How did the reporters visit Chinese distant-water fishing ships?
The reporting team shadowed Chinese ships in four fishing grounds. These were located in the South Pacific Ocean, near the Galapagos Islands; in the South Atlantic Ocean, near the Falkland Islands; in the Atlantic Ocean, near Gambia; and in Korean waters, in the Sea of Japan. Our trips at sea to visit the Chinese fleet were facilitated by hitching rides with willing partners. In some cases, national law-enforcement authorities or private fishing-boat captains agreed to taxi us to target the various fishing grounds around the world. In other instances, ocean-conservation groups, including Sea Shepherd, EarthRace, and Greenpeace, transported the team to high-seas locations of interest.
How did the reporters convince Chinese fishing captains to let us on board?
Typically, this process took several days while we were reporting at sea and occurred in stages. First, while we were in a fishing ground, we would approach a fishing ship that we hoped to engage. We would keep our distance, far enough that the vessel could see us but not close enough that they might get nervous. Sometimes the ships fled nonetheless. If they didn’t, we began communicating by radio and talking to them about the work they do and conditions on the water and how long they’ve been at sea, trying to stick to innocuous topics. We were always careful to identify ourselves as journalists and explain that we were reporting on the experience, difficulty, and importance of Chinese squid fishing.
After breaking the ice for a little while, we would then use a ship crane to put a skiff in the water, so that we could slowly and respectfully approach the fishing vessel. When the captain could see us in person, we would begin using a handheld radio to communicate with him on the bridge. The point of this was to give him a chance to see us and reassure him that we were not a threat. After this phase, we would ultimately ask his permission to visit him on board his ship, offering him a gift of fresh fruits and vegetables, cigarettes, and candies. Most of these ships lacked such supplies, and we hoped that this would encourage him to let us on board, which sometimes succeeded. Once on board, the captains typically allowed us to tour the ship at will, though almost always we had a minder who shadowed us wherever we went.
Key pieces of technology were distinctly helpful to us while reporting out in the field. To film underwater and to estimate the amount of squid being caught, we used a tethered unmanned submarine droned. To check workers’ activities and estimate their number at night during low visibility, when ships were in transit and their fishing bulbs were turned off, we used an FLIIR heat-sensing drone. To avoid Chinese ships leaving when they saw the reporting team approaching them, we used a long-range rotorcraft Schiebel Camcopter S-100 drone, which has over five hours of flying time.
How did the reporters communicate with crew on vessels they were not permitted to board?
To engage with the crew on vessels we were not permitted to board, we trailed the ships in a smaller and faster skiff to get close enough to throw onto their aft decks a plastic bottle that was weighed down with rice and contained a pen, cigarettes, and hard candy, along with interview questions written in English, Chinese, and Indonesian. We tossed bottles onto four ships: the Lu Rong Yuan Yu 668, Lu Qing Yuan Yu 279, Guo Ji 902, and Fu Yan Yu 715.
After returning to land from reporting trips, we contacted families of crew who had written notes providing phone numbers. Sometimes, those families were relieved to hear about their relatives at sea. One note tossed back to us in a bottle from the Lu Rong Yuan Yu 668 was signed by an Indonesian deckhand named “Jufrei, Cingke,” who provided a phone number for his brother, which we called on March 7, 2022. The owner of the number was Abdul Rokim. He said his older brother Jupri works for Chinese fishing vessels. Jupri is 31 years old from Cirebon, West Java, Indonesia. Abdul Rokim was relieved and thankful to hear that his brother was fine. Jupri had left home around May 2021.
How is the documentary film “Squid Fleet” meant to serve a unique purpose within the overall investigative project?
Tonally, “Squid Fleet” is meant as a counterbalance to the investigative reporting, which provides a critical assessment of the human-rights and environmental record of the Chinese fleet. The film makers—Ed Ou, Will Miller, and Michael Hsu—sought to offer a more intimate perspective on why someone might choose to take one of these dangerous, often exploitative jobs. “Our goal was not only to convey information but also capture the complex motivations of the fishermen,” Miller said. The narrator in the film is a fictional young man from China who recalls the stories that his father, a squid fisherman, has told him about life at sea. The stories are derived partly from court records, labor contracts, and dozens of conversations that we had with crew members. But the account is imagined—an amalgamation intended to capture, in a more poetic fashion than reporting typically does, an array of common experiences.
How did the reporters identify and find information about Daniel Aritonang and other crew members?
When we first heard about Daniel, all we knew was his first name, the date of his death, and the name of the vessel he worked on: the Zhen Fa 7. We also knew from eyewitness testimony that he had been left almost dead at the port of Montevideo with signs of violence and beriberi, and that he had died soon thereafter. We reported in the port and city and spoke with multiple people involved in his case and were able to collect photographs of him when he was first taken to the hospital, as well as the autopsy report. In this report as well, he was only identified as “Daniel.” Even an image we obtained of his passport did not help: it contained only a black-and-white photograph and his first name. These documents allowed us to know what he looked like, but we still needed to figure out more about who he was.
Our first task was to build on these limited data points so we could continue our investigation into Daniel’s death. Having the name of the vessel allowed us to get satellite tracking data to establish how long the vessel had been at sea and where it had traveled while he was aboard. The data showed the vessel had been on its tour since September 2019, when it departed Busan, South Korea, for the Pacific Ocean. It made no port visits between September 2019 and March 2021, when Daniel died, giving us a degree of confidence that he boarded at Busan.
Next, we searched social media—primarily Instagram, Facebook, and Tiktok—for the keywords “Zhen Fa 7,” including in its different forms, such as “ZhenFa7,” “ZhenFa 7,” and “Zhen Fa7,” looking for results that were posted between or near the dates in question. Searching for “Zhen Fa7” returned an album posted by an Indonesian named Ferdi Arnando containing over 150 images. W systematically went through the photos and Arnando’s account and found that a man named Daniel Aritonang had been tagged and photographed several times. The images matched the photograph on Daniel’s passport.
