Think back to when you took exams in school. Now imagine if your exam was sent to two different teachers of the same subject for grading and you ended up with a passing grade from one and a failing grade from the other. You would rightly wonder what is going on, wouldn’t you? This is the same situation that manufacturers and retailers have to deal with: the same children’s product is tested by two different testing labs and it produces two different test results. Since test results can mean the difference between whether you can legally sell your product or have to destroy it, this is not inconsequential.
The Consumer Products Safety Improvement Act (CPSIA) requires that children’s products be tested by a CPSC-accepted third-party laboratory to determine compliance with all applicable regulations. Congress, concerned about the costs of this testing, asked that we come up with ways to reduce costs. Members of the public have given us comments for reducing the costs of third party testing, and according to studies submitted in several of these comments, it appears standardized testing is not the touted safety panacea due to test result variability.
Below are some problems mentioned in the public comments.
“This represents a major risk for the supply chain if testing in different parts of the world provides different results . . . . The industry experiences test result discrepancy due to different sample preparation methods in laboratories” –Global Apparel, Footwear and Textile Initiative
“Eight different CPSC-firewalled laboratories found the samples’ lead content to range between less than 50 ppm to 262 ppm . . . .Each laboratory reported different results.
“. . . .
“[N]early 50% of FJATA’s members reported that when products failed, the test results were just over the target limit. In total, almost 90% reported that test results were within 10% of target limits.”–Fashion Jewelry and Accessories Trade Association
“The Commission should take action to establish some statistical level of testing error. This will help reduce retesting costs and avoid the need to destroy a batch or lot of consumer products that fails by a margin within the statistical level of error.” –Society of the Plastics Industry, Inc.
These comments and others like them are troubling since they raise basic questions about the reliability of the testing that is a basis for much of our recent regulatory work. We need to recognize the existence of inter-laboratory testing variability and find a solution that is practicable in the marketplace. One way to get at this issue is to publicly recognize a margin of error. By staying silent on this issue, we force businesses to absorb the costs of borderline failed test results. This translates into destroyed product, lost sales, and, ultimately, price
increases for consumers.
I look forward to the staff evaluations on this issue, particularly because they themselves attest to the variability of test results. Staff found certain lead content results “represent the compliance dilemma surrounding any regulatory lead limit” . I hope that we can reform these requirements to remove the confusion and burden our regulations have created.