujeres. Valoramos la biografía de las mujeres, las características del homicidio, el historial de trastorno mental y consumo de tóxicos y datos de interés del proceso judicial.

Resultados

En nuestra muestra, como diferencias destacadas entre nuestro estudio y otros, encontramos un menor número de antecedentes biográficos traumáticos entre las mujeres condenadas por homicidio, así como una menor cantidad de agresiones y amenazas previas por parte de las víctimas.

Consideraciones finales

Son necesarios más estudios para confirmar estos hallazgos.

Introduction

Within the context of the interest in the study of homicide, homicides committed by women tend to receive plentiful attention from forensic scientists as well as the media.¹ This is partly due to their relative scarcity (as worldwide, only 10% of homicides are committed by women²), as well as the characteristics of female homicides compared to those committed by men. Women usually use conflict resolution and coping strategies for stress that are less aggressive, so they do not habitually use violence as often as men. An example of this is that women were only responsible for 5%–20% (depending on the type of offence) of crimes considered to be violent in 2018.³ With respect to homicide, when they do commit this, they are usually motivated by factors which are not shared by their male counterparts.^4,5

It has been observed that homicides committed by women tend to have their partner as their victim,⁶ usually within a context of major emotional stress factors such as mistreatment or economic dependence.^4,5 In such cases, women usually seek other forms of help (going to the authorities, social, or medical services) before taking violent action.⁴ Their children too are often the victims of women, such as in the case of post-partum depression or psychosis, or in the context of an overload of care work or a delirium.^7,8 This may also occur in cases of what is known as expanded suicide, in which killing children is seen to be a way of preventing them from suffering.⁹ Of the victims who are not direct family members, women have also been found to kill closer victims (other family members, friends, or acquaintances). In the several studies that were reviewed, unknown victims accounted for 2%–19% of homicides committed by women, while in those committed by men they amounted to 16%–48%.^4,6,10

Apart from the cases in which mental illness is a decisive factor in the intention to commit homicide, as would be the case in post-partum psychosis, women who commit homicide seem to suffer a higher prevalence of mental illness than their male equivalents, together with a major quantity of stressful and traumatic events in their previous life.^4,11 One study found that 28% more of all the women had suffered an emotionally distant childhood than was the case with the men. The difference is even greater for sexual abuse, which affected 34% of the women as opposed to 5% of the men.⁴

The aim of this study is to analyse the homicides committed by women who were assessed during the hearing by the Forensic Psychiatry Unit of the Legal Medicine Unit of Valencia over a 10-year period, to discover whether findings agree or disagree with those of previous studies. Present and past perpetrator psychosocial factors were evaluated, together with their mental health history. Given that one review suggests that the presence of a mental illness and its type is an important influence on the choice of the method used for homicide,¹² we also analyse this aspect. Likewise, we evaluate the relationship between the perpetrator and their victim, as this too influences the way the crime is committed.¹³ Lastly, we evaluate some characteristics of interest in the judicial process.

Material and methods

A retrospective study was undertaken, reviewing the files of the Legal Medicine Institute, Valencia. All cases of homicide or attempted homicide from 2006 to 2016 were selected in which the suspect was female and had been subjected to forensic medical evaluation, when there was a sentence in the first instance or higher which found the homicide and its perpetrator to be proven. The forensic medical report was then reviewed, together with the appended documentation and the case sentence. The set of documentation reviewed included the clinical history of the women who were examined, their relevant sociodemographic history and the proven facts of the homicide, as well as the circumstances of the latter.

The sociodemographic data corresponding to the women studied were gathered: their age, nationality, work status at the time of the crime (employed, unemployed, or retired, as well as whether they were economically dependent on their partner or did not control their money even though they were in work. The latter point was reported by the women before being confirmed during the hearing, and their educational and socioeconomic levels were also recorded.

Their socioeconomic level was inferred from their type of employment prior to and at the time of the crime, as well as the average wage associated with the work in question combined with data on their possessions and lifestyle obtained during the hearing and their clinical history, as well as what the women said about their perceived economic status during the forensic medical examination.

Data about the crime were also reviewed: the relationship of the women with their victims, the mechanism used to cause or attempt to cause death, whether the women had been threatened or attacked by their victims prior to committing the crime, and whether they had attempted suicide after the crime. If they had attempted suicide, then the severity of this was assessed on the SALSA scale as slight, moderate, or severe. This scale was selected as it makes it possible to measure the severity of attempted suicide retrospectively on the basis of the available clinical history, without the need for additional interviews.¹⁴

The relevant psychobiographical data of the women studied were also complied: a history of sexual abuse, a history of physical or psychological mistreatment by family members, a history of alterations during their childhood (such as the traumatic divorce of their parents, abandonment by their father, or an excessively strict upbringing), episodes experienced as traumatic that could not be classified in another category (such as robberies or traffic accidents), a criminal record or significant behavioural alterations in infancy or adolescence (detentions, expulsion from school, time spent in detention centres for minors, or similar).

