We have to celebrate the call by the UN Special Rapporteur on violence against women, Dubravka Šimonović, for the creation of national observatories for femicide/feminicide, as well as the UNODC´s recent proposal for these observatories. Both the Convention of Belem do Pará and the Istanbul Convention establish the responsibility of States to collect accurate information about violence against women, and therefore, implicitly, about femicide.
When discussing the murder of women based on gender subordination, the gravity of the problem not only derives from its incidence – femicides are committed all over the world and in some countries, the rates are high enough to be considered an epidemic; 10 for every 10,000 women. Once we set the numbers aside (one is already too many), we see that the severity of the issue is due to its very nature: an extreme violation of human rights that threatens half of the population.
What is surprising, however, is how long it has taken for this UN call to be made. For over two decades, the international feminist movement, particularly in Latin America, has been protesting and reporting crimes of femicide, building theoretical frameworks to explain it, developing the methodology to investigate it, requesting transparency of information and calling for adequate responses to stop this slaughter of women, which shows no signs of abating. We need to recognize the vision of Diana Russell, who identified, named and conceptualized this issue in 1976, when there were still no mass movements to stop femicides.
As regards to femicide specifically, feminist organizations created the benchmarks for what to look at, and how to look for it. For example, there are various femicide observatories that operate both globally and regionally and there have been many investigations published since the start of the century. Feminist organizations have been keeping track of incidences of femicide for years in countries such as Costa Rica, Argentina, Honduras and Spain. Additionally, the Central American Feminist Network against Violence towards Women has a Central American Observatory for the Eradication of Femicide. Progress has been made and lessons learned. The call by the Rapporteur and the UNODC proposal should recognize these efforts and not turn a deaf ear to collective learning experiences.
The identification of femicides is no easy task, as not all the murders of women are femicides. It is necessary to investigate the circumstances and dynamics of the crimes to recognize the unequal relationships of power between men and women that undergird a femicide. The availability of such information requires specialized judicial processes that are applicable to all murders of women, something that cannot always be guaranteed. Even with the political will and resources to carry out such a task, the exact circumstances of particular crimes can be hard to ascertain.
Identifying and compiling statistics on femicides is clearly not enough. We need to learn more about them to be able to explain why there are increases or decreases in numbers, and more importantly, there is simply no way of assessing the impact, if any, of public policies designed to eradicate the phenomenon. When reduced to numbers, femicide becomes a mysterious problem, both impenetrable and unavoidable; its history and trajectory unfathomable.
Needless to say, the final goal of investigating femicides is to tackle them with increasingly efficient processes until complete eradication is achieved. In that sense, numbers must be a tool and not the sole purpose of our efforts. The risks of forgetting this have become apparent during the more than two decades of our endeavour.
The first risk is to stop identifying femicides due to the difficulty of discerning which killings of women are femicides, and instead use statistics on homicides of women as a proxy . The labelling of all homicides of women as femicides detracts from the sexism implicit in femicide. Making femicides visible is a political act that decries the fact that women are killed for being women in societies that discriminate against us. In addition, the characteristics and dynamics of femicides are different to non-femicide murders of women.
Another risk is prioritizing comparability between countries in terms of how the problem is understood in each one of them. In this respect, the operative definition we use of femicide is fundamental. When faced with the difficulty of identifying femicides it has seemed practical to restrict cases to those committed by partners and ex-partners, as they are the easiest to identify. A limited definition of this kind allows for comparability between countries but in some cases it simply does not apply. In Mesoamerica, for example, most femicides are committed in different scenarios, mainly linked to sexual violence and criminal gangs.
The third element to consider is that, although in many countries femicide or feminicide has already been classed as a criminal offence, no law deals adequately with the problem – all typologies of crime are restrictive in some way. While observatories continue to use information sources that stick to judicial definitions there will be under-recording and biases. In some countries the issue has been resolved by expanding the definition of femicide and creating mixed spaces for State and women´s organizations to identify and analyze the cases.
The difficulties and risks represent a challenge that must be confronted. The call by the UN offers an excellent opportunity to advance in each country towards a greater understanding of femicides, a necessary prerequisite for their subsequent eradication. It will encourage greater transparency from official sources, and offers the chance to establish or consolidate alliances that increase the resources and capabilities of States whilst providing the vision and experience of the feminist movement. This is an initiative to cherish; of that there is no doubt.
 When there are difficulties in measuring a particular variable, another with a close relation to the original is used – the latter is considered the proxy variable of the original.