Skip to main content

Statement of Commissioner Joseph P. Mohorovic Regarding the Commission’s Provisional Civil Penalty Settlement with Teavana Corporation

May 26, 2016

Throughout my tenure as a Commissioner, I have been concerned about the manner in which the agency calculates, imposes, and settles civil penalty demands for alleged violations of our statutes. I have shared my reservations about penalties I thought were too high[1] and penalties I thought were appropriate,[2] in both instances expressing my belief that we have an obligation – both to regulated industry and to our consumer-protection mission – to be more candid about the thought processes that underlie what we think a given allegation is “worth.”


In our most recent penalty case, I supported our record settlement because I believed the facts warranted it, even though too few of those facts were made public.[3] Today, I find myself unpersuaded by any of the facts – public or not – that our settlement amount is appropriate or that a penalty is justified at all. Additionally, I still believe we are failing in our duty to tell people why we are imposing the penalty that we are imposing.


The Commission has voted to approve a $3.75 million payment from Teavana Corporation (“Teavana”) to settle our allegation that the company failed to report timely incidents involving glass tumblers that were breaking during use.[4] For the reasons described above and detailed below, I cannot join in approving this settlement.


The Reporting Obligation


The Consumer Product Safety Act (CPSA)[5] requires companies to report “information which reasonably supports the conclusion”[6] that one of their products “fails to comply” with one of our rules,[7] “contains a defect which could create a substantial product hazard,”[8] or “creates an unreasonable risk of serious injury or death.”[9]We have interpreted that broad language to mean companies should report even where they do not believe a defect or unreasonable risk exists, where they think a product is safe to stay on the market as-is. We have told companies, “When in doubt, report.”


I fully support taking a conservative stance on the reporting requirement and sending the message that companies should err on the side of over-reporting. That is a prudent approach that emphasizes consumer safety.


But I cannot support taking that conservative approach to extremes, particularly under the new model in which our settlement agreements include very few of the facts underlying an allegation. As I wrote two months ago, each of those settlements is a teachable moment, an opportunity to help the regulated community understand which actions are in-bounds and which are out, as well as the level of sanction they can expect if they stray outside the lines.


Properly using these teachable moments not only gives us a sounder legal and moral basis for imposing punishment, it also provides companies a roadmap (albeit a roadmap to a place they do not want to go). With that roadmap in hand, they can better understand the contours of legal terrain that is intrinsically uncertain. Whether or not a product presents a substantial product hazard is hard enough to determine; figuring out whether or not a particular piece of information could reasonably support that conclusion is even more difficult.


The Teavana penalty settlement does not provide such clarifying guidance. It is another missed teachable moment.


Instruction by Analysis


The simplest tool we could use to capitalize on each teachable moment is the penalty framework outlined in the CPSA and in our rules interpreting it.[10] We could issue settlement documents that analyze – point-by-point – the facts of each case in light of that framework. What, for example, does the Commission think of “the nature, circumstances, extent, and gravity” of the penalized conduct? This would give at least some indication of how the Commission balances the multiple penalty factors and interprets the intricacies of each case.


In this instance, not only does our settlement agreement not provide a regimented analysis, but, as has become common practice, it does not provide sufficient facts[11] for an outside reader to connect the dots herself. The only facts we allege relevant to the purported hazard are that,


[b]etween January 2010 and March 2013, Teavana received numerous reports of the Tumblers unexpectedly exploding, shattering or breaking, including reports of six injuries to consumers who were cut by broken glass or burned by hot liquid while holding a Tumbler that exploded, shattered, or broke.[12]


That’s it.


We do not discuss how that breakage happened, how severe the injuries were, how many reports there actually were, or when across that three-year period the reports came in. We certainly do not provide what would be one of the most valuable pieces of analysis to the regulated community: The date by which we feel confident Teavanahad “information which reasonably support[ed] the conclusion” that they had a product safety problem.


Not only can this terseness not adequately explain the intricacies of the Commission’s interpretation of the facts in light of our statutes, it does not even make clear why the Commission believes Teavana had a reporting obligation at all. As it turns out, I do not believe the company had such an obligation.


My Analysis


It would be easy to say that, where a company has received any report of injury, the company necessarily has information which reasonably supports the conclusion that there is a substantial product hazard or an unreasonable risk and necessarily has a reporting obligation. Such a dogmatic approach would be naïve and would read the words “substantial” and “unreasonable” out of the statute. Some injuries are so minor and so unlikely to repeat that the risk they pose cannot reasonably be considered substantial, and some product-related injuries are not unreasonable. Our own rules recognize that the universe of injuries and the universe of reportable[13] injuries are not coextensive.


Defect which Could Create a Substantial Product Hazard


On the defect prong, we note that


not all products which present a risk of injury are defective. For example, a knife has a sharp blade and is capable of seriously injuring someone. This very sharpness, however, is necessary if the knife is to function adequately. The knife does not contain a defect insofar as the sharpness of its blade is concerned, despite its potential for causing injury, because the risk of injury is outweighed by the usefulness of the product which is made possible by the same aspect which presents the risk of injury.[14]


While a glass vessel is not a perfect analog for a knife, the comparison is instructive. While a knife must be sharp and capable of injuring in order to function properly, the same is not true of a tumbler. Indeed, tumblers generally function best when intact and dull-edged. However, even though that sharpness is best avoided, it is an inherent quality of the material, and – no matter how carefully made – there is always a risk that glass will break. What matters is whether or not the circumstances of a breakage suggest a defect that creates a substantial product hazard. In this instance, I do not believe the circumstances do reasonably suggest such a defect and hazard.