Their psychiatric history was investigated: diagnosed mental disorders, a family history of mental disease or substance abuse, alterations in their intellectual capacity and brain development, a diagnosis of personality disorder or suspicion of the same. For the purposes of the study, more than two references in their clinical history to the suspicion of personality disorder was classified as such, even when the clinical history contained no definitive diagnosis of this without completely ruling it out.

Additional data were gathered on their history of substance abuse: intoxication at the time of committing the crime, the habitual consumption of any substance, whether such consumption could fulfil the criteria for abuse, dependency, or remission, whether the woman fulfilled the criteria for duel disease (mental disorder together with disorder due to active consumption) and whether they had sought specific treatment for addiction.

Lastly, data were obtained on legal proceedings: if the subjects had admitted the crimes they were accused of or not, the conclusions of the forensic medical examination respecting their capacity, and the conclusions of the judicial sentence.

The data evaluated were analysed descriptively.

This study was authorised by the Legal Medicine Institute of Valencia.

Results

Sociodemographic data

Age: the women studied ranged in age from 20 to 54 years old, with an average age of 36.50 years and a median age of 37.17 years.

Nationality: the majority of the women were Spanish nationals (10 cases, 55.6%), 5 (27.8%) were from Latin American countries, and the remaining 3 were from Western Europe, Eastern Europe, and Asia, respectively.

Employment: 11 of the women (61.1%) were in work at the time of the crime. Four of them (22.2%) were economically dependent on their partner independently of their work status, two of them were unemployed and one was a pensioner due to disability.

Educational level: Five of the women studied (27.8%) had completed secondary education, four of them had finished higher secondary education or higher Technical College training. Four of them had finished primary education and three had a University degree or higher. Only two of them had not finished their primary education, although they were literate. No woman was found to be functionally illiterate.

Socioeconomic level: the great majority of the women (14, 77.8%) were at a socioeconomic level that is considered to be low, while the remaining four were at middle level. No women with a high socioeconomic level were found.

Personal histories

Alterations during childhood: three of the women (16.7%) reported events which had had a negative impact on their childhood: parental violence, an emotionally distant educational style, and the conflictive divorce of their parents.

Sexual abuse: none of the women studied had been the victim of sexual abuse.

A history of mistreatment: Four of the women (22.2%) mentioned that they had suffered mistreatment by their parents. In two cases, this was solely physical, and in the other two, it was a combination of physical and psychological mistreatment.

Other traumatic experiences: Two of the women (11.1%) reported traumatic experiences that were not included in the categories examined: one of them had been mugged, and the other refused to specify the type of event.

Criminal record: only two of the women (11.1%) had a criminal record, both of them for violent crimes.

Behavioural alterations during infancy: when they were asked about detentions or stays in reformatories or similar institutions, two of the women (11.1%) said that they had had experiences of this type (expulsion from school and a period of time in youth custody).

Characteristics of the crime

Relationship with the victim: 38.9% of the victims were not directly related to the subjects studied (two people were unknown, two were known, one was a lover of the husband, a son-in-law, a minor and an elderly person they were looking after); four of the victims (22.2%) were children of the perpetrators, three of them were their partner (two husbands and one partner they were not living with), two ex-partners, one mother, and a grandchild.

Mechanism: the majority of the women carried out the homicide or attempted homicide with a sharp instrument (7 cases, 38.9%), followed by suffocation (5 cases, 27.8%), impacts and blunt instruments (4 cases, 22.2%). There was also one case of attempted homicide using petrol and fire, and another consisted of drowning in a body of water.

Threats received previously: none of the women had been threatened previously by the victims.

Previous violence: three of the subjects studied (16.7%) had suffered violence by the victims. One of the latter was a husband, one an ex-partner and one an acquaintance.

Attempted suicide after the crime: three of the women attempted suicide (16.7%). Based on the perceived lethality of the method, its actual lethality, and the possibilities of rescue, two of the attempts were described as true suicide and one was described as parasuicide.

Psychiatric history

Mental disorder: 10 cases (55.6%) had received one or more diagnoses of mental disorder. A total of six diagnoses of mood disorder were found (five depressions and one bipolar disorder, which at the time of the crime was in normal phase), four psychotic spectrum disorders (two schizophrenia, one toxic psychosis, and one puerperal psychosis), four anxiety disorders, two eating behaviour disorders (one anorexia nerviosa and one bulimia), a conversion disorder, and a dissociative disorder.