Daniel’s Facebook page yielded a wealth of other information: He was from a town called Batu Lungun, in Indonesia’s Kaur Regency. He had traveled with his friends Hengki Anhar and Fikran. His posts also showed a love of motorcycling. Slowly, we were able to build a picture of who Daniel was and what had driven him to board a Chinese fishing ship. We made contact with the deckhands he worked with and then with his family, eventually visiting Batu Lungun to speak to those who knew him best.
Without Ferdi Arnando’s photos, it is likely that we would never have been able to tell Daniel’s story and make the connections we needed to conduct our investigation.
How did the investigation define distant-water fishing?
Other researchers have offered different estimates based on different moments in time. Other research suggested that China’s distant-water fishing fleet is far larger. Because the Yellow Sea, South China Sea, and East China Sea are disputed, China has counted its ships fishing in these waters as part of its domestic fleet. Partly because it included Chinese ships fishing in these seas, a study published in 2020 by the Overseas Development Institute found that the distant-water fleet includes roughly 16,996 vessels. But this count does not imply that all are operating currently, simultaneously or consistently in foreign or international waters.
The Food and Agriculture Organization defines a distant-water vessel as any vessel fishing in “all FAO major fishing areas other than those adjacent to the flag state.” In this definition, Chinese vessels registered with the North Pacific Fisheries Commission but fishing within FAO Area 61 would be excluded, but given their registration status and the fact that the North Pacific Fisheries Commission extends beyond FAO Area 61, we chose to include these vessels.
Tabitha Mallory provided an estimate of 1,989 vessels in 2013. Daniel Pauly provided an estimate of 3,432 vessels in 2014. Greenpeace provided an estimate of 3,432 in 2016.
How did the investigation make the comparison between the size of the Chinese fleet and that of other fleets?
The investigation sought to compare the Chinese fleet to some of the next largest or most high-profile fleets. Those included fleets from Taiwan, the U.S., and the E.U. Estimates in 2018 for the Taiwanese distant-water fishing fleet range from 1,100 vessels, according to Greenpeace, to 1,800, according to the Environmental Justice Foundation. An estimate for the size of the United States’ distant-water fleet was 225 in 2015, and maritime researchers say that it has shrunk since then. The E.U. had 259 ships in its distant-water fishing fleet, according to the 2021 Annual Economic Report on the EU Fishing Fleet.
How did the team estimate that there are 6,500 vessels in China’s distant-water fishing fleet?
An invaluable source for building this estimate was the Allen Institute for AI’s Krakken database, an enormous dataset of Chinese fishing vessels compiled by Roberto Mielgo Bregazzi, an independent fishing-industry analyst. The Krakken database, which is based on data from over 100 existing public records, contained information on active and no-longer-active fishing vessels throughout the world, including International Maritime Organization registration status, authorization records from regional fishery management organizations, national ship registers, classification societies, and IUU red lists. All the data sources used for Krakken date from 2020 and thereafter. They reflect the most up-to-date sources available.
To establish the size of China’s distant-water fleet, we extracted the following information in sequence from the Krakken database:
|All China-flagged vessels with a registered IMO number||2,541 vessels|
|All China-flagged vessels registered with one or more regional fishery management organization||350 vessels|
|All China-flagged vessels that were inspected by Chinese authorities outside Chinese waters||213 vessels|
|All China-flagged vessels registered as exported or seized by Chinese customs agents||78 vessels|
|All China-flagged vessels registered with the Chinese distant-water fishing association||966 vessels|
|All China-flagged vessels granted a high-seas fishing authorization by China’s Ministry of Agriculture or a fishing license by a third country authorizing it to operate inside a foreign exclusive economic zone||221 vessels|
|All China-flagged vessels that had a fishing presence detected outside of Chinese domestic waters via MMSI AIS signaling and/or by port or state authorities when traveling inbound, outbound or docking||723 vessels|
|All fishing vessels that have significant associations with China—that is, that are built in or previously flagged to China, and/or a current or a previous Chinese name, operator, or owner||1,477 vessels|
After gathering all of this information , we then did everything we could to clean this data and ensure that we weren’t double counting vessels. This methodology gave us a final estimate of 6,569 vessels.
How did the reporters build a master list of squid vessels in China’s distant-water fishing fleet?
The Allen Institute for AI’s Krakken database determined whether a vessel was a squid jigger based on whether it was approved for squid fishing by a regional fisheries-management organization, the gear type it was registered with, images of the vessel, and other information. The Outlaw Ocean Project team then cross-referenced this information with data from other platforms.
We checked vessel ownership information provided by Krakken using Equasis, a non-profit that provides information on a ship’s ownership, management, and safety and inspection histories. The reporting team also used this information to establish the corporate structures of the Chinese fishing industry, which was essential for connecting the vessels that catch the squid to the factories that process it. The Chinese corporate database Aiqicha was of critical importance in this regard as well.
Vessel activity was verified by checking MarineTraffic and Global Fishing Watch, with ships that showed activity tracks after June 2021 considered still active for the purposes of our investigation, using a reference period of 2018 to 2022.
The relevant regional-fishery management organizations also provided invaluable information on the Chinese squid fleet, including their identifying information, whether they were approved to fish, and up-to-date photographs of many of the vessels in the fleet.
When we corroborated the Krakken information with these various sources, and excluded vessels that were inactive, duplicate, or owned by other countries, we came up with a final total of 751 vessels.
How did the investigation match documented human-rights and environmental abuses to individual vessels on the list?
The investigation pulled from dozens of sources to create a comprehensive list of Chinese squid jiggers associated with various human-rights and environmental concerns from 2018 to 2022. These included media investigations, port records, case reports (from NGOs such as the Environmental Justice Foundation, C4ADS, Trygg Matt Tracking, and Greenpeace), satellite-tracking data, social-media posts, photos, videos, and research from regional fisheries-management organizations. We relied heavily on two remote observation tools to track ships and identify illegal or suspicious activity, including when vessels turned off their transponders for longer than seven days, a practice prohibited by Chinese law. These tools were Skylight, a fisheries-monitoring tool built by the Allen Institute for AI, and Global Fishing Watch.
