Feminist approach to Femicide

The feminist streams in Latin America claims that the main reason of the cases of feminicides are associated with structural gender inequality and the impunity of the perpetrators in the justice system, the theory places crimes against girls and women in the patriarchy and considers them the extreme of gender domination against women (^{Lagarde, 2008}). Lagarde (2000) maintains: “Femicide is genocide against women and it happens when historical conditions generate social practices that allow violent attacks against the integrity, health, freedoms and lives of girls and women” (p. 216).

Feminist approaches agree when relating cases of feminicide with sociocultural and political factors, so that violence against women is the product of a structural system (^{Carcedo, 2000}). In the same perspective (^{Chow and Berheide, 1994}) argue that societies in the world are characterized by gender inequality, which has its roots in the sexual division of labor and is perpetuated by the process of gender socialization. In the ideas of ^{Sagot (1994}) in a patriarchal society, the transmission of the ideology of oppression is the main element of socialization.

Ecological approach to Femicide

In this position, acts of feminicide are attributed to the individual characteristics of the woman (such as having witnessed violence in the home, having been a victim of violence and having had an absent father), at the microsystem level (such as male domination in the family, the control of money on the part of the man, the consumption of alcohol and marital and verbal conflicts) and the level of exosystem (such as unemployment, low socioeconomic level and friends of the criminal world in the man). The ecological approach to abuse conceptualizes violence as a multifaceted phenomenon based on an interaction between personal, situational, and sociocultural factors (^{Heise, 1998}; ^{Krug, et al., 2002}) and femicide in general. The authors point out the complexity of the factors related to violence against women:

Violence is the result of the complex interaction of individual, relational, social, cultural and environmental factors. Understanding how these factors are related to violence is one of the important steps in the public health approach to preventing violence (^{Krug et al., 2002}, p. 12).

Violence has long-term consequences for these women and their children, as well as social and economic costs for the whole of society, as WHO (2012) argues:

Femicide committed by someone without an intimate relationship with the victim is known as non-intimate femicide, and femicide that involves sexual assault is sometimes referred to as sexual femicide. Such killings may be random, but there are disturbing examples of systematic killings of women, particularly in Latin America (p. 4).

The comprehensive and multidimensional approach used by the ecological model is confirmed by scientific research on mortality in the field of public health. ^{Gómez (2008}) affirms that "violent deaths by homicide are avoidable, as evidenced by industrialized countries where homicides have notably decreased through preventive public policies aimed at reducing social inequalities through control of social, cultural and economic determinants" (p. 83)

Political-legal approach to Femicide

According to CEDAW (1992) The General Recommendation No. 19 of 1992 was historic, since it clearly framed violence against women as a form and manifestation of discrimination based on gender, used to subordinate and oppress women. Without a doubt, it brought the violence out of the private sphere into the sphere of human rights; General Recommendation No. 35 is also a milestone:

Recognizes that the prohibition of gender-based violence has become a norm of customary international law; It broadens the understanding of violence to include violations of sexual and reproductive health rights; Stresses the need to change social norms and stereotypes that support violence, in the context of a resurgence of narratives that threaten the concept of gender equality in the name of culture, tradition or religion (CEDAW, 1992, p. 2).

The 1993 World Conference on Human Rights was considered a success by feminist activists, broadening the international human rights agenda to include gender-specific violations, some of which were identified as human rights violations (^{Sullivan, 1994}, p. 152). Since 1994, the Convention of Belém do Pará was adopted, which clearly establishes that “violence against women is an offense to human dignity and a manifestation of historically unequal power relations between women and men”, and reiterated that every woman has the right to live without violence and any form of discrimination. (OAS, 1994; art. 6). Since the adoption of the Convention of Belém do Pará, the Latin American countries have entered a process of convergence and the gender perspective was included in both into public policies in general and national legislation (^{Galanti and Borzacchiello, 2013}).

