Abstract

This study analyses the relationship between femicides and indicators of socio-economic condition, demography, access to communications, and health situation, in Brazilian state capitals and large-population municipalities. It is an ecological study using the standardized mean coefficient of female mortality due to aggression as a marker for femicide in the years 2007–09 and 2011–13. The Pearson Correlation test was used for the statistical analysis between the outcome and 17 independent variables, and those that were statistically significant (p < 0.05) were introduced into a multivariate linear regression model, using backward elimination. In the first three-year period the average rate of femicide was 4.5 deaths per 100,000 women, and in the second period it was 4.9/100,000. Poverty (β = –0.330; p = 0.006), Pentecostalism (β = 0.237; p = 0.002) and male mortality by aggression (β = 0.841; p = 0.000) were associated with femicides. The negative association between poverty and feminine deaths indicates a paradoxical relationship, in that women who die in the richer regions are mostly poor. A relationship was also found between gender violence, fundamentalist religious beliefs, and urban violence.

Homicides; Women; Gender; Violence against women

Introduction

Gender violence, a frequently-occurring social fact, takes place in a continuum in which the most severe consequence is the woman’s death. The concept of femicide was identified as murder of women due to their being women, a crime defined by Diana Russel¹1. Russel D. Femicide: The Politics of Women Killing. New York: Twayne Publisher; 1992. as a type of sexual terrorism, a social mechanism for keeping women under control, in a public masculine manifestation of power²2. Pasinato W.“Feminicídios” e as mortes de mulheres no Brasil. Cad. Pagu 2011; 37:219-246..

In the absorption of Russel’s original concept, there were countries that adopted the term ‘feminicide’ and others that opted for the term femicide. In Brazil the term ‘feminicídio’ is used, and it was recently given a definition in legislation, accompanying the legislation of other countries of Central and South America such as Costa Rica, Chile, Guatemala and El Salvador³3. Chiarotti S. Contribuciones al debate sobre la tipificación penal del feminicídio/ femicidio. Lima: CLADEM; 2011..

Gender culture and hierarchies in patriarchal society, as well as structural violence, are determining factors of femicide⁴4. Palma-Solis M, Vives-Cases C, Alvarez-Dardet C.Gender Progress and Government Expenditure as Determinants of Femicide. Ann Epidemiol 2008; 18(4):322-329.. The frequency of femicide is greater when there are social inequalities, armed conflicts, migration, gender discriminations and exercise of hegemonic and aggressive masculinity⁵5. Grana SI. Sociostructural considerations of domestic femicide. Journal Of Family Violence 2001; 12(4):421-435.

6. Prieto-Carrón M, Thomson M, Macdonald M.No more killings! Women respond to femicides in Central America. Gender Dev 2007; 15(1):25-40.^-77. Taylor R, Jasinski JR.Femicide and the feminist perspective. Homicide Stud 2011; 15(4):341-362..

Femicides may be categorized as intimate, when the perpetrator is a man known to the woman; non-intimate, with or without sexual violence; by connection, in which one or more women are killed in the defense of others, and executions of women perpetrated by gangs, mafia groups, drug dealers or members of the police⁸8. Fox JA, Zawitz MW. Homicide trends in the U.S.: 2002 Update. Washington: Bureau of Justice Statistics; 2004. Crime Data Brief.

9. Dobasch RM, Dobasch RP, Cavanagh K, Lewis R. Not an ordinary killer - just an ordinary guy: when men murder an intimate woman partner. Violence Against Women 2004; 10(6):577-605.

10. Schraiber LB, Gomes R, Couto MT.Homens e saúde na pauta da Saúde Coletiva. Cien Saude Colet 2005; 10(1):7-17.^-1111. Carcedo A. No olvidamos ni aceptamos: feminicídio em centro America. 2000-2006. San José: CEFEMINA; 2010..

Women have a higher risk of being murdered by their partners than men do. In the United States 42% of murders of women were committed by an intimate partner in 2012, compared to 5% of husbands killed by their wives⁸8. Fox JA, Zawitz MW. Homicide trends in the U.S.: 2002 Update. Washington: Bureau of Justice Statistics; 2004. Crime Data Brief.. Repeated and growing conjugal violence, and also threats of death after conflicted separation, are risk factors for women⁹9. Dobasch RM, Dobasch RP, Cavanagh K, Lewis R. Not an ordinary killer - just an ordinary guy: when men murder an intimate woman partner. Violence Against Women 2004; 10(6):577-605.