To source the information, we went systematically through our list of 751 vessels, keyword searching in both English and Chinese script on multiple search engines (including Google, DuckDuckGo, Baidu, and Sogou), on social networks (including Facebook, Twitter, Instagram, YouTube, TikTok, Weibo, and Douyin), and on other databases (among them those maintained by private organizations such as Marine Traffic and regional-fishery management organizations such as the North Pacific Fishery Commission. Our searches uncovered not only cases of abuse and illegal fishing but hundreds of photographs and hours of video that gave insight into the challenges and dangers of life on board these ships.
Chinese court records offered a rare window into the labor abuses affecting Chinese deckhands, which included the trafficking of workers, typically from poorer inland regions of the country. Labor contracts provided by former deckhands from Chinese fishing ships revealed patterns of trickery and debt bondage that sometimes ensnare crew members. The team also reviewed the ships’ labor histories, ownership details, subsidy receipts, identity changes, worker contracts, litigation pasts, port-visitation records, and catch certificates.
We put asterisks in our database for vessels tied to concerns or crimes. An asterisk required: (1) that an allegation of a human-rights or environmental crime had been made against a specific vessel or operator, and (2) that the allegation came from a credible source (e.g., a crew member, journalist, human-rights or environmental watchdog, government record). Among the criteria used to determine whether to add an asterisk were:
|Human rights sub-category||Specification|
|Labor violation||Recruiter or employer practices fail to adhere to labor-law requirements of the flag or coastal state, or in the country of beneficial ownership (for vessels) or of origin (for workers).|
|Forced-labor risk||Recruiter or employer practices are consistent with one or more indicators of forced labor but do not meet the threshold of forced labor as defined in ILO Forced Labour Convention, 1930 (No. 29) and interpreted in the ILO Hard to See, Harder to Count framework.|
Vessel / vessel operator has been designated a forced labor risk by the U.S. Treasury.
|Forced labor||Recruiter or employer practices are consistent with one or more indicators of forced labor that meet the threshold of forced labor as defined in ILO Forced Labour Convention, 1930 (No. 29) and interpreted in the ILO Hard to See, Harder to Count framework.|
|Critical health and safety violation||A vessel has a health or safety violation that has the potential to imminently cause serious injury, illness, or death.|
|Crew member death(s)||At least one crew member has died on or off the vessel, and there is no credible evidence to indicate that the death was accidental and/or not the fault of the employer and their representatives.|
|Fatal occupational accident||A work-related accident that resulted in death occurred onboard the vessel.|
|Beriberi symptom(s)||An individual working on the vessel developed at least one symptom consistent with Vitamin B1 (thiamine) deficiency disorders as defined by the World Health Organisation. Primary symptoms include:Swelling and pain in the limbs;Loss of sensation and paralysis;Muscular atrophy;Shortness of breath; andCardiovascular failure.|
|Violence, intimidation, or injuries on board||Credible reports of violence, intimidation, or avoidable injuries experienced by workers on vessels.|
|Crew abandonment||Recruiter or employer has failed to cover the cost of repatriation; left the fisher without the necessary maintenance and support; or has otherwise unilaterally severed their ties with the fisher, including failure to pay contractual wages for a period of at least two months, as defined by ITF here.|
|Illegal Fishing||Unauthorized||A vessel is identified fishing without a permit or a license in a country’s EEZ, even though it lacks the permission of that state, or operates in contravention of its laws and regulations.|
|Gear||A vessel is identified using prohibited gear (e.g., drift nets or mesh size that is too small or nets when they only have permits for jigs).|
|Noncompliance||A vessel is identified infringing observer regulations, evading arrest or law enforcement, or not complying with other regulations.|
|Species and Bycatch Related||A vessel is identified keeping undersized fish or fish that are otherwise protected by regulations (e.g. shark finning) or a vessel is identified purposefully targeting species outside the scope of its licenses.|
|Transshipment||A vessel engages in an unauthorized transshipment or a transshipment that is in violation of the laws and regulations of a coastal or flag state or an RFMO.|
|Quota Related||A vessel is identified fishing without a quota or over a quota for certain species.|
|Zone/Season||A vessel is identified fishing in closed areas or during closed seasons.|
|Unspecified fishing offense||A vessel is identified transmitting AIS in the North Korean EEZ and exhibiting patterns of behavior indicative of apparent fishing activity as classified by Global Fishing Watch. UN Resolution 2397 (2017) clarifies that the transfer of fishing rights is included in the sectoral ban on seafood established under Resolution 2371 (2017).|
|Fish dumping/discarding||A vessel dumps catch in violation of a coastal or flag state’s laws.|
|Unregulated Fishing||Use of Multiple MMSIs||A vessel uses multiple maritime mobile service identity (MMSI) numbers—that is, identification numbers that are assigned by the flag state and intended to be unique.|
|Unreported Fishing||Not reporting or showing required documentation to the relevant national authority.||A vessel fails or refuses to show documentation (e.g., fish logs) to a relevant national authority.|
|High Risk for IUU||Fishing in Russian Waters||A vessel is identified transmitting AIS in the Russian EEZ and exhibiting patterns of behavior indicative of apparent fishing activity as classified by Global Fishing Watch. Risk for illegal fishing stems from Executive Order 14068, which prohibits the import to the U.S. of fish, seafood, or derived products of Russian origin.|
|AIS Transmission Gaps||A vessel did not transmit AIS for a minimum of a consecutive 7-day period. A minimum period of 7 consecutive days was selected to reduce the possibility of an equipment malfunction or lost signal being the primary cause of the transmission gap. Additionally, AIS transmission gaps occurring outside of fishing areas (e.g. in port, in domestic waters) were excluded.|
How did the investigation identify transshipments at sea between problematic Chinese squid-fishing vessels and refrigerated cargo vessels?
