December 18, 2015

As CPSC moves toward implementation of an import surveillance e-filing pilot, I would like to record more of my thinking in voting against this approach.  Electronic Filing of Targeting/Enforcement Data, 80 Fed. Reg. 50827 (Aug. 21, 2015).  In my view, the Commission’s approach continues to disregard the central concern of importers, brokers and other stakeholders, namely the difficulty of providing information from CPSC certificates of compliance as part of the time-sensitive import entry process.  My alternative proposal would have addressed this key concern and still provided CPSC staff with the information they have consistently identified as most valuable for targeting.

Is That What You Call Engagement?


The Commission’s pilot begins a new chapter in what is already a lengthy saga.  It started in 2013, before my arrival as Commissioner, with a proposal to amend the CPSC’s regulation on certificates of compliance, 16 C.F.R. part 1110, often called the “1110 rule” for short.  That proposal would have required (1) three new pieces of information in CPSC certificates of compliance (in addition to the original seven specified in the statute); and (2) electronic filing of certificates for all CPSC-regulated imports.  Certificates of Compliance, 78 Fed. Reg. 28080, 28085, 28089 (May 13, 2013).

The stakeholder response to CPSC’s 2013 proposal was overwhelmingly negative, so much so that the Commission put the proposal on the back burner.  Then, in February 2014, President Obama issued his Executive Order 13659, which directed federal agencies to streamline their import-monitoring activities in coordination with Customs and Border Protection (CBP).  This mandate does not apply to independent agencies such as CPSC, but nevertheless breathed new life into the idea of e-filing certificates of compliance. 

To address the earlier opposition, I urged the Commission staff to engage with importers and other stakeholders before deciding how to proceed.  In particular, I recommended that we hold a workshop on this topic.  This was done in September 2014, and resulted in a healthy exchange of views between the CPSC staff and the Trade.  I also recommended that before making any final decisions about e-filing, we conduct a small-scale pilot to evaluate the tentative approach.  The CPSC staff accepted this suggestion, and earlier this year conducted a series of public webinars  presenting the pilot concept to the Trade. 

Unfortunately, the most candid input from the Trade took place in two meetings held under the auspices of the Chairman’s office, and the other Commissioners were not allowed to participate in these directly.  Eventually, however, I learned that the staff’s plan for the pilot was to require participants to file exactly the same ten fields of information that would have been required by the poorly received 2013 proposal.  Despite the understandable outcry from industry, that approach remained as the centerpiece of the draft Federal Register notice that was sent up to the Commission in June 2015 (see below). 

Why Does E-Filing of Certificates Pose a Significant Challenge?


As I have listened to our stakeholders over the last two years, their #1 concern with e-filing of certificates has been the difficulty of matching a particular entry line with all the certificates of compliance that apply.  The crucial point to understand is that relatively minor variations or differences in a children’s product may require a new certificate.  For example, a doll with green eyes would have to be certified based on the basis of different tests than a doll with brown eyes or blue eyes.  If that were the only difficulty, it might be solved by having “umbrella” certificates that cite the tests for all the intended variations of a product; however, the difficulty does not end there.  A new certificate is also required whenever there is a “material change” in the production process.  For example, if a factory runs out of green paint and starts using a different batch, then new tests would be required for dolls bearing the new paint and new certificates would be required for those dolls as well.  The same goes for any other change in materials during production.

Currently, the CBP entry process does not require an importer to document such minute changes or variations in products.  Even if CPSC demands to see a certificate of compliance, as it has a right to do under the governing law and regulations, the importer has some time—usually at least 24 hours--to determine which certificates may need to be produced.  Cf. 16 C.F.R. § 1110.13 (certificates must be available as soon as the product or shipment itself is ready for inspection).  Under CPSC’s 2013 e-filing proposal, by contrast, it would have become necessary to understand exactly which versions of a product are included in a particular shipment, and which batches they are coming from.  For each version and batch included in a particular shipment, it would become necessary to find the appropriate certificate(s) of compliance so as to be able to link that information to the shipment.  And it would be necessary to complete this complex matching process by the time of entry.

The Staff’s Draft Proposal for the Pilot


The CPSC staff’s draft pilot proposal would have required the same certificate-matching process as the original 1110 e-filing proposal.  Moreover, the seven fields of information from the applicable certificate(s), plus three additional pieces of information, would have to be transcribed, for each entry line, into the Participating Government Agency (PGA) message set (in essence, an addendum to the CBP entry form, No. 3461).  Alternatively, the certificate and supplemental fields of information could be pre-entered into a CPSC “registry” and cross-referenced in the PGA message set by a single numerical entry.  While this approach would reduce the amount of data being typed into the entry form, all of that information would have to be entered into the registry even earlier, and it would still be necessary to match each entry line with all the applicable certificates. 

