CHAPTER 4 SECTION 337 EXCLUSION STATISTICS
4.1 D ETECTING S YSTEMATIC E RRORS
Systematic errors are biases caused by non-random factors. This research expects to encounter two types of systematic errors:
• Unknown Factors: This is the kind of error to be examined. We will find out what factors are more likely to affect USITC’s decision-making.
• General Exclusion Order Request Success Rate Errors: This type of error is unneeded and can be eliminated. However, elimination of this type of error requires more detailed information currently unavailable.
4.1.1 Unknown Factors
The common law and statutory criteria for General Exclusion Order relief have been introduced in Section 3.4.5. This research will detect the non-legal factors that are most likely to influence USITC’s determinations. According to 19 U.S.C. §1337(d)(2), the USITC shall issue a Limited Exclusion Order against an unfair act unless:
(A) a general exclusion from entry of articles is necessary to prevent circumvention of an exclusion order limited to products of named persons; or
(B) there is a pattern of violation of this section and it is difficult to identify the source of infringing products.
The law requires that the USITC to grant general relief upon a determination that the complainant has established a violation of either §1337(d)(2)(A) or §1337(d)(2)(B) (see also
Chart 1.3 (1)). The USITC’s issuance of the General Exclusion Order relief establishes that the complainant has successfully proven either statutory criteria for General Exclusion Order relief as to a general population of actors (both named respondents and unnamed parties).
Request for general exclusion
§1337(d)(2)(A)
§1337(d)(2)(B)
Limited Exclusion
General Exclusion Yes
Yes No
No
CHART 4.1 GENERAL EXCLUSION ORDER DECISION-MAKING (19U.S.C.§1337(D)(2)) The key question — whether to grant general relief — can only be answered by “yes”
or “no.” This research uses Pearson’s chi-square test to estimate if a non-legal factor could affect the USITC’s tendency to issue general relief. These non-legal factors, which supposedly excluded as factors supporting a legal determination, may be influential in reaching a legal determination. This research focuses on five factors that are identified below:
• Passed Pearson’s chi-square (χ²) test:
o The period of determination (grouped into 5- or 10-year periods).
• Did not pass Pearson’s test:
o The estimated remaining life of the patent-based exclusion order (normalized using the quartile function).
o The patent’s technical feature (using a published categorical system developed by Hall and Jaffe).
o The patent’s apparent value (using the “Cited By” count provided by the USPTO as of early 2006).
o The ALJ assigned to the investigation (four most prolific ones and a catch-all OTHERS category).238
238 The two currently in-office ALJs are found to be influential to the decision making.
The non-legal factors identified above are not the only factors that may affect the USITC’s tendency to issue (or not to issue) a General Exclusion Order. They also may not be the most reasonable or promising candidates. They are, however, a useful set of factors within this research’s current data collection capability, to test the available data.
4.1.2 General Exclusion Order Request Success Rate Errors
With available data, this research would systematically underestimate the success rates of General Exclusion Order issuance. The amount of this error is unknown.
To win a General Exclusion Order, the complainant must request it (Case A).239 The online request data is incomplete. This research used all unfair acts as the denominator (Case B), rather than using the number of actual requests (Case C). This overestimated denominator would make the success rates too low and more difficult to reject the null hypothesis. The magnitude of the error cannot be estimated with a particular degree of accuracy. This systematic error could be corrected if data about actual requests were available. Perhaps the range of relative uncertainty may be corrected in an extension of this research if additional sources of corroborating information sources become available.
quests clusion Re General Ex
clusions General Ex
ions All Exclus
clusions General Ex
≤
CHART 4.2 GENERAL EXCLUSION ORDER SUCCESS RATE ERRORS
4.1.3 General Exclusion Order Review Model
The correctness of USITC’s determinations may be verified by a complete de novo review of each investigation’s questions of fact. For a number of practical reasons —
239 19 C.F.R. 210.16(c)(2).
including, but not limited to, the confidential nature of most USITC investigations — this task is beyond any researcher’s ability. To determine the rejection of General Exclusion Order requests — a more manageable task — this thesis presents a four-tier review model:
• USITC’s internal review: The original General Exclusion Order determination.
• Administrative review: Presidential veto (very rare).
• Appellate court review: Federal Circuit/Supreme Court review.
• International trade agreements: GATT/WTO finding.
CHART 4.3 GENERAL EXCLUSION ORDER REVIEW AND DISPUTE RESOLUTION PROCESS
4.1.3.1 General Exclusion Order Determination
Within the USITC, Commissioners vote to determine if a General Exclusion Order.
can be issued. The USITC’s fact finder is the ALJ. That is why our research initially assumes that the ALJ factor may influence the final determination. If a Commissioner has a strong personal view that is different from the opinion, he or she can file a dissenting opinion.
For example, in Large Video Matrix Display Systems and Components, Inv. No. 337-TA-75, Commissioner Stern proposed to issue a Cease-and-Desist Order instead of the first Limited Exclusion Order (see Section 3.4.5). This remains a viable — yet unlikely — option.
4.1.3.2 Presidential Veto
Generally, presidential review, through the auspices of the USTR, focuses on the policy reasons supporting the issuance of exclusion order relief, which typically has only a weak correlation with the statutory criteria for issuing a General Exclusion Order.
In
Multi-ply Headboxes, Inv. No. 337-TA-82, President Reagan vetoed an exclusion
order in favor of issuing a Limited Exclusion Order (see Section 3.4.5). Such vetoes are rare, and the veto exercised in this instance no doubt exerted substantial influence on the USITC’s development of Limited Exclusion Order relief.4.1.3.3 Federal Circuit/Supreme Court Review
The issuance of General Exclusion Order may be unchallenged. Respondents may be inclined to accept a General Exclusion Order if, for example, it tends to be equally disruptive for potential competitors as for them. A Limited Exclusion Order, on the other hand, would exclude named respondents, while permitting unnamed suppliers to enter the U.S. market, if they do exist. Under this scenario, a named respondent could be shut out of the market, as a practical business matter, even if the Limited Exclusion Order were later overturned on appeal.
If a self-claimed non-infringing third party’s products were excluded because of general relief, it could argue non-infringement or patent invalidity in a Customs proceeding.
Absent intervention in an investigation, it would appear impossible for an infringing third party to request for a Limited Exclusion Order. The existence of unnamed infringers can be used, however, to prove a “pattern of violation.” One known case illustrating this pattern was
Canadian Tarpoly, discussed in Section 3.3.2.2, below.
4.1.3.4 GATT/WTO Finding
Other than launching trade retaliation, recourse to the GATT/WTO provides the only regular means of securing international intervention in a USITC investigation. In the past, such recourse has been shown to be highly influential. Using the WTO’s dispute settlement does not require a complainant to exhaust local remedies.240 However, only member states can participate GATT/WTO dispute settlements. This path is not available to U.S. importers.
When a non-U.S. respondent wants to challenge the Section 337 in the GATT or WTO, it has to persuade its national government to represent itself.