10. Schraiber LB, Gomes R, Couto MT.Homens e saúde na pauta da Saúde Coletiva. Cien Saude Colet 2005; 10(1):7-17.^-1111. Carcedo A. No olvidamos ni aceptamos: feminicídio em centro America. 2000-2006. San José: CEFEMINA; 2010..

Studies show that between 60% and 70% of murders of women are femicides and the victims are young, poor, members of ethnic minorities, migrants or sex workers – that is to say, those affected are predominantly vulnerable women¹⁰10. Schraiber LB, Gomes R, Couto MT.Homens e saúde na pauta da Saúde Coletiva. Cien Saude Colet 2005; 10(1):7-17.^,1111. Carcedo A. No olvidamos ni aceptamos: feminicídio em centro America. 2000-2006. San José: CEFEMINA; 2010..

Worldwide, femicides are tending to grow⁸8. Fox JA, Zawitz MW. Homicide trends in the U.S.: 2002 Update. Washington: Bureau of Justice Statistics; 2004. Crime Data Brief.^,1212. Waiselfisz JJ. Mapa da violência 2012. Atualização: Homicídio de Mulheres no Brasil 2012. Brasília: Centro Brasileiro de Estudos Latino Americanos; 2012.. Central America is one of the locations with the highest level of occurrence⁶6. Prieto-Carrón M, Thomson M, Macdonald M.No more killings! Women respond to femicides in Central America. Gender Dev 2007; 15(1):25-40.^,1111. Carcedo A. No olvidamos ni aceptamos: feminicídio em centro America. 2000-2006. San José: CEFEMINA; 2010.; in the United States femicide is one of the primary causes of death in women, and the highest in Afro-American women¹³13. Campbell J, Glass N, Sharps PH, Laughon K, Bloom T.Intimate partner Homicide. Review and implications of research and policy. Trauma, violence and Abuse 2007; 8(3):246-269.. In Brazil, in the last 30 years, murders of women have increased significantly: the rate has risen from 2.3/100,000 to 4.6/100,000 women¹²12. Waiselfisz JJ. Mapa da violência 2012. Atualização: Homicídio de Mulheres no Brasil 2012. Brasília: Centro Brasileiro de Estudos Latino Americanos; 2012..

Femicides occur more frequently in urban conglomerations and cities or regions with higher population density⁵5. Grana SI. Sociostructural considerations of domestic femicide. Journal Of Family Violence 2001; 12(4):421-435.. This study is an analysis of the relationship between femicides and a series of indicators in Brazilian cities with large populations.

Method

This is an ecological study that analyses the relationship between female mortality due to attack and variables indicating socio-economic and demographic status, access to communications and health, in capital cities of Brazil’s states, and in Brazilian municipalities with a population of more than 400,000 – a group of 58 cities. Secondary data from the Health Ministry (DATASUS), the Brazilian Geography and Statistics Institute (IBGE/SIDRA/PNAD) and the Economic and Statistical Foundation (FEE) were used. The data are for the years 2007–2009 and 2011–2013. The rates of female mortality due to aggression were standardized, using the standard population supplied by the World Health Organization for the period 2000–2025. The standardization makes it possible to carry out comparisons between regions with different demographic and age patterns, as is the case for the Brazilian state capitals and municipalities selected.

This study is part of a survey financed by the CNPq entitled “Femicides, and Other Murders Relating to Gender in Rio Grande do Sul”, in which three ecological studies were made with different territorial cross-sections. In the first, the distribution of feminine deaths by aggression was analyzed by states of the Brazilian Federation¹⁴14. Meneghel SN, Hirakata VN.Feminicídios: Homicídios Femininos no Brasil. Rev Saude Publica 2011; 3(45):564-574.; in the second, deaths were distributed in the micro-regions of the State of Rio Grande do Sul¹⁵15. Leites G, Meneghel SN, Hirakata VN. Homicídios femininos no Rio Grande do Sul, Brasil. Rev Brasileira de Epidemiologia 2014; 17(3):642-653.. This present study investigates the variables associated with feminine deaths by aggression in the most urbanized and densely populated regions of the country.