 

 

My Proposal for the Pilot


To avoid this complex matching process, I started with the principle that the purpose of requiring any additional information for imports is to improve CPSC’s targeting and enforcement.  CPSC currently uses a Risk Assessment Methodology (RAM) to assign a risk score to all products of interest.  We use the risk score to guide our choices of which products to set aside for examination.  If additional information would enrich our risk assessment, we could focus more narrowly on products that pose a significant risk to consumers and avoid detaining products that do not.

In CPSC’s current system, the most valuable information for import targeting is the identification of the foreign manufacturer.  For apparel products, that information is already captured by the Manufacturer Identification (“MID”) code on the CBP entry form.  For other products of interest, however, the MID code may or may not reflect the identity of the foreign manufacturer.  That is because CBP allows importers of non-apparel products to construct the MID code using the name of the shipper rather than that of the manufacturer. 

My proposal would have required importers volunteering for the pilot to enter the name and address of the foreign manufacturer in the PGA message set (unless it was already captured by the MID).  It would also have required importers to provide a more specific identification of the product being imported.  This could be done, for example, by including the Universal Product Code (UPC) or Global Trade Identification Number (GTIN).

In short, my proposal would have expanded the availability of the most valuable information for targeting, but with far less difficulty on the part of importers than is required to match each entry line with all applicable certificates.  In my earlier example, for instance, all of the dolls were being produced by a single manufacturer regardless of the different colors and batches of paint or other materials used.  To be sure, some importers and brokers might need to adjust their current processes in order to make the manufacturer identification available at entry.  That adjustment has already been made by the entire apparel industry, however, and it is surely easier to obtain the name of a product’s manufacturer than it is to invent the detailed accounting procedures required to match certificates with entry lines.

The Chairman’s “Compromise”


After my proposal for the pilot, the Chairman advanced a different approach, which was ultimately approved by the Commission.  The Chairman’s approach will require fewer than ten fields of information to be entered in the PGA message set (or the CPSC registry) and therefore is an improvement over the staff’s original proposal.  Even so, I remained steadfast in my opposition because the approved approach will often, if not always, require the complex certificate matching process that was the Trade’s greatest concern.  For example, it will still require the identification of all laboratories who conducted the testing on which certification of an imported product depends.

At our decisional meeting, the Chairman said his conclusion as to which pieces of information should be required in the pilot reflected the CPSC staff’s “best judgment as to what is really necessary in terms of targeting and enforcement.”[1]    Yet during the public meeting, the import surveillance staff said that it had “no quantitative information” on the usefulness for targeting of test lab identification.[2]  How, then, can there be any basis for the conclusion that this information is “really necessary”?

In my opinion, using test laboratory identification as a basis for targeting is problematic to say the least.  Consider the familiar concern of the “golden sample”—where a manufacturer provides a test lab with samples that are specially made of the best materials so as to get a passing test but then uses substandard materials to manufacture the products that actually go to consumers.  If the CPSC later determines that the product fails to comply, should the test lab be faulted?  If such cases would count against a test lab’s track record, then the leading labs, who do the most testing, are most likely to suffer.  The perverse result is that the involvement of the world’s most experienced test labs would more likely increase risk scores than would the involvement of smaller, less accomplished labs.

Two points add to my frustration at the inclusion of this factor in the pilot.  First, the CPSC compliance staff already has information, in the form of thousands of certificates, which could be analyzed to determine whether test lab identification might be useful for targeting.  No one has attempted to make that case.  Second, it is almost certain that the pilot, as designed, will shed no light on the usefulness of this or any of the information collected because one of the criteria for eligibility to participate in the pilot is that an applicant “[h]ave a history of compliance with CPSC requirements.”  80 Fed. Reg. at 50830.  Likewise, at the decisional meeting, the Chairman himself lowered expectations on this point, cautioning other commissioners that the pilot “is not primarily intended to reach conclusions about each data element.”[3]   

Conclusion


In sum, CPSC’s import surveillance pilot, like the original 1110 proposal, will require participants to develop new tracking systems for matching customs entries with certificate information.  My approach would have secured information that is known to be useful for targeting without requiring the exquisite tracking of variations and batch information that is made necessary by the majority’s approach.     



[1] Decisional Meeting: Electronic Filing of Certificates of Compliance – Pilot Program, https://www.cpsc.gov/en/Newsroom/Multimedia/?vid=74237 [2:13:35].

[2] Id. [19:57]

[3] Decisional Meeting: Electronic Filing of Certificates of Compliance – Pilot Program, https://www.cpsc.gov/en/Newsroom/Multimedia/?vid=74237 [1:44:10-32]