ByΒ Rowan Philp
If there is one type of lesson that journalists never forget, it is the kind that follows a strategic reporting mistake that stalls an otherwise promising investigation.
GIJN asked a diverse group of nine investigative reporters to share one memorable misstep theyβve made in an investigative project, and the key lesson they learned.
What is striking about their responses is that most of these errors involved dealing with uncooperative human sources β and most of the takeaways emphasize the value of extra preparation before interacting with them.
What investigative mistake have you made that others can learn from?
Vuyisile Hlatshwayo β investigative reporter at the Inhlase Centre for Investigative Journalism in Southern Africa (Eswatini)
Courtesy: Vuyisile Hlatshwayo
Vuyisile Hlatshwayo, director of the Inhlase Centre for Investigative Journalism in Eswatini. Image: Courtesy of Hlatshwayo
βIf there was one investigative mistake that I wouldnβt like to repeat in my career, it was to give too much information to a public relations officer at a government regulator. I learned this with a local regulatory body in a story about a school uniform pricing cartel. I was asked by that body β a competition commission β to send a questionnaire so that they could provide written responses. But, lo and behold, two days down the line, the commission issued a notice in the national newspapers notifying the public that it was investigating price fixing of school uniforms! It went on to invite concerned members of the public to make submissions. I was sidelined in the whole investigation and didnβt even get the report at the end of the so-called investigation. Thatβs how I learned β the hard way β that you donβt have to give too much information to your accountability sources because they have a tendency to kill stories.β
Axel Gordh HumlesjΓΆ β investigative reporter at public broadcaster SVT (Sweden)
SVT investigative reporter Axel Gordh HumlesjΓΆ. Image: Screenshot
βWe did a lot of undercover stuff with our latest story, and it was a success and a failure. The investigation was a bribery story about the worldβs biggest security company, which has about 400,000 employees. We had a lead in London about a paid-for trip, so we wanted the hotel bill, and the hotel wouldnβt give it to us. So I created a Gmail account in the name of the public official, with some numbers after it, and called the hotel and said: βIβm doing my bills; can you send me my bill for my latest stay there?β I gave them the email address, and they said theyβd send it right away. But I never received the email β it went to the real guyβs email! I called again, and they said theyβd sent it to the email registered on their database. I tried saying, βOh no, thatβs an old email,β and then I got the bill to the new address, which was good β but unfortunately the official was now alerted that we were on to him.
βThe lesson learned is that creativity is good, but you have to do role-playing with colleagues to see what could be the outcome. I could have told them the email they had was old in the first call. When you think two steps ahead, you have to think of the third and the fourth steps that can happen.β
Martha Mendoza β Pulitzer-winning investigative reporter at the Associated Press (United States)
Associated Press investigative reporter Martha Mendoza. Image: Courtesy of Mendoza
βI always tell people what Iβm investigating β Iβm not good at being sneaky about it. On Seafood from Slaves, we tracked a bunch of seafood to the US. I went to a great seafood expo, and I interviewed the director of the expo on day one, first thing in the morning. I thought: I want to ensure I get this guy on camera, as he was in charge of seafood imports into the US at some level. But afterward, he put out a blog post to all the companies there, along with my photo β and said, βMartha Mendoza from the AP is here, writing about labor abuse; you donβt need to talk to her.β Then, every time I would approach the owner of a seafood company, they would just be, like, βNo, weβre not going to talk to you.β Before I could even open the door, and say βI have a dossier here with your companyβs export records, and we tracked the seafood; we have X photo.β I couldnβt even open the conversation. I should have waited until the end of the meeting to talk to him.β
Mia Malan β Editor-in-Chief at the Bhekisisa Centre for Health Journalism (South Africa)
Bhekisisa Centre for Health Journalism editor-in-chief Mia Malan. Image: Courtesy of Malan
βIβve learned to never ever trust a second-hand source for statistics. That may sound simple, but many journalists donβt consider, for instance, a health minister quoting international statistics β so, not from his own or her country β or, say, a mental health organization issuing a press release on the proportion of people who have depression, a second-hand source. Iβve been amazed at how often those sources are wrong. Unless youβve checked the original source theyβre quoting, do not believe the statistic and donβt quote it. The same goes for a journal study β no matter how prestigious the journal β quoting another studyβs results. Go and check the original study and see if itβs been interpreted correctly. You will be amazed to see how often study authors βbendβ findings to suit their arguments.β
David McSwane β investigative reporter at ProPublica (United States)
ProPublica investigative reporter David McSwane. Image: Screenshot
βIβve made the mistake of talking to somebody, and theyβll say, in passing: βWas that off the record?β Iβll say, βWell, weβd have to agree to be off the recordβ β and then thinking that was clear. And later theyβll say: βI was under the impression we were off the record.β Iβve learned we canβt assume source proficiency with journalist rules, and itβs better to just lay it out very explicitly: βThis conversation is on record. If you want to talk off record, we can do that another time, or we can talk off record now, and be on record later.β You donβt want anyone to feel tricked.