According to a study conducted by ^{Carcedo (2000}) indicates that “the regulation of a crime should provide a legal instrument that allows women to access protection and request help from the authorities when they are subjected to violence” (p. 72). At that regard Saccomano (2017), concludes:

It is found that the criminalization (or typification) of femicide is not significant to predict the rate of femicide; On the other hand, low levels of the rule of law and the lack of representation of women in decision-making bodies, such as national parliaments, appear as the most relevant factors to explain the variation in trends in femicides (p.51)

Magnitude of the cases of attempted and femicide in the country

Femicide cases in the country

According to the country's literature, the first data recorded show that most of them are perpetrated by couples, ex-partners or close relatives. According to Centro de la Mujer Peruana Flora Tristán points out that "more than 64% of the victims at the time of the attack had a sentimental, affective or intimate relationship with their aggressor" (Citado en Defensoría del Pueblo, 2010)

Chart 01 Number of feminicide victim cases recorded according to years in the country 

There are few studies on femicide-related factors in Peru, most of which are reports of statistics and specific measures or proposals (^{Villanueva, 2009}; ^{Viviano, 2010}) and have a legal approach (^{Dador, 2012}; ^{Estrada, 2011}). The largest cases of femicide in the country are in the regions such as: Metropolitan Lima, Arequipa, Callao, Cusco, Huánuco, La Libertad and Puno are the regions (see Chart 01). According to (MIMP-CEM, 2019), the evolution of femicide cases in the country are quantitatively high between 2009 and 2019, among those years 1292 women of different ages and social status have lost their lives.

Table 1 Cases of femicide victims by area of occurrence 

Area	2019 a /		2018
	No.	%	No.	%
Urban	90	64%	76	51%
Rural	44	31%	31	twenty-one%
Marginal urban	6	4%	19	13%
It is unknown	0	0%	2. 3	fifteen%
Total	140	100%	149	100%

a / Cases of victims of femicide that occurred as of October 31, 2019

Source: Data registered by the CEM in the country 2019

When the phenomenon is analyzed according to the urban and rural context, the highest percentage of cases occur in the urban area with 64%, while in rural area it represents 31% and a percentage less than 4% in the marginal urban area (see table 1). Now, considering data from the registry of (MIMP-CEM, 2019) that counts people who are victims of femicide and attempts, it is noted that a use of violence with a strong gender bias (men towards women) prevails in the country. The modality of cases of victims of femicide, the highest percentage is by stabbing (23%), followed by asphyxia / strangulation (29%) and the modality by bullet shots represents (15%) respectively (see table 2).

Table 2 Modality of the case of the victim of femicide 

Modality	Femicide
	No.	%
Stabbing	32	23%
Various blows	16	11%
Bullet shot	21	15%
Poisoning	5	4%
Unraveling	0	0%
Choking / strangulation	41	29%
Run over	1	1%
Burn	4	3%
Other	20	14%
Total	140	100%

Source: Data registered by the CEM in the country 2019

Another aspect to consider is the scarcity of accurate data on the conditioning factors of femicide in the country and in the Puno region. National statistics do not record the motive, nor the different types of violence that women suffered before being murdered. However, the data is available according to the relationship that the victim had with her perpetrator. Faced with such a situation, it is necessary to specify that 68.48% of the cases, the aggressors (femicides) are their current partners, and 26.19% of the aggressors have been their ex-partners and 18.13% of those reported are someone known to the victim and 16.11% are people unknown to the victim (see graph 02).

Chart 02 Cases of victims of femicide according to relational link 

Conditioning factors of femicide cases in the Puno region.

From birth, men and women participate in different socialization processes. Men are taught to adopt values, assumptions, behaviors and stereotypes that are assumed to be inherent to their sex. Likewise, within this socialization process, they are taught how to relate to other people (^{Tamayo, 1990}; Defensoría del Pueblo, 2010). In general, the men who commit acts of feminicide in the country do not report having consumed any psychoactive substance, including alcohol or drugs, the figures are evident in this regard, 51% have not consumed alcohol or drugs at the time of committing the act of femicide, while 27% of the aggressors have been under the influence of a drug such as alcohol (see table 3).

For analysis of the factors related to the phenomenon of “feminicide”, the bivariate logistic regression model was designed, having as the dependent variable the total number of cases of the people who were victims of feminicide and the independent variables will be each of the sociodemographic variables, socioeconomic, individual.

𝑌 = 𝛽0 + 𝛽1. 𝑋_ij + 𝜀_ij

Table N ° 4 shows the risk values of being the object of violence and with subsequent femicide it increases with a higher level of education. In other words, women with a year of higher education have a 2.9 times more likely to be victims of feminicide than those with a lower level of education. Something similar happens with women who have a high school education, have 2.4 times the risk of being subjected to feminicide than those who have a primary school.

The values of the bivariate logistic regression model show us that younger people have a higher risk of being victims than older people. The values indicate that the 18 to 29 year old group has a 4.7 times higher risk of being victims of femicide than the 15 to 17 year old group. Another group of ages between 30 to 59 years of age also registers Odds ratio (OR) values of 2.1, which implies that women who are in this age group are 2 times more victims of femicide than those who are younger than 17 years. Similarly, women in the age group 60 years of age and older have a 0.8 times lower probability of being a victim of risk of femicide than other age groups (see table 5).