China’s fleet of reefers regularly transships with vessels that have documented histories of illegal fishing or human-rights abuses, adding a layer of opacity that is difficult to penetrate. Our goal was to establish exactly how often these transshipments took place. To start, we used data from the satellite firm Global Fishing Watch, a database called Krakken, and C4ADS, a research firm, to build a spreadsheet listing all Chinese-owned reefers.
Then we turned that data over to the satellite firm Skylight, which created a log of every time a reefer on our list was tracked in a position adjacent to a squid vessel in a pattern that suggested transhipment over the past five years. The Skylight algorithm registers an encounter when two transmitting vessels travel within 250 meters of each other at approximately the same speed. That gave us a total of over 12,000 transhipment events since 2018. We then cross-referenced the unique vessel identifier codes of squid vessels we had identified as engaging in problematic behavior with that data to establish the total number of transshipments involving vessels that we had identified as worthy of concern. This yielded a total of 2,640 transshipments between problematic Chinese squid-fishing vessels and refrigerated cargo vessels.
How did the investigation connect problematic fishing vessels to refrigeration vessels that were transhipping to seafood-processing plants?
The most difficult connection to make in this supply chain was from reefer to processor. In one case, we hired investigators in China to watch a cargo ship unloading in Shidao port and covertly follow trucks as they delivered the seafood to factories. In most other cases, we connected the dots from vessel to exporter through what is known as corporate network analysis. In order to make connections from fishing vessel to exporter, we started by drawing up a list of all the vessel operators. This list included owners of problematic fishing vessels as well as owners of reefers that had documented transshipments with fishing vessels that had engaged in abuses.
Next, we identified the owners of these vessel operators and examined associated company structures. Where necessary, we mapped ownership layers up to the fifth tier (the owner of the owner of the owner of the owner of the vessel operator) until we identified actual people—the “beneficial owners” who ultimately control and profit from vessels involved in IUU fishing and labor-rights abuses. In most, but not all, cases there was only one beneficial owner who maintained majority control.
We used Aiqicha, a website with publicly available information about businesses in China, to examine corporate ownership structures and relationships between individual managers, directors, and beneficial owners. For selected corporate groupings, we mapped the corporate structure in more detail, using graphical representations. This enabled us to better identify interlinkages between groupings of companies, as well as connections between companies based on shareholders and key personnel.
We concluded that seafood caught by a fishing vessel had probably been supplied to a factory if certain conditions were met, notably:
- If a fishing-vessel operator, reefer operator, and processing factory were all controlled by the same holding company, or ultimately owned by the same individual(s).
- If a fishing-vessel operator and a processing factory were controlled by the same holding company, or ultimately owned by the same individual(s).
Where the reefer operator and the processing factory had a common ownership distinct from that of the fishing-vessel operator, we recorded this in our findings as a weaker connection that was more open to interpretation—weaker because many smaller distant-water fishing-vessel operators contract out the logistics of the freezing, storing, and transferring of frozen seafood (often called “cold-chain logistics”) to cargo vessels owned by third parties.
What other methods did the investigation use to trace problematic fishing vessels to seafood-processing plants?
The investigation carried out keyword searches online to find imagery of squid catch being processed at factories in Shandong in social-media posts, corporate websites, and news articles. Those searches included running through each of the names on the list of fishing vessels found linked to problematic behaviors to find footage of catch from those ships being stored at seafood-processing plants or on the factory floor for processing. More general keyword searches uncovered videos and images of bags of squid inside processing factories. This imagery was examined to identify the vessel names on those bags, and these names were then checked against the list of problematic vessels. If no match was found on the list of problematic vessels, the names were also checked against the wider list of squid vessels in order to see where squid from different fishing companies was being sent for processing. The investigation also looked at identifying features in the source material to corroborate where the videos or photos had been taken.
How did the investigation identify North Koreans working at seafood-processing plants?
U.S. law prohibits the import of products from factories using migrant North Korean labor. Multiple UN Security Council Resolutions also ban the use of this labor beyond 2019. However, North Korean overseas labor is still being dispatched to parts of China, including Dandong. We reviewed multiple reports and analyses from South Korean news agencies, non-profits, and North Korea–focused sites like Daily NK. Multiple reports said that in 2022 up to 80,000 North Koreans lived in Dandong and worked at clothing and seafood processing plants.
A 2022 report from the Committee for Human Rights in North Korea gave some more specific leads, naming two seafood processors, Dandong Taihua Food and Donggang Julong Food, as having North Korean workers on site. Aside from that, details were scant, and information on North Korean workers in China was heavily policed by Chinese censors.
The Chinese government has largely scrubbed references to these workers from the internet, but certain unorthodox search terms on Douyin, the Chinese version of TikTok, yielded results. But, using the search term “North Korean beauties,” my team and I found several videos on Douyin, the Chinese version of TikTok, that appear to show female seafood-plant workers, most posted by gawking male employees.
The videos offered a rare window into the average Chinese person’s view of the issue. The status of the women, and the conditions they are subjected to, are the subject of heated discussion by Chinese underneath the Douyin videos. In just one series of videos, showing North Korean workers at a tool factory in Jilin province, hundreds of comments were posted. One commenter says, “Don’t underestimate these girls, they have a strong sense of nationalism and identity and are self-disciplined!” Another commenter rebukes them: “Haha! They have no choice; you go ask them if they yearn for the United States or South Korea.” The original poster of the video then responds: “We can’t talk to them! Even if we do, they ignore us!” Other comments claim that the workers have no choice but to obey orders, or “their family members will suffer,” while several highlight the fact that the wages paid to the workers go straight to the North Korean government. Others comment that the workers have been “brainwashed.”
How did the investigation calculate the total tonnage of exports from Chinese seafood plants that use North Korean workers?
We searched across publicly available trade data for each Chinese seafood plant that we identified as using workers from North Korea. For each company, we collected, structured, and cleaned available export data between 2017 and 2023. The data fields available for each shipment included product description, importer, exporter, gross weight, arrival date, and more. The total tonnage of exports for each company was calculated by using the total of gross weight reported in each shipment.
