Session notes from the Investigative Practice Series
Chair: Chris Woods
Founding Director, Airwars
Freelance reporter and founder of the Yemen Data Project
Journalist and author of No Good Men Among the Living: America, the Taliban and the War Through Afghan Eyes.
War reporting is changing: reams of data and munitions expertise can now be brought to bear without ever setting foot in a war zone, while the access of officially ‘embedded’ reporters is heavily subscribed, leaving it to freelancers to take the huge risks necessary to get the story out. Analysis of data and social media opens up unprecedented new opportunities for warzone reporting, but does it also open journalists up to manipulation?
In this seminar, Investigative Practice asked how online research and on-the-ground reporting can most fruitfully work together.
"The role of the field correspondent is crucial in not just telling stories but also framing our understanding of civilian harm."
The need for on-the-ground reporting
On-the-ground reporting is in decline. This is a fact made evident by an Airwars report, examining the media’s coverage of civilian casualties in Iraq. What they found was that during the first two years of the war against ISIS in Iraq, not a single report from a mainstream media outlet was carried out from the ground. The reason why this is so problematic is because the Pentagon uses mainstream media field coverage as a metric in determining the extent of civilian harm.
This gulf then has to be filled by people from those countries. For instance, Iraqi and Syrian civilians self-reported, predominantly through social media, more than 26,000 casualties from Coalition actions since 2014. A similar number of civilian fatalities reported in the same way from Russian actions in Syria alone. Why aren’t we listening to those voices?
So in this world where our governments fight remote wars, given the general absence of reporters on the ground but with so much information online how do we ensure that the effects of those wars are properly understood? And what are the risks of these new methods?
"It’s about having a specialism."
Iona Craig has been reporting from on the ground in Yemen for several years, and also working with the Yemen Data Project so her work spans both the methodologies in question.
Reporting using data
In using data to form stories, anonymity is often crucial. This is something Craig found key when working with the Yemen Data Project. A lot of their data was open-source, taken from Whatsapp groups, Youtube and Twitter but also from local media and local activists. A difficulty they found in establishing the project was gaining the trust of international media and journalists. Some were concerned that the data was being used to push a political agenda: ‘a front for the Saudis or the Houthis.’ Their way around this was to be as transparent as possible about their methodology and its possible flaws.
More generally, the demand to report on her findings as a journalist was sensitive to changes in global politics. For instance, Craig pitched a story about disappearances in Yemen to five different publications around the time of Trump’s election, and had no uptake. Another problem in spreading these stories is that there seems to be an audience bias towards foreign correspondents, having a foreign correspondent going in is somehow seen as more legitimate than a local voice telling that story. For example, when writing a piece about the navy SEAL raid on Yakla in 2017, Craig found that her article had a lot of traction in spite of its slow release and the fact the Bureau of Investigative Journalism had already covered the story with a Yemeni reporter. This seems to be a cultural issue within the media industry and in the media consumption of the wider Western public.
Reporting from the ground
The obvious advantage of reporting from the field is that you gain first-hand experience of the territory you are reporting on, importantly of local dynamics, politics and internal biases. Craig found that building relationships with locals and having local knowledge gave her special access to stories she wouldn’t have had otherwise. When the Saudi Coalition air-campaign began, she was able to get into Yemen by boat when many other journalists were refused. Having spent a lot of time there, she was given permission from tribal leaders to go to certain places because they had a clearer understanding of her intent and trusted her.
The other advantage her time spent and connections in Yemen afforded her was the knowledge of local power structures and different political factions. This allowed her to unpick the different biases and contradictory claims that are evident in reports coming out of the country in a way that others without that local experience and knowledge would not be able to. Having put in this time and work to understand the place and the people living there provides advantages in reporting both on the ground and remotely using data and open-source evidence.
"In pursuing ground-based stories, there are a variety of obstacles to navigate."
- Funding: It is difficult to get stories funded that involve going into the field. Anand Gopal’s project, The Uncounted, on US-led coalition actions in Iraq was rejected by multiple outlets before it was taken up, for instance. His project visited 103 sites over the course of three years, demanding a lot of resources.
- Lack of infrastructure: Gopal found that in Iraq, unlike in Syria, there was no active local media network. This meant that events were not being reported online, making ground investigation even more crucial.
- Complexity of fighting: On the ground, it is difficult to decipher responsibility for civilian harm especially in conflicts with many different actors such as the war against ISIS in Iraq. Gopal got round this by visiting every single impact site to determine what munitions and artillery type were used for each strike and who was responsible.
- Incomplete records: Gopal’s method was to send the coordinates of the strike sites to the Coalition to ask solely if they bombed that location. The majority of those strikes did not involve civilian harm; they were either empty structures or ISIS targets. So by asking simply whether a given location was targeted, and not addressing the reasons for targeting or the nature of what they were trying to hit, they were able to access a different response mechanism from the Coalition bureaucracy which then gave them yes or no answers for each of the sites that they had studied. Unfortunately, the records those answers were based on turned out to be inaccurate or incomplete, as demonstrated by the Coalition claims that they had not struck a location for which they themselves had uploaded video evidence of striking.
