Why Sheffield Hallam Matters

A closer look at what is driving U.S. forced labor trade enforcement.

[AUTHOR’S NOTE: This post was updated on October 2, 2023 to correct a misstatement of the number of detentions that remain pending from July, August and September 2022.]

A few weeks ago, I was speaking with a member of the Foreign Policy Establishment, and the supply chain and human rights research¹ published by the Helena Kennedy Centre for International Justice at Sheffield Hallam University (SHU) came up in conversation.

My interlocutor, a true Washington sophisticate, asked if I thought companies were taking the Sheffield Hallam research seriously, and, going a step further, asked why they should.

I laughed and assured him that, uh, everyone I know is taking it seriously. I explained that a company referenced in a SHU report (or with a supplier referenced in the SHU report) would ignore it at their peril.

Playing devil’s advocate, he pressed again. Why should a company care about what Sheffield Hallam University, a “whatever-ranked university” has to say about anything? Ouch. But okay, I hear you.

More recently, I was addressing an audience with interests in the solar industry, and they were trying to make sense of one of the recently published SHU supply chain reports, which purports to “create[] a new model of assessment for the solar industry and its supply chains to identify exposure to state-imposed forced labour in the Uyghur Region as supply chains become increasingly opaque[.]” (Emphasis added.)

Essentially, it ranks solar panel manufacturers against the risk that they could export goods to the U.S. in violation of the UFLPA. Ambitious! But also, if you’re an observer of the industry, a little confusing.

Say you’re a solar field developer, trying to buy solar panels from Malaysia for an installation in Texas. Your supplier assures you they are going to send you panels that are traceably proven not to contain any content from the Uyghur Region. But here’s SHU, claiming that this supplier has a “High” or “Very High” risk of sourcing content from the Uyghur Region. What gives?

Also, does this matter? If my supplier *knows the rules*, and swears they’re supplying me *clean merchandise*, and gives me *the paperwork to prove it*, does it matter that an academic research team thinks the supplier presents a high risk of UFLPA violations? Shouldn’t actual proof of the supply chain trump evidence-based inferential reasoning, such that an interested party could justifiably ignore the NGO allegations?

These are all great questions. After all, why should the world be attentive to research published by a group that the Washington Post recently described as operating on a “shoestring-budget”, affiliated with a school the U.S. News and World Report ranks as tied for 1317th Best Global University?

This week, to answer these questions, I want to propose how you should think about SHU research. Then, I want to explain how the law — that is, the UFLPA as currently drafted — causes SHU’s research to have a dramatic impact. Finally, I’ll try to answer the above questions, and provide advice on how companies with a stake in international trade should be responding.

How to Think About SHU Supply Chain Research

The research teams affiliated with Sheffield Hallam University investigating supply chain linkages to probable forced Uyghur labor are producing reports that constitute informal inferential reasoning.

Donning my economist hat for a moment, I would explain that there is a difference between drawing inferences from data using statistical analysis, and engaging in informal inferential reasoning. The former involves inferring properties of a broader population by harvesting a sample of data, and formally testing hypotheses against that sample using statistical or econometric techniques. The latter involves (according to Wikipedia) “making a generalization based on data (samples) about a wider universe (population/process) while taking into account uncertainty without using the formal statistical procedure or methods (e.g. P-values, t-test, hypothesis testing, significance test).” (Emphasis added.)

In a wonky nutshell, that’s the SHU formula. Use sample data to generalize inferences about about a wider universe, and ballpark the right discount for uncertainty.

A longtime reader and friend of FLT once described this newsletter as “come for the legal deep dive, stay for the epistemology.” Sick burn! But in this week’s update, I’m belaboring a lesson in inductive logic because it’s crucial to understanding the contributions SHU is making to this world, and why you cannot look away.

The point is this. SHU is not an outfit that is hyping academic credentials as a proxy for trustworthiness. It isn’t relying on its long-forged connections to the Establishment as a reason it expects the world to take notice.

It isn’t even engaging in statistical analysis, a more empirically rigorous means of proving the properties of a broader population (like all global trade activity) based on samples of trade data.

In a plain nutshell, SHU is looking at publicly available data about supply chains and forced Uyghur labor, and reasoning toward conclusions about where there’s smoke, and where there’s fire.

Consistent with this method of reasoning, the end result of SHU’s reports are probabilistic statements—based on the evidence recited (which varies in nature and character) here is a supplier or brand with “high risk” of Uyghur Region content. Here is a supplier that “could be” sourcing from the Uyghur Region, even “inadvertently”. And lots of claims about what is “likely” or not.

Informal Inferential Reasoning: The Sole Driver UFLPA Enforcement

Now. For the six of you who haven’t fled to greener pastures of Instagram or TikTok after that arduous, hopefully not-too-obvious description of SHU research, here is the point.

People engage in informal inferential reasoning all the time because it’s what makes life interesting. If you’re good at it, you can peer into a space with limited visibility, and end up enhancing your life, because you can correctly predict which neighborhoods are likely to have the best estate sales, or which lawmakers are likely to be corrupt. If you’re really good at it, you can make a killing at DraftKings.

