Repost: Reporting and context – an example from 2021

Originally published on September 21, 2021. Republished with minor grammatical corrections. Although the article at the heart of the piece is now rather dated, it nevertheless serves as a notable example of how the press routinely fails to contextualise statistical data for the sake of eliciting an emotional response from the reader.

On September 20, 2021, the Guardian published on its website the following headline: “Ten women and girls killed every day in Mexico, Amnesty report says”1. The headline was sufficiently eye-catching for me to click through to the article, and at least at first, the story seemed just what one would expect. Women murdered, police dragging its feet, families forced to carry out their own investigations, Amnesty International publishing a “scathing report”, and so on. This is all actually quite terrible stuff, certainly from the point of view of Mexican women, though one is unclear as to how impactful this single report by Amnesty is ultimately going to prove.

And yet.

Going through the piece, one could not help but notice that every single figure and statistic cited concerned female murder victims – 3,723 women killed in 2020, 940 of these cases investigated as femicides, roughly 400 women murdered in a single Mexican city during the 1990s. Nowhere were these figures placed into any conceivable context. Is the figure of 3,723 victims a lot, or a little? What was the total number of murders recorded in Mexico during 2020? What is Mexico’s murder rate per capita, whether for women only or in general, and how does it compare with, say, its northerly neighbour, the US? Should we, the readers, not have any frame of reference through which to view these numbers? Is that not worth even half a sentence in a story that runs for 19 paragraphs?

Perhaps the Amnesty International report upon which the article is based can provide some answers, I said to myself, with the Guardian‘s editors and writers simply deeming them as unnecessary for the emotional thrust of their article. I thus turned to Amnesty’s press release on the subject2, only to find that just like the piece in the Guardian, it provides absolutely no context for the 3,723 figure. Nor does the 60-page report itself fare any better. To be sure, on pages 16-17 it does provide a table that includes per-capita data on female murders for each of Mexico’s constituent states, though, curiously, not for the country as a whole. Aside from this, however, no significant added statistical information appears to be included; rather, much of the report is devoted to an exhaustive description of four individual murder cases, and a slew of recommendations for the Mexican government, virtually none of which are likely to be heeded in practice. As well, there are not less than three pages devoted to defining the term “feminicide” and explaining why it is a human rights violation, though again, without bothering to give the readers any sort of feminicide statistics outside of Mexico. Surely in a 60-page report one could find even a little room for this? Surely we can at least try and compare Mexico’s record on violence against women against that of even a single other country, if we cannot be bothered to contextualise the figure of 3,723 itself? Why, why would Amnesty International, whom the Guardian basically parrots in this instance – because not asking basic questions is what real independent and facts-based journalism is all about! – take this particular tack?

To provide one possible answer, we turn to the data. Alas, I am not as yet proficient in Spanish, which precludes me from digging through the Mexican government’s own reports on crime rates. Nevertheless, a quick Internet search reveals several relevant data points:

  • A Statista page3 with figures for homicides in Mexico over 2015-2020. Specifically in 2020 the country recorded 43,265 killings of all kinds, including both murders and manslaughters.
  • A Council of Foreign Relations “Global Conflict Tracker” page for Mexico4, stating that during 2018 there were 33,341 drug-related homicides in Mexico. It must be noted that both Amnesty and the Guardian openly scoffed at Mexican authorities chalking up murders of women to “organised crime”.
  • A dual-language website seemingly specifically devoted to crime in Mexico5, filled with all sorts of informative charts and tables that, among other things, show a total of 35,532 homicides in 2020 based on the number of opened police investigations. This figure includes 3,792 female homicides and 975 feminicides, as opposed to Amnesty’s figures of 3,723 and 940, respectively, which suggests that different primary sources are utilised6.