In Brazil, femicide is not specified in the death certificate, making it impossible to identify this crime through the secondary data obtained from the mortality system. Thus, the total of feminine deaths by aggression obtained in the SIM/DATASUS was used as an ‘approximate marker’ for femicide. The possible overestimation of deaths is compensated by sub-notifications of life events, due to the failings in the coverage of the information system.

Feminine homicides correspond to the range X85 to Y09 in the External Causes chapter of the 10^th edition of the ICD-10. The explanatory variables comprised 17 indicators, most of them expressed in percentages, considered in their totality or differentiated by gender and grouped in four thematic blocks:

• Economic: Gini Index; Human Development Index; average monthly income; poverty; women as head of family.

• Demographic: Woman’s marital status; illiteracy; people not born in the municipality; race/color; religion.

• Access: Home has internet; homes with fixed or mobile phone.

• Health: AIDS mortality rate; cervical and breast cancer mortality rate; male rate of mortality by aggression; proportion of deaths from undefined causes; doctors per 1,000 population.

The statistical analysis was made using the SPSS program, version 18.0, using as outcome the standard mean coefficient of female mortality by aggression for the two three- year periods. The statistical association between the indicators was verified using the Pearson Correlation test, and the variables with statistical significance (p < 0.05) in the bivariate analysis entered into the multiple linear regression model, using backward elimination. Having in mind that the variables that remained in the final regression model were the same for the two three-year periods, the more recent statistical analysis of the outcome (2011-2013) will be presented.

The survey was conducted within the standards required by the Helsinki Declaration, and the project was approved by the Ethics Committee of the School of Public Health of Rio Grande do Sul.

Results

There were 4,368 female deaths by aggression in the years 2007-2009 and 4,834 in the years 2011–2013, in the 58 municipalities of this study, representing an increase of 10% between the two three-year periods. These locations comprised 1% of Brazilian municipalities, 33% of the feminine population and 39% of female deaths by aggression.

The average coefficient of female mortality by aggression in the first three-year period was 4.5 deaths/100,000 women, and in the second period was 4.9/100,000 women. In 58% of the municipalities there was an increase in the rates between the two periods. The lowest rate found was 0.7/100,000 in the municipality of São José do Rio Preto, São Paulo State, and the highest was 16.3/100,000 in the municipality of Serra, Espírito Santo State, both in 2007–2009.

The coefficient was higher than five deaths per 100,000 women in 20 cities in the first three-year period, and in 25 in the second: 10 of these were capital cities of Brazilian states: Rio Branco, Porto Velho and Boa Vista in the north of the country; Maceió, Salvador and João Pessoa, in the Northeast; Cuiabá in the Center-West; Belo Horizonte and Vitória in the Southeast; and Curitiba in the South. The State of Espírito Santo had the municipalities with the highest national rates: Vila Velha and Serra, although they showed a reduction from the first period to the second (Table 1).

Table 2 gives the averages of the two three-year periods for the social-democratic characteristics of the victims and the location of frequency of the crime. There was a predominance of black women (whose risk of dying was twice that of white women), single women (70.3%), young women (72.0% aged 10-39) and women with low level of schooling, although in 6.3% of the crimes women had more than 12 years of schooling. 22% of the deaths took place at home.

Table 3 describes the independent variables of the study listed in four blocks: socio-economic, demographic, access to communications, and health. The means, standard deviations, maximum and minimum values, and the source of the data were presented for the two periods of the study.

Table 4 presents the results of the Pearson correlation test between coefficient of female mortality by aggression and the independent variables in the three-year period 2009–2011. There was a negative association between female mortality by aggression with women who were separated, and proportion of poor women. A positive association was found between female mortality and negro population, people of Pentecostal religion and masculine mortality by aggression.

Table 5 shows the results of the multivariate analysis between femicides, and the variables that were included in the final regression model: poor, black, separated women, Pentecostal religion, and male mortality by aggression. After the adjustment of the regression model, three variables remain significant; there was negative correlation with poverty (p = 0.006), and positive correlation with Pentecostal religion (p = 0.002) and male mortality by aggression (p < 0.000).

Discussion

This study showed an increase in femicides in Brazilian state capital cities and major cities in the period analyzed, principally among young, poor, black and single women and those with a low level of schooling. Further, a relationship was found with poverty, evangelical religion, and male homicide. The association between femicide, areas with high prevalence of evangelical religion, and male violence has been discussed in the past in other surveys¹⁴14. Meneghel SN, Hirakata VN.Feminicídios: Homicídios Femininos no Brasil. Rev Saude Publica 2011; 3(45):564-574.