What definitions did the investigation use to determine that seafood workers transferred from Xinjiang are victims of state-imposed forced labor?
The United Nations, human-rights organizations, and academic experts generally agree that since 2018, the Chinese government has systematically subjected Xinjiang’s predominantly Muslim ethnic minorities to forced labor across the country through labor-transfer programs, which use coercive methods in worker enrollment and deny workers the ability to leave employment. This is known as state-imposed forced labor. The U.S. has actively enforced an import ban on goods produced through all forms of forced labor since 2016, and in 2021 issued the Uyghur Forced Labor Prevention Act, which creates a legal presumption that Xinjiang minorities involved in the production of export goods are subject to state-imposed forced labor. The U.S. law places the burden of proof on business actors to demonstrate that their supply chains are free of forced labor.
Labor transfers, both within and outside of Xinjiang, are a well-documented component of China’s systematic subjugation of ethnic minorities in the region. Experts believe that the number of transfers began to increase in 2019, after the decline of the mass-internment and political re-education system, and then have continued to grow since the advent of the pandemic.
Shandong Province is one of the country’s main seafood-processing hubs and is located thousands of miles from Xinjiang. The investigation documented the Shandong government’s role in the so-called Xinjiang Aid program. Under the auspices of this program, Shandong authorities have constructed securitised internment facilities, boarding schools, “welfare parks” for the young and elderly, and industrial complexes across southern Xinjiang since at least 2010. Officials encourage Shandong companies to build “satellite factories” in Xinjiang at which Uyghurs and other minorities have been subjected to forced labor, according to NGO and media reports. Shandong government authorities have also encouraged these companies to oversee labor transfers thousands of miles east to worksites there.
Experts and academics say that coercion is built into the Chinese labor-transfer program within and from Xinjiang. As part of the process by which the government forcibly recruits Xinjiang workers, party cadres regularly visit many rural Uyghur households. Unemployed villagers are matched with labor-demand quotas from eastern factories, rounded up early in the year at recruitment fairs, and dispatched east by coach, train, and plane. People who refuse transfer to waged employment risk attracting the attention of authorities and being detained. Full employment is required in all households under the current five-year plan for Xinjiang, and 400,000 party cadres are assigned to unemployment monitoring that is intended to prevent transferred Uyghurs from leaving work.
Transfer workers undergo “militarized” vocational training delivered by private and state-run firms. Uyghurs transferred in groups of 30 or more are typically accompanied by security guards and minders. We documented transfers to 10 Shandong seafood plants, some of which involved uniformed police, security guards, and members of the Anti-Terrorist Detachment of the Chinese security services.
By using Xinjiang labor, the seafood-processing companies aren’t doing anything illegal, because transferring Chinese people from one province to another is not a violation of Chinese law. But under U.S. law, any product handled by workers from Xinjiang is banned from import.
How did the investigation get visual windows on labor transfers from Xinjiang to other provinces in eastern China?
Videos uploaded to Douyin by government accounts offered views on the recruitment and transfer process. The videos showed labor transfers in southern Xinjiang, from areas such as Kashgar, Hotan, and Bayingolin Mongol Autonomous Prefecture. These regions were targeted because they had been identified in our research as areas working with the Shandong government through aspects of its Xinjiang Aid program.
The investigation identified government Douyin accounts sharing footage of so-called “village work teams,” the party cadres and sometimes state-owned company representatives who implement grassroots recruitment efforts. For example, a May 2023 video (described in the story) showed people in neat lines, signing contracts while being monitored by officials in army fatigues. A February 2022 video (also described in the story) showed thousands of Uyghurs attending a “job fair” at an internment camp in southern Xinjiang.An April 2023 video (described in the opening sequence of the story) showed more than eighty men and women, dressed in matching red windbreakers, standing in front of the train station in Kashgar, a city in Xinjiang, China, awaiting their transfer.
Another source documenting the transfers were videos uploaded by Uyghurs themselves after they had been transferred to Shandong to work in seafood. In one case, a Uyghur woman from southern Xinjiang uploaded a 2021 video to Douyin showing her in a departure lounge among a group of people all dressed identically in patterned red windbreakers with red flowers pinned to their chests. Red flowers are a common feature of labor transfers, according to experts. The same red flowers pinned to the chests of men and women transferred to one Shandong seafood plant by plane during the early months of the COVID-19 pandemic. A local news article about the transfer says these were migrant workers from Xinjiang. Local news shows an image of a China Southern Airlines flight attendant assisting a Uyghur wearing a red flower with their landing card.
How did the investigation identify Xinjiang labor transfers to Shandong seafood-processing plants?
We undertook targeted Chinese language searches—on Google, Baidu and Sogou—to investigate links between companies known to be involved in human-rights abuses at sea, and Xinjiang forced labor. We combined company names with a range of indicator words and phrases, such as “Xinjiang,” “labor dispatch” and “migrant worker.” Next, we did broader searches, using terms such as “seafood processing factory” instead of company names. We found company press releases and state media articles detailing labor transfers from Xinjiang to seafood-processing factories in Shandong over a period ranging from 2008 to February 2023. We had these materials translated by native Chinese speakers.
The investigation identified eight Shandong seafood corporations implicated in the Xinjiang labor-transfer program. Together these corporations operated a total of 58 processing plants, the majority (44) authorized to export to the U.S., with a little under half (21) authorized to export to the E.U. Local news and company materials, including videos, helped us identify labor transfers to ten specific seafood-processing plants located in four Shandong cities: Weihai, Rizhao, Yantai, and Qingdao, the final two of which are both global whitefish-processing hubs.
How did the investigation connect Xinjiang Zhongtai Group to Uyghur forced labor in Shandong seafood-processing facilities?