McSwane adds a second takeaway: βYou really do need to pick the long-term stories about which youβre passionate, because itβs going to be a slog β if itβs a topic you donβt care about that much, you wonβt do a great job.β
Rosa Furneaux β investigative health reporter at The Bureau of Investigative Journalism (United Kingdom)
TBIJ investigative health reporter Rosa Furneaux. Image: Screenshot
βI have erred on the side of being more cordial than I ought in the accountability interview with someone I suspected was a bad actor, and done that interview too soon. Iβve had situations where I know I have a tricky source whoβll sit down with me, and Iβve had to decide whether to immediately whack them with hard questions, or try to butter them up to get a second interview. Iβve misjudged that, where Iβve gone to a source a bit too early β where, lacking evidence, Iβve been afraid to whack them with hard questions, knowing they wonβt come back to me for a second interview. They say: βWell, Iβve already talked to you once.β Whereas if Iβd waited a few more weeks, Iβd have known the exact tough questions to put to them, and, when I got that one opportunity, I could get the answers I needed and not need to worry about a second interview.β
Juliana dal Piva β columnist for UOL NotΓcias, former investigative reporter at O Globo (Brazil)
UOL NotΓcias columnist Juliana dal Piva. Image: Screenshot
βOne mistake that I learned a lot from was when I was checking data on political disappearances at the time of Brazilβs military dictatorship. A few years ago, this data was very disorganized. One day, I was asked by my newspaper to produce a report β in 24 hours β on the cases of political disappearances from the dictatorship specific to Rio de Janeiro. This meant compiling about 30 cases out of a total of 500 victims β among the total number of disappeared and people who were executed β but whose families were able to bury the bodies. Unfortunately, when looking at the data spread across different surveys, I forgot a name.
βA few months later, when investigating documents from the dictatorship that had been released for consultation for the first time, I ended up finding the first document from the dictatorship in almost 40 years that acknowledged the death of the murdered guerrilla. Iβm sure I only noticed his name in the huge list of names because I had made that mistake, and I never forgot the name and history of MΓ‘rio Alves. Mistakes teach you.β
Alexandre Brutelle β freelance investigative reporter and director of the nonprofit Environmental Investigative Forum (France)
Environmental Investigative Forum director Alexandre Brutelle. Image: Screenshot
βDuring our investigation into unlawful fracking operations in southern Tunisia by a French-Tunisian company, our task was to βhuntβ for water ponds located in the middle of the Tunisian part of the Sahara Desert β quite a big area of interest. We were already doing a poor job of looking online for the official limits of the companyβs oil and gas concession in order to restrict our searches, but could not find anything. So, for weeks, we reviewed a suspected area of 40,000 square kilometers that could be located anywhere in the south of the country, mapping everything we could. At the end of the project, we realized the ETAP (Tunisiaβs national oil agency) website already had a map showcasing these concessions. Instead of looking for βCompany X oil concessionβ on Google, we learned, instead, to always look for the data source first. So try to Google the category β βoil concession registry,β or similar other queries β directly at the start of an oil industry investigation!β
Dan McCrum β investigative reporter at the Financial Times (United Kingdom)
In a recent GIJN webinar, McCrum emphasized two related mistakes he made while investigating massive fraud at Germanyβs giant Wirecard electronic payments company.
Financial Times investigative reporter Dan McCrum. Image: Courtesy of McCrum
βIn early 2016, a new set of short sellers came along who decided to publish their allegations anonymously β what is known as a short-seller attack, hoping to drive the share price down. They shared a copy of their dossier with me in advance, which I thought might force German regulators to investigate. But I got a bit carried away. I was determined to βownβ the story, and be the first to write about it β and I felt the dossier was out in the public domain. I made a mistake, where I agreed that I wouldnβt ask anyone for comment until it was published, because they didnβt want to tip off the company. So I couldnβt test it properly. So I just wrote a very short blog post, which just drew attention to the fact that this report had been published β and with a link to the report.
βWell, the phrase we use at the FT is, if you ask the question: βShould we run this by a lawyer?β then you already know the answer. Because the blog post was just a couple of sentences, I thought it would be safe, and that was quite a big mistake β because Wirecardβs lawyers started sending threatening letters, saying the Financial Times was responsible for publishing this entire report, and that therefore we were on the hook for everything in it. That meant Wirecard could file a libel suit at any point in the next year. What it meant was I couldnβt write about anything in this 100-page report; it would have been daring Wirecard to sue us. I think that chilled other people investigating it. I gave up for a bit. But I always had the backing of my editor on the team. Luckily, stories get stories, and eventually whistleblowers came forward.β
Additional Resources
Investigative Tactics That Reporters Love
Megha Rajagopalan: What Iβve Learned About Investigative Journalism
Rowan Philp is a reporter for GIJN. He was formerly chief reporter for South Africaβs Sunday Times. As a foreign correspondent, he has reported on news, politics, corruption, and conflict from more than two dozen countries around the world.