15. Leites G, Meneghel SN, Hirakata VN. Homicídios femininos no Rio Grande do Sul, Brasil. Rev Brasileira de Epidemiologia 2014; 17(3):642-653.^-1616. Machado MDC.Representações e Relações de Gênero nos grupos Pentecostais. Rev Estudos Feministas 2005; 13(2):387-396.. The inverse relation with poverty is a paradoxical finding, in that various surveys on femicides indicate that poor women are more affected¹¹11. Carcedo A. No olvidamos ni aceptamos: feminicídio em centro America. 2000-2006. San José: CEFEMINA; 2010.^,1313. Campbell J, Glass N, Sharps PH, Laughon K, Bloom T.Intimate partner Homicide. Review and implications of research and policy. Trauma, violence and Abuse 2007; 8(3):246-269..

This contradiction can be explained by the fact that in more conservative communities, such as in small towns in the interior of Brazil, the rules of gender are more rigid and women remain submissive, strictly complying with the roles designated to them by the culture. This condition causes them to put up with situations of violence for longer periods of time, to avoid conflicts and reduce the risk of dying, in contrast to the large cosmopolitan cities, where the traditional standards of gender are more flexible¹⁷17. Sagot M, Carcedo A. Ruta crítica de las mujeres afectadas por la violencia intrafamiliar en América Latina: estudios de caso de diez países. Washington: Organización Panamericana de la Salud; 2000..

In a patriarchal society, femicide tends to be a punitive and disciplinary act, practiced against the victim who has become vulnerable as a result of having carried out an attack against male honor, or because they do not have protection, or, further, due to behaving in a way considered to be morally inappropriate. Thus, assassinations of women take place against a women who has left her place, that is to say, her subordinated position in a status-based system. The movement of the woman to a position not destined for her in the hierarchy of the traditional model challenges the position of the man in this structure and the moral codes stipulate that she should be punished or even killed¹⁸18. Saffioti HIB.Contribuições feministas para o estudo da violência de gênero. Cad. Pagu 2001; (16):115-136..

In today’s world, in various countries of the Americas and many regions of Brazil, in areas of extreme inequality, armed conflict, land-grabbing, frontier regions, shanty towns and circumstances where the law is that of a second state, femicides have come to be practiced as a form of exemplary punishment, demonstration of power or a message to women to behave themselves and to other men to show who is in the command¹¹11. Carcedo A. No olvidamos ni aceptamos: feminicídio em centro America. 2000-2006. San José: CEFEMINA; 2010.^,1414. Meneghel SN, Hirakata VN.Feminicídios: Homicídios Femininos no Brasil. Rev Saude Publica 2011; 3(45):564-574.^,1515. Leites G, Meneghel SN, Hirakata VN. Homicídios femininos no Rio Grande do Sul, Brasil. Rev Brasileira de Epidemiologia 2014; 17(3):642-653.^,1919. Gartner R, Baker K, Pampel FC.Gender stratification and the gender gap in homicide victimization. Social Problems 1990; 37(4):593-612..

Women who acquire sexual and economic autonomy test the traditional patterns of gender and there is a greater risk of femicides, because they place themselves against, or rupture, situations of subordination. Thus, the change in the traditional roles of gender in countries, regions or cities where there is not economic and gender equity increases the risk of victimization or death¹⁰10. Schraiber LB, Gomes R, Couto MT.Homens e saúde na pauta da Saúde Coletiva. Cien Saude Colet 2005; 10(1):7-17.^,1919. Gartner R, Baker K, Pampel FC.Gender stratification and the gender gap in homicide victimization. Social Problems 1990; 37(4):593-612..

Violence operates as a mechanism of control and subordination, and the ideology of gender, which is present in the discourses of the social institutions, naturalizes and cements this system. Thus, there is a patent contradiction between increased gender violence in richer locations, major centers, centers of industrialization and development – where groups of women make use of the more egalitarian and autonomous situations – and the situation in which the number of poor women, in a situation of vulnerability and social inequality, the principal victims of femicides, increases²⁰20. Faludi S. Backlash. O contra-ataque na guerra não declarada contra as mulheres. Rio de Janeiro: Rocco; 2001..