Family history: four of the women (22.2%) stated that they had a family history of mental illness or problematic consumption of psychoactive substances.

Intelligence: 17 of the 18 women (94.4%) had a normal level of intelligence, and only in one case was the level of intelligence borderline.

Neurological development alterations: one case (5.6%) had a history of epilepsy, while the other women had undergone normal neurological development.

Personality disorder: two cases (11.1%) had been diagnosed with personality disorder. One of these was in cluster A (paranoid) and the other was in cluster B (limit). Six cases (33.3%) had a history of personality disorder after review of their clinical history and forensic medical examination. Nevertheless, there was no confirmatory diagnosis and no suspicions were ruled out in the clinical history. In four cases, there was a suspicion of cluster B disorder, one of cluster B and one of cluster C.

Drug consumption history

Intoxication at the time of the crime: after the forensic medical analysis, four of the women (22.2%) were found to have been intoxicated by one or more substances at the time of the crime (two cases of alcohol intoxication, one case of alcohol + THC and one case of alcohol + THC + cocaine).

Current habitual consumption: eight of the women (44.4%) habitually consumer a psychoactive substance.

Abuse/dependency/remission: in three cases (16.7%), the women were found to fulfil the criteria for the abuse of one or more psychoactive substances. One of these cases was also found to fulfil the criteria for dependency. No disorder due to withdrawal was observed.

Dual disease: in one case (5.6%), criteria for dual disease were observed (depression + alcohol, THC and cocaine abuse and dependency).

History of treatment for consumption: three of the women (16.7%) had been treated in the past by Addictive Behaviour Units, and one of them had also made one attempt at hospitalised detoxification.

Legal process

Admission of guilt: 12 women (66.6%) admitted that they had committed the crime before their trial. 5 of them (27.8%) denied it, and only 1 (5.6%) referred to amnesia about the situation.

Conclusions of the forensic medical report: no circumstance that could modify criminal responsibility was detected in nine of the cases (50%). In four cases, a mitigating factor due to psychiatric reasons was found (22.2%), in three cases, there was an incomplete mitigating factor (16.7%) and in two cases, there was a complete mitigating factor (11.1%).

Conclusions of the sentence: all of the sentences except two (11.1%) coincided with the forensic medical reports. In both of these cases, although the forensic medical report concluded that mitigating factors existed for psychiatric reasons, the sentence concluded that the said mitigating factors did not exist.

Discussion

Characteristics of the crime

Interesting results are found if we analyse the circumstances of the homicides. Studies tend to find that the homicides and murders committed by women share certain characteristics.^4,15 In the series of cases which we analysed, we find certain results that are consistent with these studies, such as the lack of planning. We can only be sure that one of the cases had been planned, and there are doubts about one other case, while the others (n = 16) were committed without any previous planning. We also found that the majority (n = 16) of the victims were known to the perpetrators of the homicide. The majority of victims were found to be dependent (n = 7), as well as partners and ex-partners (n = 5).

However, on the contrary to the said studies, very few of the perpetrators had been attacked by their victim in the past (n = 3), and we found no case in which they had been threatened. Respecting attempted suicide after the crime, it should be underlined that both attempts which were considered to be serious (with a mechanism perceived to be highly lethal, a highly lethal mechanism was used, and with little expectation of rescue) took place in cases where the victim was a child of the women. This leads us to think that there was an intention to commit expanded suicide.

With respect to the mechanism used, our study seems to partially confirm the findings of previous studies.¹² Sharp instruments were used in all of the cases except one (case 14, puerperal psychosis) in which there were psychotic symptoms. This relationship does not seem to be so marked in cases of mood disorders and strangulation/asphyxia/drowning; the latter methods, together with blows, seem to be more associated with physical superiority over the victim (7 of the 10 victims of these methods were children or dependent adults).

Sociodemographic data

If we compare the ages of the women in our study with the latest report on homicide prepared by the Ministry of the Interior, we find that their average age is closer to that of men than it is to that of women. We also find that the percentage of Spanish perpetrators in our study is lower (55.6% vs 63.5%).¹⁶ It should be pointed out that, although the majority of the women were in work, almost all of their jobs were precarious. This, combined with another series of factors (poor management of the family economy in case number 7, or the irregular situation of case number 11) means that the majority of the women were at a low socioeconomic level. A low socioeconomic level is a stress factor that has been associated with increased violence, especially against a partner.^17,18 Their educational level was higher than expected: more than half of the women had gone higher than secondary education, and this is unusual in the samples of individuals who have committed homicide.^19–21