We found newsletters on the website of a Shandong seafood conglomerate, Chishan Group, discussing how corporate management met with party officials in 2020 and 2021 to resolve worker shortages at several Chishan factories. One of these newsletters detailed a meeting in mid-2021—attended by representatives of a company called Xinjiang Zhongtai Zhihui Modern Service Co. Ltd., along with government officials from Xinjiang—during which the parties agreed to accelerate labor transfers to the Chishan Group. We used a Chinese corporate-registry aggregator and open-source intelligence gathering to identify Xinjiang Zhongtai Zhihui Modern Service as a division of the state-owned Xinjiang Zhongtai Group, which is involved in recruitment and labor-dispatch activities in partnership with the Chinese government. This was the first indication that Xinjiang Zhongtai Group—a Fortune 500 company notorious for its involvement in labor transfers and forced labor within Xinjiang—implements transfers from Xinjiang to other provinces.
How did the investigation calculate the total number of seafood plants connected to Xinjiang forced labor?
We found user-generated content posted in the last 12 months showing Xinjiang minorities working at ten seafood enterprises, for which we also had state media and/or company statements describing Xinjiang labor transfers. The ten plants are operated by five corporate entities, and each group owns two facilities:
|Company name||Corporate group|
|Qingdao Lian Yang Aquatic Products Co. Ltd.||Tianyuan|
|Qingdao Tianyuan Aquatic Products Co. Ltd.||Tianyuan|
|Rizhao Jiayuan Foodstuff Co. Ltd.||Shandong Meijia|
|Rizhao Meijia Keyuan Foods Co. Ltd.||Shandong Meijia|
|Rizhao Rirong Aquatic Products Co. Ltd.||Rongsense|
|Rizhao Rongxing Food Co. Ltd.||Rongsense|
|Rongcheng Haibo Ocean Food Co. Ltd.||Chishan|
|Shandong Haidu Ocean Food Co. Ltd.||Chishan|
|Yantai Longwin Foods Co. Ltd.||Sanko|
|Yantai Sanko Fisheries Co. Ltd.||Sanko|
How did the investigation estimate the minimum number of Xinjiang ethnic minorities deployed to Shandong seafood factories since 2018?
Our reporting cites “over a thousand” Xinjiang ethnic minorities. The real number is likely much higher.
We created an estimate of the minimum number of people transferred based on numbers detailed in state media reports and company materials. But precise figures on transfers were only available for some companies. Consequently, the companies referred to in this estimate are different from the companies at which we verified the presence of Xinjiang workers. Where there were no available figures for processors (Qingdao Lian Yang, Yantai Longwin, and Rongsense Group), we added numbers to reflect known Uyghur Douyin users uploading from the plant.
We estimated that at least 1,380 people had been transferred by the government from Xinjiang to work in the following Shandong seafood companies:
|Company name||Number of workers transferred from Xinjiang|
|Weihai Wendeng Xinghe Foods Co. Ltd.*||500|
|Yantai Sanko Fisheries Co. Ltd.||346|
|Qingdao Tianyuan Aquatic Products Co. Ltd.||165|
|Shandong Meijia Group||139|
|Qingdao Lian Yang Aquatic Products Co. Ltd.||5|
|Yantai Longwin Foods Co. Ltd.||4|
(The calculation for Weihai Xinghe Food was based on a 2020 article about the company that said it has taken in 2,200 Uyghur workers from Xinjiang since 2009; approximately 200 a year. At this rate, intake since 2018 would be 1,000; we have conservatively estimated 500 Uyghur workers since 2018.)
The investigation captured video footage of Uyghurs working at seafood-processing plants that could not be identified through geospatial and comparative analyses, suggesting the deployment of Xinjiang workers to Shandong seafood-processing factories is widespread. This is supported by local news sources, which reported that in 2022, 469 individuals from “Xinjiang ethnic minority” groups were deployed to seafood factories in the city of Weihai alone. An article from the previous year, citing deployment to the same district of seafood processors, refers to as many as 800 “foreign workers.”
How did the investigation calculate the total tonnage of exports from Chinese seafood plants using workers from Xinjiang?
We searched across publicly available trade data for each Chinese seafood plant we identified using workers from Xinjiang. For each company, we collected, structured, and cleaned available export data between 2018 and 2023. The data fields available for each shipment included product description, importer, exporter, gross weight, arrival date, and more. The total tonnage of exports for each company was calculated by using the total of gross weight reported in each shipment.
Due to strong evidence of group-level deployments, we opted to include two additional Shandong Meijia Group factories (Rizhao Meijia Aquatic Foodstuff Co. Ltd. and Rizhao Jia Tian Xia Foods Co. Ltd.) and three additional Chishan Group companies Rongcheng (Jiamei Seafood, Rongcheng Jinyuan Aquatic Food, and Rongcheng Runtong Aquatic) in these calculations.
How did the reporters find Douyin videos uploaded by Uyghurs working in seafood?
Following the discovery of a video posted to Douyin, the Chinese version of TikTok, featuring a Uyghur male working at a seafood plant, we watched thousands of clips to identify additional Uyghur users. We found a total of 72 Uyghur Douyin users who had worked at – or were working at – ten seafood processing plants, and archived over 300 videos uploaded by these accounts. Archived Douyin material spans September 2018 to July 16, 2023.
Two Uyghur researchers independently reviewed each of the videos archived by the investigative team, listening to voice clips, reading uploaded text and translating song lyrics. The analysis indicated that Uyghurs working at Shandong seafood facilities were using coded content to convey critical perspectives on their experiences through humor, poetry and song. Posts alluded to strong sentiments of loss and separation, and offered direct and indirect references to coercion and involuntary transfer, as well as commentary on poor working conditions, and the relationship between Han Chinese and Xinjiang ethnic minorities. The independent researchers also scanned the clips for clues about where transferred workers had come from, and were able to pinpoint various locations in southern Xinjiang’s Kashgar, Hotan, and Bayingolin Mongol Autonomous prefectures.
How did the investigation verify the location of Shandong seafood-processing plants receiving Xinjiang workers?