The analysis of religious inclinations indicated, both in this study and in the previous one¹⁴14. Meneghel SN, Hirakata VN.Feminicídios: Homicídios Femininos no Brasil. Rev Saude Publica 2011; 3(45):564-574., that there are more feminine deaths in territories with a high presence of people practicing evangelical Pentecostal religion. The Pentecostals share an identity of traditional values, in which the women are controlled by the men of the religious community. This vigilance maintains the gender order and, if they do not obey the rule, they will be punished, with the consent of the group²¹21. Almeida R.Religião na Metrópole Paulista. Rev Brasileira Ciênc Sociais 2004; 19(56):15-27.. Among evangelicals there is a patriarchal division of the genders, which reinforces women’s position of subordination to masculine authority in all instances: home, work and church. The regulation of conduct provided by the evangelical religion aligned to the patriarchal gender order is usually permissive on aggressions, and complacent with the perpetrators of crimes, a fact which stimulates the emergence and continuation of acts of violence¹⁷17. Sagot M, Carcedo A. Ruta crítica de las mujeres afectadas por la violencia intrafamiliar en América Latina: estudios de caso de diez países. Washington: Organización Panamericana de la Salud; 2000.^,2222. Rojas ZO. La iglesia pentecostal: Comunidade e Mujeres. Revista Cultura y Religión 2009; 3(2):51-66..

The phenomenon known as backlash²⁰20. Faludi S. Backlash. O contra-ataque na guerra não declarada contra as mulheres. Rio de Janeiro: Rocco; 2001. emerged in western society in periods of greater conservatism and influence of religious fundamentalism, and is a discourse of attack on women’s equality. This movement, which is considered to be a retrograde anti-feminism, acts through recrudescence of the conservative rhetoric that proposes keeping women in the domestic environment and social control of sexual and reproductive behavior. The objective is to maintain the hierarchies of gender, which are necessary to make the system of subordination and exploration of women, as made more acute in capitalism.

The relationship between femicide and male death by aggression, which showed the highest correlation, was found in the studies made in this survey and indicated that where there are more murders of women there are also more murders of men¹⁴14. Meneghel SN, Hirakata VN.Feminicídios: Homicídios Femininos no Brasil. Rev Saude Publica 2011; 3(45):564-574.^,1515. Leites G, Meneghel SN, Hirakata VN. Homicídios femininos no Rio Grande do Sul, Brasil. Rev Brasileira de Epidemiologia 2014; 17(3):642-653.. The neoliberal economic model in the peripheral countries has sharpened both economic and gender inequalities. This model resulted in increased exploration of women, principally the young, poor and black, exposing them to increasingly severe acts of violence. The new gender division of work placed a large proportion of women in the informal market, on routes of migration, and in sex exploration and illegal work. They have more people under their responsibility, less mobility and need to accept the worst work, the lowest salaries and the most dangerous conditions. The increase in the structural violence of the patriarchal system produces some of the new scenarios of femicides¹¹11. Carcedo A. No olvidamos ni aceptamos: feminicídio em centro America. 2000-2006. San José: CEFEMINA; 2010.^,2323. Segato RL. Território, soberania y crimenes de segundo estado: la escritura em el cuerpo de las mujeres asessinadas em Ciudad Juarez. Brasilia: Universidade Nacional de Brasilia; 2004. Serie Antropológica 362..

This study has limitations in its ecological design, which uses secondary information from information systems that may be partial or biased, although the sample space chosen is the state capitals and municipalities with population of more than 400,000, where the quality of information is more accurate. This is because, unlike the large-scale municipalities, in small towns there is a greater possibility of low capture of information on deaths, through the civil registry, via notaries’ offices. The sub-recording of life events is greater in these areas, among other reasons due to non-recognition of the importance of the document²⁴24. Frias PG, Pereira PMH, Andrade CL, Szwarcwald CL.Sistema de informações sobre mortalidade: estudo de caso em municípios com precariedade dos dados. Cad Saude Publica 2008;24(10):2257-2266..

This investigation, which is part of a larger survey, evidenced the structural problems of understanding murders of women in large urban centers. It found: an inverse relationship between gender violence and poverty – indicating higher rates of death due to gender in richer regions, although the victims are mainly poor and unprivileged women; and higher rates in unequal territories where there is an accentuated presence of evangelical religion, poverty and structural violence.

References

Publication Dates

History