Clearly, depending on the source, women and girls account for somewhere on the order of 10% of the total number of homicide victims in Mexico. One supposes that someone at Amnesty must have thought providing this kind of contextualisation might well detract from the report’s emotional impact. As well, given that Mexico’s per-capita murder rate appears to be five to seven times greater than in the neighbouring US7, one might well suggest that four thousand female murder victims per annum in Mexico, while tragic and regrettable, is hardly an unexpected development. Finally, it is rather more difficult to dismiss the excuse employed by some Mexican authorities to avoid investigating female homicides, that it’s all due to “organised crime”, when, quite apparently, each year tens of thousands of murders in Mexico are connected to organised crime.

On the other hand, surely not a single one of Amnesty International’s main points – that a number of female homicides are not properly investigated, that feminicide is a problem in Mexico, et cetera – would be rendered invalid by any of these figures. More nuanced, perhaps, requiring some further elaboration as to why, in a country with a less than encouraging criminal environment8, this particular kind of crime ought to be singled out. As things stand, however, Amnesty basically appears to airbrush all of this context out of the picture, which smacks of “ends justify the means” sort of ethics that the organisation should ostensibly stand in opposition to.

And there is more, of course. For if Amnesty had bothered to contextualise rather than propagandise, for example by comparing female homicides in Mexico with its neighbour, the US, it might have asked the following question: how is it that female homicides in Mexico amount to roughly 10% of the annual total, when according to the FBI’s data for 20199, women accounted for 21.2% of all murder victims in the US that year? Is the feminicide situation in the US twice as bad as it is in Mexico, notwithstanding its much lower overall crime rates? Or, more likely, is the magnitude of drug-related violence in Mexico such that far more men are killed every year than would be expected in a relatively peaceful society? And does this ongoing, seemingly never-ending drug war somehow skew the government’s responses and investigative priorities, which in turn ends up disproportionately affecting female victims of crime?

Alas, such questions were not to be. In any event, Amnesty is an advocacy-focussed NGO, and so can be expected to at least occasionally present its findings and recommendations in a way which some may interpret as…slanted. The real issue in all of this is the conduct of the Guardian. To partly quote the paper’s solicitation for donations appended to every single article posted on its website:

Tens of millions have placed their trust in the Guardian’s high-impact journalism since we started publishing…With no shareholders or billionaire owner, we can set our own agenda and provide trustworthy journalism that’s free from commercial and political influence, offering a counterweight to the spread of misinformation. When it’s never mattered more, we can investigate and challenge without fear or favour.

One wonders precisely how a 19-paragraph piece that parrot’s an NGO’s press release almost verbatim and fails to ask even basic contextual questions fits into this model of “investigating and challenging without fear or favour”.


  1. Angren, D., “Ten women and girls killed every day in Mexico, Amnesty report says”, Guardian, September 20, 2021, retrieved September 20, 2021.[]
  2. The release, which includes links to the report itself, can be found at https://www.amnestyusa.org/press-releases/failings-in-investigations-of-feminicides-in-mexico-violate-womens-rights/, retrieved September 21, 2021[]
  3. Namely https://www.statista.com/statistics/959787/mexico-number-homicides/, retrieved September 21, 2021[]
  4. See https://www.cfr.org/global-conflict-tracker/conflict/criminal-violence-mexico, retrieved September 21, 2021[]
  5. See https://elcri.men/en/, retrieved September 21, 2021[]
  6. One is curious how Amnesty accounts for the fact that a simple Internet search on “homicides in Mexico” reveals more female murders than they claim in their “scathing report” on the subject.[]
  7. Specifically, 24.6-28.4 in Mexico over 2020, depending on the month, per https://elcri.men/en/, compared with US murder rate of 5.0 for 2019 calculated by the FBI in https://ucr.fbi.gov/crime-in-the-u.s./2019/crime-in-the-u.s.-2019/topic-pages/tables/table-1, both sites retrieved September 21, 2021[]
  8. Well – the criminals certainly might feel encouraged![]
  9. See https://crime-data-explorer.app.cloud.gov/pages/explorer/crime/shr, retrieved September 21, 2021.[]