The Bureau of Investigative Journalism has focused on improving military transparency and pushing for accountability around US operations in Afghanistan since 2015. However, in the past year, we feel we have made significant progress, increasing the public’s visibility on the scale of the airwar and trends in it, but also highlighting what this means for the people on the ground.
The Bureau began tracking strikes from January 1 2015 onwards, when the US and Nato’s combat mission in Afghanistan came to an end. At the time we began recording strikes, we relied on open source data, mostly using reports in the local news. We would then apply our methodology, recording strikes into a public database.
Over time, this prompted the US military to engage with us and they began releasing official figures on the number of strikes carried out each month. However, in September 2017, this ended and a year long blackout began. We continued to question this decision, publicising the blackout through a series of articles and engaging with various concerned parties.
And in October 2018, the US began to release monthly strike totals for Afghanistan once again. The new data provided much more detail than previously available, including information on where and what strikes hit. However, two months later, the US military quietly stripped out information on the strikes’ targets. Resolute Support, the US-led Nato mission in Afghanistan, told the Bureau the sudden withdrawal of information was due to operational security concerns.
The decision to withhold the information followed shortly after the Bureau published a story highlighting the number of buildings destroyed in US strikes. Our analysis found that over 60 buildings had been destroyed by these strikes in a single month. The story quoted civilian casualty expert Larry Lewis who said the strikes were the “riskiest kind” for civilians – they have been subject to an almost ban since 2009 due to their link with civilian deaths.
When the US military decided to withhold data on how many buildings they were hitting, we knew we wanted to demonstrate the impact these types of strikes can have on the ground. The data they had briefly released had showed an alarming number were being targeted.
We worked with talented photo-journalist Andrew Quilty who was at a hospital in Helmand when ten children from the same family were rushed in – their house had been hit in a US strike. Our photo story documented the family’s struggle to recover. But it highlighted a wider issue – this data is vital in holding the US military to account.
What makes this project innovative?
Instead of highlighting the issue in a single story, we have shown the value of sustained coverage. As a result of us having followed this issue for the years before, we were in a unique position to not only notice the decline in transparency and trace its origins, but to ask for change. Central to our efforts is the guiding principle that this data should be available to the public. And this can be seen throughout our years of covering the topic. All confirmed strikes are recorded in a publically available spreadsheet, which includes details such as their location and the number of people they killed. Our website lists these entries with further details on the incident and sourcing information, alongside strikes we have not been able to confirm. We also have a tool that allows users to create graphs and charts from the spreadsheet data (this was developed before the award period). Beyond this, we engage frequently with other journalists and organizations to ensure the data is widely used. And we collaborate - we try to make sure every story is published with a partner, including local partners.
What was the impact of your project? How did you measure it?
I believe our military transparency campaign has had tangible impact, particularly in the past year. Thanks to our consistent coverage, we noticed the decline in transparency and were able push for the release of the data, providing the public with important metrics to evaluate the war in Afghanistan. When this data was released again with more detail, we were able to highlight trends that concerned us and translate what those figures meant for people on the ground. It is difficult to provide particularly accurate audience engagement metrics as we often publish with partners and do not have access to their data on this. Our publishing partners for the stories in this entry include outlets such the Daily Beast and Just Security. But importantly, we have also worked with local media in Afghanistan such as Pajhwok and Etilaat Roz, expanding our reach to those affected by what we are reporting on. This includes expanding beyond English language media outlets – our first story in Dari was published in January. We can offer some audience engagement metrics specific to our site. For example, the Bureau’s work on Afghanistan has been cited 375 times in the last year, including mentions by organisations such as Amnesty and CIVIC, but also by news outlets across the globe. Our website page for the Afghanistan data in 2019 has had 1,979 views so far, with 1,860 of those being unique. Readers spent over three minutes on average on our photo story, with 4,182 unique views out of 4,546. We are also pleased that we have a readership in Afghanistan - after the US and UK, most users that read our story on the CIA-backed Afghan special forces unit, for example, came from Afghanistan.
Source and methodology
The data we receive is official data from the US military. Currently the data provided does not include information beyond the date of a strike and its location (province level). We try to provide additional details on these strikes by searching though information published by national and international media outlets, NGO reports, and sometimes conduct our own field investigations. Sometimes we are unable to match a strike reported in the media with the list provided by the US military. When this happens, or when a strike has been reported by unnamed or vague sources, we list it on our website as a C strike, signifying that is a possible strike but not yet confirmed. Even within a single report there can be contradictory information on how many individuals were killed or what the target was – for example the report might say it was ‘either a house or a vehicle’ that was hit. Reconciling accounts from multiple sources can be even more difficult. Where credible sources differ over how many people were killed we provide a minimum and maximum count of the number of people reported killed. This is why our casualty counts are a range.
When the data was released, we needed an efficient way to import it into our spreadsheets. We used a computer script written in python and using python's requests, bs4 and csv modules to gather and structure into a spreadsheet data format the information contained in the military's monthly afghan strikes reports. Not only did this made it easier to analyse and interrogate the reports, but as NATO resolute support removes the previous report each time a new one is published, it ensured that all this information would remain publicly accessible.
Jessica Purkiss Abigail Fielding-Smith Andrew Quilty Emran Feroz Charles Boutaud