This article was first published by The Conversation. Space.com's Expert voices: Op-Ed and Insights was contributed by the publication.
Gavin McDonald Senior Project Researcher, University of California Santa Barbara
Economically speaking, high seas fishing is somewhat mysterious. These open ocean areas that are not under the jurisdiction of any country are often considered to be high-effort and low-payoff fishing areas, but fishers continue working in them.
My job as an environmental data scientist is to use data and analysis techniques to answer crucial questions about natural resource management. My colleagues from the Environmental Market Solutions Lab discovered that high-seas fishery is often a very unprofitable venture. Even when you take into account government subsidies, this is the case.
Yet, high-seas fishers continue to catch in record numbers. This suggests that financial support is available beyond government subsidies.
Open ocean fishing has a problem with forced labor. However, it is not easy to track the scale of this problem. Our team was puzzled by the number of vessels fishing high seas, even though it isn't profitable. We speculated that many of these vessels may be subsidised through low labor costs. If the vessels used forced labor, these costs could be even zero.
Combining our data science expertise with satellite monitoring and input from human rights professionals, we created a method to predict whether a fishing vessel is at risk of being forced labor. The study found that between 2012 and 2018, up to 100,000 people may have been forced to work on these vessels.
Magazines Direct: Incredible Earth $22.99 Have you ever wondered why or how these amazing things happen? How did the Earth get its name? How can we predict weather? What is the process of fossil formation? How do earthquakes occur? Which animals glow in darkness? View Deal
Unusual behavior from forced labor
The International Labour Organization defines forced labor as any work or service that is demanded of anyone under threat of any penalty. It also includes any work or service for which the person is unable to stop working, and any work done on the high seas. Unfortunately, forced labor is well-documented in the fishing industry, but it remains to be seen how widespread the problem is.
Our team wanted more information about forced labor in fisheries. We were able to answer a key question that drove the project: What if vessels using forced labor behave fundamentally differently from vessels that use it?
We first examined 22 vessels that had used forced labor to answer this question. Global Fishing Watch, a non-profit organization that promotes ocean sustainability through near-real-time fish data, provided their historical satellite tracking data. We used it to identify commonalities in the behavior of these vessels. We met with representatives from Liberty Shared and Greenpeace to learn more about what to look out for in satellite monitoring data.
These indicators include vessel behavior such as spending more time at sea, traveling further from ports than other vessels, and fishing longer hours than other boats. These vessels could be seen at sea for several months at a stretch, which can lead to suspicion.
Once we had an idea of the risks that could indicate forced labor, our team used Google data scientists to help us use machine learning techniques to search for similar patterns in thousands more vessels.
It is shockingly common
Data from 2012 to 2018 was used to examine 16,000 fishing boats. We examined 16,000 fishing vessels using data from 2012 to 2018. Between 14% and 26% showed suspicious behavior, which suggests that they may be exploiting forced labor. As many as 100,000 people could have been forced to work in these six years. We don't know if those boats are still in use or how many high-risk ships may be present on the oceans today. Global Fishing Watch reports that there were almost 13,000 vessels in the squid-jigger, trawler, and industrial longliner fleets as of 2018.
Squid jiggers use bright lights to lure their catch; longliner boats follow a line with baited hookeds; and trawlers drag fishing nets through water. The highest proportion of vessels that showed signs of forced labor were found to be Squid Jiggers, closely followed by longliner fishing boats and, in a lesser degree, trawlers.
Our study also revealed that forced labor violations occur in all major ocean basins. These violations can be found both at the high seas as well as within national jurisdictions. In 2018, high-risk vessels visited ports in 79 countries, with most of these ports being in Africa, Asia, and South America. Canada, the United States of America, New Zealand, and several European countries are also notable for their frequent visits. These ports are both potential sources for exploited labor and transfer points to seafood caught under forced labor.
Our model, as it stands now is a proof-of-concept that needs to be validated in the real world. We were able show that 92% of suspicious vessels were flagged by the model after it had assessed vessels already captured using forced labor. Our team plans to continue validating and improving the model by collecting more information about known cases of forced labor.
Turning data into action
Our team developed a predictive model to identify vessels at risk of being forced laborers. Our results can be used to complement existing efforts to promote transparency in supply chains and combat human rights violations. Our team currently uses individual vessel risk scores for determining forced labor risks for certain seafood products.
We hope to be able to use the model in the future to free victims of forced labor in fishing, improve working conditions, and prevent other human rights violations.
Global Fishing Watch is now helping us to find partners from governments, enforcement agencies, and labor organizations that can benefit from our findings to better target vessel inspections. These inspections provide opportunities to catch offenders as well as more data that can be fed into the model to improve its accuracy.
This article was republished by The Conversation under Creative Commons. You can read the original article.
Follow Expert Voices to keep up with the debates and issues. You can also join the conversation on Facebook and Twitter. These views are the author's and may not reflect those of the publisher.
