Cyber Power Decision Keeps the Lights On Origin

The Court of International Trade has issued the much anticipated (at least by me) decision on the merits in Cyber Power Systems (USA) Inc. v. United States. It's a bit of a roller coaster ride. The plaintiff clearly hits some bumps along the way, but there is a thrilling conclusion. So, buckle up.

The case is about the country of origin of five models of uninterruptible power suppliers ("UPS") and one surge protector. The UPS is essentially a glorified backup battery for computers and other devices. In the event of a power failure, the UPS kicks in to power the device and may assist in a "graceful shutdown." Modern UPS include printed circuit board assemblies that monitor the battery to ensure it is fully charged, monitor the available supply of electricity, switch to the battery when needed, provide status reports and other functions. All of that takes some firmware stored in chips on the boards. The surge protector is a simpler, but still electronic device, that smooths out spikes in powerline voltage that might damage electronics.

Cyber Power imported the UPS and surge protectors from the Philippines and declared the that to be the country of origin. Customs determined the country of origin to be China and excluded the merchandise. Cyber Power protested the exclusion. This decision, following a trial to establish the facts, is the review of the denied protest.

The trade bar has been watching this case very closely. Since the Trump Administration imposed Section 301 duties on products from China, there has been a renewed interest in country-of-origin determinations. Companies have been moving final production out of China while continuing to source components from China, leading to the question of whether the final assembly is enough to produce a product of the new country rather than a product of China that was assembled in, say, Mexico.

The traditional statement of the substantial transformation test in an assembly context that the assembly operation must be complex and produce a new and different article of commerce with a new name, character or use. This rule can be traced back to 1908. Over time, the Courts and Customs have added a "gloss" to the test that increases complexity and reduces certainty for companies trying to make origin decisions. This includes decisions that focus on the origin of those components that embody the "essence" of the finished article and whether the parts had a pre-determined end-use prior to the assembly (as opposed to being generic parts with multiple potential uses). 

In my mind, the fundamental flaw with the current approach is that it compares the condition of the parts before and after assembly rather than comparing the parts to the finished article. A zipper from Japan, for example, will only be used as a zipper and will still be a zipper after sewn into a coat in Mexico. The zipper would seem to not be substantially transformed. On the other hand, the zipper is no longer a separate commercial entity. It is wholly integrated into the coat and should not prevent the coat from originating in Mexico. 

Normally, I would jump right to the good stuff, but there is an important bit of background: in court, someone has a burden of proof. Typically, the burden of proof falls on the plaintiff. There are two parts to the burden of proof. The first is the burden to produce, which means the plaintiff has to give the court something to go on. The second is the burden of persuasion. In this case, the burden of persuasion is to establish the facts "by a preponderance of the evidence." If the plaintiff fails to meet the burden, it will lose. 

This case involves six items. For five of them, plaintiff ran into proof problems that were sufficiently severe that it lost on those claims. It is important to understand what happened because this decision shows how Customs and possibly the Department of Justice might comb through records to identify gaps and anomalies that might form the basis of an argument that an importer has failed to prove origin. Let's look at how that happened here.

In an effort to prove its case, plaintiffs presented witness testimony and documents. Apparently, the job of the principal witness was to corroborate the documents. The plaintiff submitted the documents one would expect including spec sheets describing the imported merchandise, bills of material showing the components and their countries of origin, bills of material for printed circuit board assemblies (PCBAs), the production process timeline, a manufacturing process flowchart, and standard operating procedures for the assembly workers.  There was also evidence showing that firmware coded in Taiwan had been "burned" into the PCBAs. There was some evidence showing the firmware upload occurred in China for PCBAs made there or in the Philippines for other boards. 

Plaintiff was presumably using this evidence to show what happens in the Philippines and to establish that it is a complex process producing a new and different article of commerce with a new name, character, and use. In other words, to show that a substantial transformation occurred. 

The Justice Department raised questions about the documents and the testimony. In particular, the government pointed out that the witness' signature was not consistent across the documents, raising questions of whether he actually signed them. The witness was also unclear on the corporate structure as it relates to a supplier that is related to the producer. Finally, the witness made errors in his explanation of a video introduced to illustrate the production process. None of that was good, but it does not seem to have been fatal. 

