Visual investigative journalism about China — Q&A with Muyi Xiao
This is a conversation with Muyi Xiao. In a different era for media in China, she wrote features and took photographs for Tencent News. Now she’s a China reporter on the Visual Investigations team at The New York Times.
![](https://thechinaproject.com/wp-content/uploads/2023/04/image1-2.jpg)
Muyi Xiao works on some of the most interesting stories coming out of the New York Times as part of the Gray Lady’s Visual Investigations team. She began her career about a decade ago as a photo editor for Reuters, and visual journalist for the internet company Tencent.
Times have changed, and much of the work she was able to do back then is no longer possible. She now has an equally intriguing job doing investigative pieces on China that involve all kinds of forensic work. She’s at the cutting edge of China journalism, and we had a fascinating chat about it last week.
This is an abridged, edited transcript of our conversation.
—Jeremy Goldkorn
What exactly do you do as an investigative visual journalist for the New York Times?
The unit I’m working at is called Visual Investigations. The type of investigative journalism that our team does, its uniqueness is that it combines traditional journalism like interviewing people, and developing sources with advanced digital work that includes video and visual forensics, like taking a close examination of satellite imagery, CCTV footage, and social media photos.
We might check it frame by frame, we might take a video and then match it with satellite imagery, those kinds of things. We also deal with nonvisual evidence, like audio, like police radio scanners and intercepted phone calls.
Right.
We also use more traditional things like documents, we do that a lot too.
I would say the type of investigative reporting our team does is very rooted in material evidence, versus the more traditional kind of investigative journalism which often heavily relies on human sources.
Does your team operate independently? Or do you work together with the more traditional type of journalists? Do you combine the two approaches?
We combine. We do work very closely with the more traditional journalists, and we ourselves also do very traditional journalism.
I actually came from a traditional journalism background. So we often work with beat reporters, or as part of a traditional kind of team that is very collaborative. It really depends on the specific story, for example, the balloon story
Ah yes. The balloon story.
For that one, the backbone of the story is the satellite imagery of the balloon and satellite data that we used to calculate the altitude of the balloon.
But that alone wouldn’t be a story. We had to also interview balloon experts, find the right weather experts, and find the right geospatial analysis expert to give us background and analysis. In addition to that, we also needed to work with a national security beat reporter who has been covering the balloon since the beginning of the saga. His sources are government sources who initially released a lot of information about the balloon.
We didn’t approach it assuming those intelligence sources were right. We did our independent reporting using material evidence. We brought these findings, material evidence, back to the reporter and worked with him so that he could conduct more effective interviews with his sources.
Let’s talk a little more about the balloon story. I mean, it was very beautifully presented and I think it was the first time that I had seen documentation of it actually taking off from Hainan?
The U.S. government officials were the first people to tell the world that the balloon took off from Hainan, but they didn’t say anything else. I might be wrong on who reported first, but from my memory I think the Washington Post and New York Times were the first outlets that published that. So [our report] was the first time that we had independent corroboration that the balloon actually did launch from Hainan.
What’s your takeaway from that story?
When I cover this kind of story, clarity is something that is really important for me. This whole thing about the balloon, people have reasons to think that it should be a big thing and also have reasons to think that this shouldn’t be a big thing. But it doesn’t matter what they think. It ended up becoming a spotlight of public attention. I think for events like this, I crave for a lot more clarity. There was all this information floating around about the balloon, I wanted to see things that were verifiable.
So the service to the reader is to really just establish what actually happened?
Yeah, exactly, because there was so much speculation about where it is from, what it is for, what’s the capability, how it’s maneuvered, all of that.
There was speculation and there was information released by both governments. I think providing an independent, verifiable source to the public is very important.
Can you talk about what seemed to be a disinformation campaign during the White Paper protests that preceded the end of China’s COVID-zero policy. When you searched on Twitter for Chinese city names in Chinese, looking for information about protests, what turned up was mostly spam for sexual services.
Your story was titled “How bots pushing adult content drowned out Chinese protest tweets.” What did you find in your investigation?
The takeaway from that story is that the Chinese government was probably not behind it!