In order to get a fix on the processing facilities, we used cell-phone footage from workers in seafood plants that had been published on Douyin, combined with a variety of open-source imagery including company promotional materials, photos in local news, satellite imagery, and Baidu street view. Douyin videos offered internal and external views on seafood factories. The identity of seafood plants was verified through a combination of geospatial analysis, comparative analysis (e.g., comparing workshop features, setups, and worker uniforms with corporate promotional material), text translation (some videos showing legible text provided strong indication of factory identities) and, in a minority of cases, Douyin user comments on videos and the appearance of tiny company logos in the backgrounds of clips uploaded by Uyghur workers. Geospatial location offered firm verification for videos showing external views of all ten facilities. Comparative analysis verified some plant interior scenes, but the majority of plant interior footage was posted by Uyghur users who also had uploaded views on plant exteriors to their Douyin reels. Some Douyin videos that were collected depict seafood plants that remain unidentified.
How did the investigation find out about social audits conducted at Shandong seafood-processing plants using Xinjiang labor?
We communicated our findings to hundreds of North American and European companies buying seafood from Shandong plants using workers from Xinjiang. In many cases, companies pointed to social-audit reports to assert there was no evidence of forced labor at the implicated factories. We asked importers and their customers to tell us when and what types of social audits had been conducted, who had conducted them, and what they had found with respect to Xinjiang workers. Although in most cases, companies declined to answer our enquiries, usually referring to commercial confidentiality, some buyers confirmed audit dates, auditor identities, and the standard used (Sedex Members Ethical Trade Audit, or SMETA). A few even shared audit reports. In order to further ascertain whether Xinjiang workers were being detected by social audits, we spoke to the auditors, and standards and certification bodies, about our findings.
How did the investigation identify companies importing seafood from Chinese processors connected to abuses at sea and on land?
Trade data allowed us to track exports from processing plants to stores and restaurants outside of China. We obtained data from a variety of sources, including Chinese customs and private aggregators of import data from North American and European countries. We also searched company websites for information about customers and export approvals. Footage posted to Douyin by workers in seafood plants often featured seafood packaging showing useful details like vessel names or brand labels. We also used optical character-recognition searches to look for examples of packaging and documentation featuring the unique export codes of Chinese processing plants (export approval codes and health marks issued by government authorities, certification codes issued by the Marine Stewardship Council and the Aquaculture Stewardship Council).
How did the investigation connect companies importing seafood tainted or associated with crimes or other concerns to consumers?
Trade data told us which companies were importing seafood from Chinese suppliers of interest, but in most cases we needed to look at the next link in the chain: the customer of the importer. We searched through importers’ websites, product catalogs, and social-media profiles in order to ascertain who they were supplying. We used OpenCorporates, an open database of companies, to identify the ownership and corporate structures of companies in the U.S., Canada, Europe, and elsewhere. We also used a U.S. government trademark database and a global brand database to expand our list of brands owned by companies importing seafood tainted by forced labor or illegal fishing to search across catalogs and online stores for major grocery chains and food service groups.
We identified unique codes for importers—health marks issued by market state authorities, certification codes issued by the Marine Stewardship Council and the Aquaculture Stewardship Council—and conducted optical-character recognition searches for those codes on product packaging and commercial documents. We visited dozens of stores in 12 states (Alabama, Arkansas, California, the District of Columbia, Florida, Georgia, Hawaii, Massachusetts, New York, South Carolina, Texas, and Virginia) and several countries (Australia, France, Ireland, and the United Kingdom) to obtain images of seafood packaging in order to establish, using those unique codes, the origin of the seafood. We reviewed product listings on major retailer and foodservice distributor websites to identify seafood products that were produced or sold by target importers and that matched the types of seafood sourced from Chinese processors of interest.
Finally, communications with seafood importers and their customers—and, in some cases, the next tier of buyers—helped clarify our findings on the connections among fishing vessels, processing plants, and global consumers.
How did the investigation connect companies importing seafood tainted by or associated with potential crimes or other concerns to public procurement chains in North America and Europe?
We looked at government-contract databases, such as the European Union’s tender database, which contains detailed records of tenders and contracts for all European Union countries and European agencies, and USASpending.gov, which provides federal spending data, to identify the main companies supplying frozen seafood to government agencies. We used trade data to identify any companies that received procurement contracts and also imported from Chinese companies tied to seafood associated with potential crimes, including those using Uyghur labor. We also investigated major government suppliers’ product lists and catalogs to ascertain if they were supplied by companies that imported seafood associated with potential crimes and risk indicators.
In the UK, whitefish is supplied by companies associated with our investigation to public institutions such as schools and hospitals. The supply is typically through intermediaries, working under what are known as “framework agreements” that identify government-approved vendors. Implicated seafood suppliers were identified through reference to brand names and the use of unique Marine Stewardship Council (MSC) codes on primary school menus for 2023 and other documentation.
U.S. public procurement rules have various exemptions that allow local-level buyers for school-lunch or other federally supported programs to purchase food and other products if they are looking for better options in terms of price, quality, quantity or availability. In the U.S., half of the fish sticks served in public schools have been processed in China, according to the Genuine Alaska Pollock Producers, an industry group. They said their research was derived from a review of purchasing records of their members. States and large school districts have historically used USDA grants to buy seafood directly from commercial vendors, much of which is sourced through China, the organization said. Foods purchased by the USDA have only accounted for about 20 percent of what is served in schools, according to the organization, which means the remaining 80 percent is purchased mostly by local buyers.
What is an example of one method that was used to connect the Zhen Fa 7 to consumers globally?
The investigation conducted surveillance in Shidao port in China as the Lu Rong Yuan Yu Yun 177, a reefer that had transshipped with the Zhen Fa 7, as it unloaded its catch. The Lu Rong Yuan Yu Yun 177’s catch is transferred by an orange crane that lifts tons of bags of frozen squid in nets and drops the bags onto the dock, where about 20 workers in protective clothing—baggy white overalls and red hard hats—stack the bags of squid and load them onto trucks.