The real problem seems to have been that the documents submitted in evidence were not clearly connected to the production of the imported goods. The spec sheets, for example, contained no information about the manufacturing process. The bills of material provided lists of the components used to assemble the finished product but did not explain how the parts came together or in what quantity each part was present. The bigger problem may be that the bills of material were undated. The production timeline and manufacturing process flowchart were also undated. There is more to this, but I will leave it with the fact that the witness testified that he was "familiar" with the process but had not observed it personally.

The Court reviewed the evidence in detail and, ultimately, could not connect it to the imported merchandise. As a result, the Court concluded:

that Plaintiff has failed to carry its burden of producing evidence to show that five of the six models of subject merchandise . . . were substantially transformed in the Philippines. Plaintiff’s evidence does not establish what operations occurred in the Philippines to produce these subject devices to permit the court to conduct a substantial transformation analysis.

The fact that the plaintiff lost on these claims cannot be interpreted as saying anything about the correct analysis of substantial transformation beyond the threshold question of what evidence the Court considered important.

But, the Court did give us something on that front as well. For the item for which the plaintiff met the burden of proof, the Court has some good things to say. First, the Court agreed with prior decisions that substantial transformation is reviewed on a case-by-case basis. Next, the Court reiterated its prior decision rejecting an approach that asks whether only essential or critical components have been substantially transformed. The Court also rejected an approach that performed a component-by-component analysis. Finally, while not mentioning predetermined end-use, the Court reiterated that "the intended use of components may provide some insight as to whether the assembly of those components into the finished merchandise accomplishes a change in use that indicates a ‘substantial transformation,’ such a consideration is but one of many for the court to consider as part of the ‘totality of the evidence." 

Instead, the Court looked to evidence of "the potentially transformative processing" in relation to the nature of the product. That means whether the finished product has a different name, character, and use than the parts from which it was made. The decisive paragraph is this:

Because the court finds that the entirety of the [manufacturing] occurred in the Philippines, the court need not make a determination as to whether [the] UPS assembly process alone constituted a substantial transformation. The CP600LCDa began its manufacturing journey in the Philippines as a set of components not yet functional as a power source of any kind. . .. After several stages of manufacturing, each involving numerous steps directed toward changing the electronic properties of the device as a whole, the CP600LCDa left the Philippines as a fully functioning UPS. It is undisputed that that the CP600LCDa is capable of providing “battery backup (using simulated sine wave output) and surge protection for desktop computers, workstations, networking devices, and home entertainment systems,” and that, thanks to its programming, “is able to provide real time status and alerts of potential problems. ... Even without detailed evidence describing the assembly stage of UPS production, the change from all of its components to its ultimate finished product as a UPS device is a change so marked as to shift the burden of proof in Plaintiff’s favor.

The Court concluded:

Indeed, the CP600LCDa’s Philippine manufacture satisfies all three prongs of the substantial transformation test: a change in name (from a set of PCBA and UPS component parts to the finished, functioning UPS Model No. CP600LCDa), a change in character (from component parts not yet capable of being electronically programmed to a device capable of performing a number of intelligent functions), and a change in use (from component parts to a device geared towards a specifically identified purpose: protecting against power outages).

If this decision holds after the inevitable appeal, this may be a meaningful reset of the law of substantial transformation. 

There are, of course, limits that remain in place. Simple assembly operations should raise concern. Manufacturers should continue to be prepared to show that the assembly process requires a number of steps, time, tools, and some degree of skill. Adding components to an otherwise largely complete product will not suffice. A radio without knobs has the same name, character, and use as a radio with knobs. [Yes, I know many of you have never seen a radio with knobs. Exhibit A.] Drawing the exact line is hard, which is why these are always done on a case-by-case basis. On the other hand, the Court seems to have rejected the essence or critical components tests, including the notion that the location where the main PCBAs are assembled will generally be the country of origin. It also seems clear that the proper point of comparison is between the parts and the finished products rather than the parts before and after assembly.

That is good for importers trying to manage their supply chain. Clarity, consistency, and predictability are always helpful for compliance. It is too soon to make major decisions based on this opinion, but this is a positive step.