Which is something that I do as well: When the Chinese government is not responsible for something, but a lot of people are speculating they are, it is also our work to make sure that it’s not a misanthrope.
Some initial data analysis suggested it might be a coordinated campaign. The timing, the data seemed to show a spike of adult content bot activity overlapping with the start of the protests.
But that was a very limited data analysis, and it intrigued me. So [to understand what was going on], we needed data from way before the protests to compare, we needed a control group, data about cities that were not relevant to the protests, data during the protests and data from before the protests.
So we started to do this. And then the results actually showed that there were no significant differences between cities with protests and those without protests. All major cities are…
When you search those cities’ names, you find a lot of adult content and gambling content, all kinds of commercial bot content.
The problem we had is that we didn’t have much historical data. But we learned that [researchers at] Stanford University were doing a similar analysis, and they had historical data because their Twitter API is better. So we went to them and they provided the missing pieces of the puzzle, which showed is that this activity was pretty rampant even before the protest.
So we could basically prove that this whole phenomena had nothing to do with the protests. So it’s a bad thing that is happening on Twitter. But it is highly unlikely that it’s the government.
We actually interviewed one of the people who posts these advertisements, a sex service provider. And it’s a straightforward sex service.
So they’re using Twitter just because it’s not censored and people looking for illegal services often will use a VPN because?
Yeah, that’s my understanding.
That is very interesting. And I just did a Twitter search on Beijing and Wuhan and…
It’s still going on…
You wrote a story about workers at COVID test kit factory in China not being paid, and how COVID exacerbated the problem of factories not having the money to pay. Talk to me about that.
Yeah. That story is another example of me following evidence.
I came across some videos of worker unrest at a factory. I started to [investigate] where the videos were filmed, whether it is what people say it is, when it was filmed, that type of basic video verification stuff.
As I looked into it more, I searched to see if there were any other COVID test kit factories having similar problems. And I came across a second one, and a third one, and a fourth one.
I [asked myself]: Is there some kind of trend or new phenomenon going on that is different from the usual lunar new year worker unrest [that happens nearly every year around the country when workers have not been paid their wages before the holiday]?
I went to social media, and saw that about a few months before the worker unrest there were a lot of hiring videos posted by labor agencies. They posted a lot of videos trying to hire large quantities of workers for some of these companies, and promised a very good pay for making those COVID testing kits. Shortly before the unrest, there were other videos warning people about those hiring posts that they could be exaggerating compensation and job security.
So I realized, okay, maybe that is why: It’s not only that they didn’t get paid, it’s also that they were promised that the work would be long lasting, stable, and very well paid. But in the end, they were just let go. Here, they got hired with big promises when the COVID testing kit demand was surging. But that demand didn’t last.
But what I saw from social media and by analyzing the videos and social media posts, it was not enough. One of our China reporters, David Pierson, was also looking into it. And we also worked with Keith Bradsher, who actually went to, who checked out those factories, who actually interviewed people onsite. So that was very important for this story as well.
How do you approach the daily work of looking for stories? What’s a typical day in the life of Muyi Xiao?
My typical day actually changes a lot. There were almost two years I was “stuck” in this long project analyzing more than 100,000 government bidding documents about China’s surveillance state.
But right now, we are very much focused on news. Every day, I just get up and take a scan of the news. I check a bunch of sites that in the past have helped me with reporting, databases with procurement documents, government sites, satellite imagery provider sites…
When news breaks, the first thing is to see if there is a visual investigation angle for us. Is there visual evidence or any type of other evidence in this thing? Are there any questions that are lingering that people are keen to answer and evidence that can help answer?
I think as an investigative reporter, we operate by chasing after questions, chasing after our own curiosity. When there’s news going on, we read a lot and if we come across things that intrigue us, then we see if we can find any way to figure something out, figure out some puzzle.
Can you talk about the differences between the Anglophone media and Chinese language, China-based media when it comes to news.
I’ve worked in both and I think there’s definitely a huge difference.
When I was working at Tencent News, I didn’t feel like I had to explain many things to my fellow Chinese readers, who are familiar with those local subjects, or national subjects.