- Bureaucracy: Getting answers, even inaccurate ones, as to why a region was targeted, is an even more difficult process. A common route is to make an FOI request but the issue here is that they can take an extremely long time to process. The way they circumvented this is by stressing the danger that some of the victims were in. For example, in the case of Basim Razzo, one of the survivors of an airstrike reported on in The Uncounted, the Coalition records and the footage of the airstrike they published stated that Razzo’s home was an ISIS IED (Improvised Explosive Device) factory, and the deaths of his wife and children were allocated to a column recording the deaths of ISIS fighters.This left Razzo at huge risk of being further targeted by the Iraqi authorities for his alleged ISIS sympathies.
This approach did work in terms of getting answers from the Coalition bureaucracy much more quickly, within two or three months, but a further issue is that these loop-holes close up when they are used too often, which meant the government stopped fast-tracking Gopal’s FOI requests, despite there being hundreds of other civilians in similarly dangerous situations. He then resorted to suing the US government to speed up the process.
Q Is there a bias towards stories from foreign correspondents over those from local journalists, and does that apply to all international journalists or just those based in Western countries?
A What seems to make the difference is if the story is covered by an American mainstream media outlet, specifically. For instance, if you look at Colonel Ryan, the spokesperson for the Coalition, all the outlets and journalists he follows on Twitter are American. In this sense, it is not so much about getting a journalist from a western country, as it is about getting the backing of a US outlet.
It is important that when you are pitching a story to a US outlet, you get some interest even if not a direct commission prior to your trip because they will want an editorial input throughout the investigation: which characters to focus on or what angle you will take, for example. It is also good to have some kind of institutional support whilst you are on the ground, even when you are freelancing, though this may of course be easier said than done. There is also the issue that you cannot ever be certain, especially in conflict reporting, whether there will be a story there when you arrive or what form such a story will take, which makes this process of getting institutional support prior to the trip that much more difficult. This is less of an issue with TV, since there are greater resources available and you can often get ‘project development’ to cover such exploratory work, but when working in print that funding is almost never available.
Q How do we effectively report on stories that are so fast moving?
A In the case of the US, they are dropping bombs faster than they can build them, allegedly one every twelve minutes. When the numbers are so staggering it can be difficult to keep monitoring each airstrike. In the case of Raqqa, which the UN has described as the most destroyed city in Syria, 20,000 munitions were dropped by the Coalition forces alone on the city in less than five months. There was very little coverage of this, though Quentin Sommerville of the BBC and Amnesty have been doing good work trying to understand the events after the fact.
But what can be done, as Airwars, Iraq Body Count and others have shown, is to monitor civilian harm from the perspective of civilians, recording claims of civilian casualty events, run rudimentary assessments and then leverage this information for change. The challenge is in getting governments to engage with that information, in order to introduce at least some accountability into the process.
The US military have actually proved reasonably responsive to such pressure and have improved their methods for reporting civilian harm, whereas in the UK the government does not want to concede any civilian deaths at all. The UK is currently involved in wars in Iraq, Syria, Afghanistan and Libya and have admitted two civilian casualty events in the past decade, one in 2008 and one in 2018. These are systemic and cultural challenges which we have to contend with.
To some degree, in counting bombs dropped, you will have to round-up or alter the metric by which you classify an attack. For example, in the case of the Yemen Data Project, they will characterise several strikes on the same target as a single air-raid if it happened over a short time. In terms of air raids, after three and a half years, the records are nearing 19,000 air raids, but in terms of the individual bombs being dropped, it’s just not feasible to count.
Q How can you verify your sources if you are not on the ground?
A Part of it is intuitive and self-evident, such as in the case of Airwars investigating civilian harm in Syria. They found that the nature of underrepresented stories is that they lack any kind of wider audience in the international community, which means that the very locally specific communities that are reporting most of these incidents have little incentive or agenda to fabricate posts. Airwars also found that these communities were good at self-regulating. They saw this in how local communities fact-checked both inflammatory ISIS claims and the reports of local outlets, through the comments section.
In certain areas such as Yemen it is very difficult to have any system of validation because there is no online community surrounding them to report what is going on. In this situation, you have to rely on word of mouth and pool from your existing contacts. Alternatively, you may have to just visit the location yourself. Craig reiterated that this is a scenario where it is really important to have a speciality because you will avoid the obvious pitfalls of disinformation. This proved essential in questioning the Houthis’ inflated reports of civilian casualties or deciphering media biases when you have media organisations with the same name, but directly opposing biases.
Q How do we validate and verify reports from smaller humanitarian agencies and charities when people are reluctant to trust them?
A You may be able to assess reports from smaller organisations through speaking to larger sponsors or UN agencies that are working with those organisations on the ground and will have a more informed understanding of the agenda or trustworthiness of each agency.
The other method is to validate content coming from small agencies by cross-referencing against other reports online and using standard verification methods such as geolocation. Ultimately, understanding the credibility of those agencies may come down to a mix of intuition, local knowledge and the experiences of other journalists working in the area.
Bringing the methodologies together
Ultimately what is probably required in complex conflict reporting is a hybrid of both on-the-ground reporting and remote data analysis. In bolstering the credibility of the story, it seems that drawing on both data and on-the-ground reports is the best way forward. It is possible that someone on the ground may exaggerate their experience but it is also possible that official statistics can be inaccurate. As such, it is best to use them in tandem to get a clearer picture of events.
The question really is where the ideal balance, or the merge point exists between these two approaches. In marrying the two, it may call for journalists to elevate and refine new skills, as well as an increase in cross-disciplinary collaborations, especially between the methods and practices of both journalism and social science.
This discussion was part of the Investigative Practice series at the Logan Symposium, Oct 2018