Cargo from the vessel was loaded onto one truck in bags marked “Rongcheng Wangdao Dayang Aquatic Products Co., Ltd,” which is the registered owner of the Zhen Fa 7 and, according to records held by the China State Administration for Industry and Commerce, is a subsidiary of Shandong Jiekou Fishery Group Co. Ltd. The truck was subsequently followed from Shidao port to a squid processor, Rongcheng Xinhui Aquatic Products Co. Ltd. Rongcheng Xinhui Aquatic Products supplied Lagoon Seafood Products with six shipments of squid, totalling 118 tons between 2021 and 2022, according to bills of lading. Lagoon Seafood Products is based in Quebec, Canada, and its Blue Tide brand of frozen squid rings are for sale via the Costco Wholesale website (in Canada). Since 2018, Cal Ranch has imported at least four shipments of squid from Rongcheng Xinhui Aquatic Products Co. Ltd., according to export records. Performance Food Group sells Cal Ranch’s Pacific Ranch brand squid. (Although data suggest that it is possible that squid caught by the Zhen Fa 7 was used in the Pacific Ranch brand, the connection does not definitively establish that the Zhen Fa 7’s squid was used by Cal Ranch.)
The Zhen Fa 7 is also connected to the Rongcheng Zhenhai Processing Plant via shared ownership with the Jiekou Group, and to the Rongcheng Nanguang Processing Plant and the transshipment vessel Lu Rong Yuan Yu Yun 177, both of which are owned by the Shidao Group. Additionally, it is connected to the Rongcheng Haibo Processing Plant, Shandong Haidu Processing Plant, Rongcheng Runtong Processing Plant, Shandong Shuangdu Processing Plant, and Rongcheng Jiamei Processing Plant, all of which are owned by the Chishan Group, whose reefer, the Lu Rong Yuan Yu Yun 009, transhipped with the Zhen Fa 7 in 2018. At least two Chishan plants, the major global exporters Shandong Haidu and Rongcheng Haibo, use Xinjiang labor.
Roughly a third of all squid processed in China is handled by Haibo and Haidu.
We tracked where squid from Chishan processors ends up using trade data, corporate publications, and social-media footage taken inside the plants, with the shipments sent around the world for distribution to major retailers, foodservice distributors, small corner stores, and local restaurants. For example, Pecheries Oceanic Fisheries Inc, a Canadian seafood importer and distributor, and the U.S. company Pacific American Fish Co Inc (PAFCO) have each been supplied squid by two Chishan processors: Shandong Haidu and Rongcheng Haibo. Ruggiero Seafood, headquartered in New Jersey and a supplier to the major foodservice providers Sysco and Performance Food Group, has imported scores of shipments of squid from Shandong Haidu and Rongcheng Haibo since 2018, and strongly suggests that Haidu and Haibo squid imports went to these suppliers, among others.
The squid processor Rongcheng Haibo, owned by the Chishan Group, supplies Costco in Australia and New Zealand, according to a source at the company in 2022. Grand BK has also received squid from Rongcheng Haibo, with import records showing two shipments totalling 40 tons in 2021 and 2022. Grand BK Corporation “serves as a major food distributor for HMART Group, which is one of the largest Asian-style chain supermarkets in the U.S.,” according to its website. Rongcheng Haibo has also supplied squid to the Spanish company Froxa, with trade records and authenticated video from Chinese social media showing the supply-chain relationship. Froxa imported at least four shipments of squid from the processor between 2021 and 2023, according to bills of lading.
For more detail on these and other supply chains connected to our investigation, see “Bait to Plate.”
What method was used to decide which countries to include in the “Discussion” database?
As part of our supply-chain research, we contacted hundreds of companies around the world who were in some way linked to our investigation. We wanted to give the companies we found at various stages of the supply chain—from fishing and processing to importing and retailing—an opportunity to hear about and respond to the investigation’s findings. In some cases, those companies were direct importers of product from plants at which we found evidence of Xinjiang labor, while others were contacted because they are customers of distributors whose supply chains included plants or fishing companies connected to illegal fishing or human-rights abuses at sea.
Corresponding with companies further down the supply chain helped to clarify ongoing supply relationships and also helped inform our understanding of the wider traceability problems in the seafood industry. We also reached out to the companies that owned the fishing vessels we had found to be potentially tied to human-rights, environmental, or other concerns, and we contacted the owners of the processing plants to which we had traced catch from those vessels. The processing plants at which we found evidence of forced labor were contacted for comment, as were the auditing bodies who carried out social audits at the plants or created the framework for conducting those audits.
We compiled this correspondence in “Discussion” to show what we asked of these companies and organizations, and to provide their responses. “Discussions” is not intended as an exhaustive array of all those we contacted: we excluded a small number due to factors including export dates falling outside the timeframe of our investigation, a very low number of shipments within that period, or the business ceasing to operate.
To be clear, a company’s inclusion in “Discussion” does not mean it engaged in or is connected to wrongdoing. In many cases, we contacted companies because we were investigating a high number of supply-chain lines in order to build a comprehensive picture of seafood’s global distribution after leaving factories in Shandong, China, and not because we necessarily believe they were, or are, involved in illegal or illicit activity.
What method was used to decide which companies to include in the Ubiquity graphic?
We compiled our supply-chain research into a consolidated graphic to demonstrate the widespread distribution of seafood potentially tainted by or associated with crimes or other concerns to groceries, foodservice groups, and governments across North America and Europe. It is not intended to be a comprehensive illustration of every company we identified in our supply-chain analysis. We excluded certain companies due to factors including a very low number of shipments within our investigative timeframe or because we could not demonstrate the exact seafood product or species imported was distributed at the supermarket, foodservice group, or government agency.
In other cases, we didn’t include certain small and medium-sized grocers, distributors, and foodservice groups because they didn’t meet our threshold for market reach. For example, some grocery stores identified are limited to one location or one municipality. Additionally, the Ubiquity graphic does not expand beyond the first-level buyer from an importer. In some instances, we identified further buyers downstream. For example, seafood distributed to certain foodservice companies or government agencies ultimately ended up in federal prisons, public-school cafeterias, hospitals, or restaurants.
For a more comprehensive overview of our supply chain analysis, see “Bait to Plate.”