The values of the bivariate logistic regression model, it can be observed that relatively young men have a greater probability of committing the act of feminicide towards a woman. The values indicate that the age group from 30 to 59 years old is 8.2 times more likely to commit the act of feminicide, while the age group from 18 to 29 years old is 5.0 times greater than the age group from 14 to 17 years old. (see table 6).

Work activity is one of the variables of equal importance for the study, table 7 shows us that as the labor insertion of women in work activities increases independently, acts also increase the risk probability of being a victim of femicide. In other words, the values indicate that women who worked independently have been victims of femicide in 2.2 times more than other occupations. In the same way, women who were engaged in domestic activity have been victims of feminicide in 1.9 times more than other occupations. Therefore, as women become independent from dependence on men by obtaining a salary, the risk of suffering as a victim of violence with the risk of femicide is high; which indicates that as women's wages increase, the risk of being a victim increases.

The regression estimates on the individual factors of the aggressors are mostly statistically significant (p = 0.000) which indicates that if the occurrence of feminicide in women can be predicted according to the silver model. Table N ° 10 shows the values of Odds ratio (OR), the factors as the aggressor's state condition (in a sober state) at the time of the act of femicide is 3.7 times greater than the consumption of some psychoactive substance (alcohol or drug), in the same way, the factor “having been a victim of violence in childhood” is 2.1 times greater than the other factors under study. Another factor that is significantly valid is “having witnessed violence in home” showing that aggressors who have witnessed violence in home are 1.3 times more likely to commit the act of femicide (see table 8).

Men who have a formal occupation are less likely to commit acts of feminicide, however, the opposite happens with those who are inserted in occupations or informal activities, the logistic regression values - Odds ratio (OR) indicate that the probability committing increases 3 times higher in those men who have informal occupations. Next are those people (men) without occupation with a value of 2.8, which means that as the man has a stable formal job, the probability committing acts of femicide will decrease (see table 9).

Motivations and beliefs of the aggressors

All kind of violence against women, are always followed by a risk of femicide, so it is important to analyze the causes and consequences. Violence against women is a historical and socio-political construction. Men do not always beat their women and, if they do, the beatings do not always respond to the same reasons. Definitions of what constitutes unacceptable or acceptable physical aggression vary with changes in notions of sexual roles and the sexual division of labor, and their implications for the natural and social order (^{Bourdieu, 2000}), with the transformation of the organization of sexuality in the family and in society (^{Tinsman, 1995}). The motivations and beliefs of the aggressors are more frequently motivated by jealousy, infidelity and revenge respectively (see table 10).

The model Multivariate shows the results of odds ratio (OR) similar to those described above, which includes the beliefs and motivations of men who attack women until committing femicide. The regression values indicate that jealousy and alleged infidelity of woman are triggers to commit the act of feminicide in the region. This probability is very high according to the values estimated by the logistic regression model (6.6 and 5.1 odds ratio), this means that jealous men are 6 times more likely to commit femicide compared to those who do not have that kind of feelings and 5 times greater probability of risk committing the act of femicide motivated by infidelity acts from the couple (woman). Other motivations such as a feeling of revenge, separations and refusal of a romantic relationship are also grounds for male sociodicea, to justify acts of this type (see table 11).

The harmony between objective structures and internal subjective ones, hide the historical and social conditions that made possible its realization under the veil of toxic experience. Comparing the results of the studies on the attempt and femicide, most of the authors coincide in stating that such behaviors are abnormal and associated with structural, individual, and psychosocial factors (^{Berger and Luckman, 1986}; ^{Bourdieu, 2000}). When the causal factors are verified, some authors already mentioned above, agree that the phenomenon of feminicide has as its base history similar characteristics such as jealousy, sense of ownership, violent response to infidelity or the end of the relationship. From the theoretical point of view, femicide has not become independent from what has been formulated in relation to general violence against women (^{Ellis and Dekeseredy, 1998}; ^{Lagarde, 2008}). On the other hand, studies such as ^{Shalhoub-Kervorkian and Daher-Nashif (2013}); Fulu and Miedema (2015) argue that in the absence of theoretical independence, opinions in favor of having objectively relevant explanatory frameworks have not diminished, which helped decolonialize the cultural approach to femicides and integrate them into causes linked to the processes of globalization and of the world system.