I know that my readers know all the context information they need to know. I wouldn’t say they know everything, but they probably have a pretty decent understanding of things. I don’t have to explain. For example the kind of things you would explain in the New York Times or even at The China Project…
Like you don’t have to tell people what is a yóutiáo 油条 [Chinese fried breadstick] or whatever?
Yes.
So that is a big difference
Also, in Chinese media, I think partially because of the censorship and limitation on the types of stories we can do, the stories that we tend to be able to tell are human and individual-focused stories. In the Chinese media, we’re passionate about doing features, human stories.
That’s why I spent half of my career doing things like photo essays, being very intimate with subjects, with the people I interview, following their daily life and telling those kind of stories. Which I do think can illustrate big social issues, but we often can’t directly talk about the bigger issue.
Chinese journalists develop the skills to navigate [the constraints]. You can still tell very engaging stories, you can still get a lot of attention from readers. And hopefully, those who are paying attention can reflect on the bigger systematic issues behind them. And that’s how we kind of operate.
So maybe you don’t talk about the dangerous coal mine, that will get censored. You profile the coal miner who’s got a lung disease?
Exactly. I literally did one like that!
I enjoyed doing those stories that allowed me to get on the ground, to really get to know all these individuals in China from very different backgrounds. That’s something that has influenced me, even now when I’m not doing too many human stories.
Now, working for international media, I have to remind myself all the time that the majority of readers are not Chinese people.. It’s a little bit hard to reconcile with, because Chinese readers really don’t have that many high quality publications to read from. There are certain stories that I wish…For example, when I was working on the [COVID whistleblowing doctor] Lǐ Wénliàng 李文亮 story, someone that Chinese people care so deeply about, I really wanted to tell it for Chinese people.
When we are telling this for the New York Times readers, who are mostly American readers, we have to do a lot of introduction and explain why we’re even doing a story about Li Wenliang, why he is important, how he is important. We have to justify that we are doing an in-depth investigation about this figure.
In international media, we are mostly serving readers who are not from the region we cover. How do you balance telling stories about a different place, how do you avoid “othering” the people from that place?
How can [we tell those stories] in a way that doesn’t alienate people in the country, but still acknowledge that there are vast distances between the reader, and the region and the people the story is about?
You’re an editor, you definitely deal with that every day.
Yes, that is a constant challenge. You don’t want to “other” the Chinese people. But when you do critical journalism, that is in fact often the effect. In the U.S., people read stories about something horrible going on in China, and they’re like “Look at those Chinese, doing that terrible thing again.”
It’s difficult to do critical journalism in a way that doesn’t create that effect.
Yeah.
Last question: Since you’ve been working at the New York Times, has it become more difficult to access evidence and data from China?
It seems to me that the kind of journalism you do is becoming more important every day because it’s one of the few ways we still have a direct line into China.
Yes, The Washington Post is currently hiring someone doing this kind of job, which I think is great. I’m very happy to see that another major media is establishing a position like that. Unfortunately, the reality that China is restricting ground access, this kind of reporting is important.
Many researchers and journalists are realizing that this has become like one of the primary reporting methods to cover China. And so they’re using all these databases, but it is not unnoticed. The Chinese authorities, they are noticing it and they are implementing more and more restrictions on those sites.
A very small example, some of the corporate databases that we used previously, you could access from the U.S. Nowadays, you need a Chinese phone or Chinese social media accounts to log in, and those Chinese accounts are usually linked with your Chinese identity numbers. Not only that, they also implement geoblocks, meaning that they can detect your IP address. And you can only use [the websites and databases] if your IP address is mainland China. You can’t even be in Hong Kong. Or Macau.
Right.
And not many secure VPNs provide IP addresses in China. You have all these VPNs that you can use in China to access services outside of China. But it’s not easy to do the reverse.
And partly because of scams and stuff, there’s more detection on the origin of phone calls from overseas to people in China. So when foreign numbers now call people in China, some phones won’t even let it get through, or they will let it go through, but they will show a warning that it might be a scam number.
And then sometimes the police would check on people who receive overseas calls. So it’s also becoming more difficult just to do basic interviews.