But in this context, informal inferential reasoning doesn’t just purport to tell you which nefarious corporations are profiting from slave labor. It is the sole driver of forced labor trade enforcement. CBP doesn’t need anything more than an inference to decide that a particular supplier should be targeted for UFLPA enforcement. In other words:

Informal inferential reasoning is not just the most important precondition for the enforcement of U.S. forced labor trade laws, it is the only precondition.

As I’ve explained in previous posts, the UFLPA was predicated on a notion that CBP can identify goods that were manufactured wholly or in part with forced labor. The operative provision in the law - Section 3(a) - kicks in “with respect to any goods, wares, articles, and merchandise mined, produced, or manufactured wholly or in part” in the XUAR or by a UFLPA Listed Entity. The law then prescribes the treatment for such goods.

But 99.99%+ of all UFLPA enforcement to date has unfolded around a conceptually prior question: whether or not CBP has stopped goods with the requisite supply chain connection to the XUAR or a UFLPA Listed Entity. CBP tries to pick the best targets it can; importers invariably argue CBP got it wrong.

This process is not contemplated by the UFLPA. It unfolds according to an extra-legal process, governed neither by statute nor regulation, and so far, without the benefit of input from the judiciary.

How, then, does CBP go about trying to pick the best targets it can? A blue ribbon of excellence to anyone who guessed informal inferential reasoning. Or as the UFLPA Strategy Update put it last month : “CBP employs a dynamic risk-based approach that prioritizes the highest-risk goods within a sector based on current data and information.” (Emphasis added.)

For this reason, a widely-read, obsessively-researched and thoroughly-documented write-up from SHU is just what the doctor ordered. CBP, like your uncle during COVID, does its own research. But both the UFLPA Strategy Update and the update to the Xinjiang Supply Chain Business Advisory (published earlier this week) are littered with references to the criticality of NGOs and academic research in informing the present state of enforcement.

You must understand how utterly unique this all is in the landscape of laws enforced against traded goods. The Food Safety Modernization Act prescribes minimum safety standards for manufacturing, processing, packing, and holding of human food, including food that will be imported. But what if it didn’t? What if instead it just charged CBP with inferring, from publicly available data, which imports might contain salmonella, and then . . . . That’s how the UFLPA is written.

There are a million ways you could craft and enforce forced labor trade laws. Under the model currently in effect in the United States, the only thing that drives enforcement are the suspicions that can be inferred from publicly available information.

Advice to Companies

This is why, when speaking to my interlocutor from the Establishment, I told him what I tell all companies. You have to consider carefully what information you disclose publicly about your supply chain.

If a values-driven, internally-motivated researcher can compare your disclosed data to other publicly available data with just the help of a laptop computer and internet connection (never mind a large language model), you have to consider what inferences it will draw. And while your first instinct is probably less is more, remember that in this ecosystem, the less information available to that values-driven researcher, the more likely they are to draw an adverse inference.

A claim about your company or your supplier might strike you as unfairly speculative. Of you might think it is based on information that is out of date. Or maybe you just think you owe nothing to the tied-for-1347th-ranked university. Or perhaps Winnie the Pooh has your tongue. Just don’t forget: allegations of forced labor suffer from the DeBeers effect. Facts change, and supply chains evolve. But an allegation of forced labor lives forever. Even if SHU updates its website.

The most aggressively funded individual trade enforcement priority in U.S. history² isn’t waiting for an elite institution to level imperious allegations. If it smells like forced labor, an army of CBP enforcement personnel is going to probe it. And I must stress above all—in this weird ecosystem, possessing proof of compliance with respect to your own shipments won’t necessarily help you if there is published informal inferential reasoning that confounds you.

Sure, you might win release from a UFLPA detention . . . eventually. But your kids could be in college before you do. Even if you’re not currently planning to have kids.

[Author Note: This paragraph has been updated to correct a misstatement in the number of pending decisions. While still significant, the number of unresolved detentions is not as high as originally described.] According to CBP’s UFLPA Detention Dashboard, as of today—September 29, 2023—there remain 22 detentions from July 2022 that are STILL pending decision by CBP. And another 15 from August 2022. Almost 44 from September 2022, and so on. All of these are now pending for over a year. And the government hasn’t even shut down yet.

The point is, the best proof in the world about how your imported goods are compliant is of limited value if you can still get crushed in an interminable, extra-legal enforcement action, just because some interested third party researchers found reasons for suspicion about your supply chain that they couldn’t resolve based on publicly available information.

If you’re a foreign supplier that wants to preserve your access to the greatest consumer market in the world, this is a reality to recognize. And if you’re a domestic party reliant on global trade—whether you’re a developer of that Texas solar field, or an importer and retailer of stretchy socks—you’re going to want to encourage your foreign suppliers to play ball. And just in case anyone desperately needs an early-aughts internet catchphrase to drive the point home:

Pics or it didn’t happen.

Best wishes to the fine civil servants at CBP, DHS, and elsewhere as you navigate the impending government shutdown.

1. The SHU research in question focuses on documenting known and potential supply chain links to forced Uyghur labor, often tracking connections to the supply chains of globally-known brands and companies.

2. I’m asserting this. Someone find me a counterexample.