Personal history

This is one of the sections in which we found a greater difference between our sample and the models established by the above-mentioned studies. We found no case of sexual abuse and few cases of mistreatment (n = 4), relevant alterations during childhood (n = 3), and other events which are considered to be traumatic (n = 2). This contrasts with findings that indicate that women who commit homicide have a higher number of traumatic psychobiographical factors or that they have been the victims of abuse.^4–6 We find few previous convictions (n = 2) or significant behavioural alterations during infancy (n = 3). This finding is consistent with studies which indicate that women who have committed homicide tend to have fewer previous convictions than men.^22,23

Psychiatric history

The abundance of psychiatric diagnoses among the cases studied (n = 10) agrees with the findings which indicate that women sentenced for homicide are more likely to be diagnosed with a mental illness than the equivalent men.¹³ Nevertheless, distortion of the sample must be taken into account, given that not all cases are subjected to expert assessment, which only occurs when there is the suspicion of a mental disorder. There was an active psychosis in cases 5, 9, 14 and 16, and the motive for the homicide was associated with the psychotic symptoms of the women who were studied. Case 17 underwent a dissociative episode involving religious delirium which led the women to commit the crime in question.

The prevalence of a family history or alterations in intelligence and neural development are slight and seem to be irrelevant. Respecting personality disorders, we only found one diagnosis of a cluster A (paranoid) disorder and another cluster B (limit) disorder. This contradicts findings which chiefly associate cluster B with propensity to use violence.²⁴ However, when the individual histories are analysed, there is the suspicion of an undiagnosed cluster B disorder in four additional cases, while there is only one case with the suspicion of undiagnosed cluster A and C disorder for each cluster. This may be linked to the difficulties in diagnosing a personality disorder without prolonged examination, or the suspicion may reflect the observed tendency to overdiagnose disorders in women who have displayed criminal forms of behaviour.^1,5

Drug consumption

There are very few cases of acute intoxication by substances at the moment of the crime in comparison with cases of homicide in general.²⁵ Respecting the habitual consumption of substances, this is similar to that in the general population of women in Spain.²⁶ It is striking that of the four cases when there was intoxication at the moment of the crime, three of the women also met the criteria for the abuse of at least one substance, while one of them also fulfilled the criteria for dependency and dual disease, which may be linked to increased risk of violence.²⁷ Of the three women who fulfilled abuse criteria for one or more psychoactive substances, two had been treated in an addictive behaviour unit, and one of them had undergone detoxification in hospital, followed by relapse. No information was supplied on the causes of this relapse during her examination.

Legal process

The fact that two-thirds of the women admitted their responsibility is in marked contrast to what has been observed in male samples, where only 50% admit this.²⁷ The lack of statements that they had forgotten, which is suspected to be a defence strategy, is also striking. Only the woman in case number 9 stated that she had forgotten the events, and this was a case of toxic psychosis in which forgetting is a symptom that can be expected. 25% of a male sample stated that they had forgotten the events, with the suspicion that some of these consisted of what is known as “judicial amnesia”.²⁷

Discrepancies were only found between the forensic medical assessment and sentencing in two cases, in which the forensic report considered that there was an analogous mitigating mental health factor which the sentence ignored. This level of discrepancy between the report and sentence is similar to what is found in male samples.²⁷

Limitations

Several limitations arise when interpreting the results, and these are intrinsic to studies of a series of cases. The small sample size (n = 18, in spite of having taken the samples over a relatively long period of time, 10 years) prevents us from performing a statistical analysis, which would have given more conclusive results. The selection of the sample was also distorted by the choice of cases in which women sentenced for homicide had also been assessed by the Legal Medicine Institute of Valencia. Although both of these inclusion criteria are useful in that they allow us to know all of the details of the case and legal process, they exclude cases in which there was no expert assessment from the study together with those pending trial or sentence, and those which are still being heard. Nor was it possible to access the files of other legal medicine institutes in Spain. Lastly, the retrospective nature of this study restricts the information to what is shown in the documentation that was consulted, preventing the performance of tests which may have enriched the study, such as scales or inventories.

Final considerations

Regarding the differences found between our study and previous ones, it stands out that we found very few histories of aggression or threats by the victims against the women sentenced for homicide. We also found few cases with a history of trauma. Although it would not be ridiculous to think that in a society with increasingly flexible gender-based roles, gender characteristics in cases of homicide would also become less rigid (with increasing characteristics intrinsic to homicides committed by men in those committed by women, and vice versa), we have no objective data that would support this hypothesis. Further studies are required, preferably prospective one with a larger sample, to study the findings in this series of cases.

Financing

This study received no specific grant from any agency in the public, commercial or not-for-profit sectors.

Conflict of interests

The authors have no conflict of interests to declare.