Several of the factors in our rules weigh against finding a defect, here. The utility of the product[15] is clearly high, while the nature of the risk of injury[16] is low. The obviousness of the risk[17] – depending on how we define the risk – is fairly high, as well. If the risk is the fact that the product is breakable, it is hardly earth-shattering news that glass breaks, and the tendency of glass to break when exposed to temperature changes is reasonably obvious, as well. However, if the risk is that a particular glass breaks too easily compared to its peers, that difference may not be obvious and may constitute a defect.


Even if the breakage is a defect, however, it may not be a substantial product hazard, and in fact I do not believe Tevana’s product presented one.[18] The factors in that analysis are similar to the defect analysis, and they concern the “pattern of defect,”[19] the “number of defective products,”[20] and the “severity of the risk.”[21]Again, glass breaks, and there is no evidence the breakage in this case created a severe risk; we have only a few, seemingly minor injuries. As to the number of defective products, we do not detail the total number of breakage incidents or the proportion of the product lines that suffered from the purported defect.


Weighing these factors, I do not believe this product presented a defect that created a substantial product hazard. This is not to say that no breakage of a glass product could ever be a substantial product hazard. Indeed, the Commission is litigating a broken-glass case right now.[22] In that case, the company had reports of serious injuries requiring surgery and causing permanent loss of function, and there were at least five incidents of the product breaking in customers’ hands at the store.[23] Again, glass breaks, but glass breaking that easily – and that frequently – at least strongly suggests a defect that presents a substantial product hazard, triggering the reporting requirement. This stands in stark contrast to glass breaking in the presence of hot liquids, a well-known phenomenon that still may indicate a defect but does not lead inescapably to that conclusion.


Unreasonable Risk of Serious Injury or Death


As for unreasonable risks, we acknowledge that


[t]he use of the term ‘unreasonable risk’ suggests that the risk of injury presented by a product should be evaluated to determine if that risk is a reasonable one . . .. The Commission expects firms to report if a reasonable personcould conclude given the information available that a product creates an unreasonable risk of serious injury or death.[24]


Taking these terms in reverse order, there is no evidence that Teavana’s tumbler created any risk of death or serious injury, which the Commission describes as injuries “which require actual medical or surgical treatment, fractures, lacerations requiring sutures, concussions, injuries to the eye, ear, or internal organs requiring medical treatment, and injuries necessitating absence from school or work of more than one day.”[25] Our settlement agreement notes six injuries, but it does not characterize their severity or suggest any customer required medical treatment.[26]


Even if we consider the cited cuts and burns to be serious injuries, there is still nothing to suggest there was an unreasonable risk of those injuries occurring. The circumstances of the breakage – hot liquids – are a well-known risk of glass breaking, so there is no hazard latency problem. I could still see the risk as unreasonable if it were occurring more often with a particular product than the average similar product, but I do not see that fact, here. Again, our settlement agreement does not discuss how many tumblers Teavana sold, but it does note the company imported 445,000 units.[27] If even half were sold, I have no reason to believe that .003 percent of glass tumblers breaking when exposed to hot liquid tumblers is an unreasonable rate.


In sum, I do not believe Teavana ever had “information which reasonably support[ed] the conclusion that [its] product . . . contain[ed] a defect which could [have created] a substantial product hazard . . . or create[d] an unreasonable risk of serious injury or death.” I do realize that another reasonable person could look at these facts and come to the opposite conclusion. However, such reasonable disagreement demonstrates that, rather than having a clear obligation to report, Teavana was, at most, in a very gray area, and I do not believe it is appropriate to penalize in such a case.




Borrowing our Chairman’s words, our penalty authority is intended to enable us to identify, punish, and shame into compliance those companies who “have bad intent.”[28] In the case of the reporting obligation, it is intended to admonish companies who knowingly withhold information that could have helped us keep consumers safe. It is not a Letter of Marque to exact vengeance for reasonable judgments that, with the benefit of hindsight, we might have made differently. Even if we are going to impose penalties on the latter category, we need to make clear – through our rhetoric, through the details of the cases, and ultimately through the dollar figures – the distinction between those and the actual bad actors.


One of the most esteemed attorneys in the CPSC community recently asked whether, in the agency’s eyes, it was possible for an honest company to make an honest mistake, or whether every alleged failure to report was equally sinful.[29] As discussed, I am not even convinced Teavana made a mistake in this case, but, assuming they did, there is nothing that suggests to me it was not an honest mistake on a very difficult judgment call with scant information. As such, I cannot support penalizing that mistake, certainly not to the tune of $3.75 million. [30]


Moreover, even if I did agree that Teavana had – and failed to fulfill – a reporting obligation, I would still be concerned about how the sparse details of our settlement agreement and how little they do to define that obligation for the benefit of the regulated community. The CPSA and our rules under it do not establish a clear line for when a company must report, but at best a blurred zone of indecision. An agreement that does not analyze the facts under our statutory and regulatory frameworks – that does not even define when Teavana should have reported based on what the company knew at that time – only makes the line between non-reportable quality problems and reportable safety issues even harder to find.


There is, among some in the regulated community, a perception that the CPSC’s focus is “big-game hunting,” prioritizing even minor violations by companies with significant resources over perhaps more significant violations by less well-heeled firms. Whether or not the criticism is accurate, that narrative is only furthered by what has been a constant post-CPSIA drumbeat for higher and higher penalties, with no attempt to link those bigger numbers to more egregious violations or to even explain what we mean by “higher.” Are we simply multiplying old precedents by the same factor as the CPSIA increase? Are we calculating proportional equivalents? Are we just throwing out the old precedents and getting whatever we can?

We are leaving too many of these broader questions just as unanswered as the fact-specific questions about individual cases. Combined with our almost-exclusive pursuit of failure-to-report cases – which are inherently malleable, fact-specific alleged violations – the growing opacity of our approach both to specific cases and penalties writ large only creates distrust between the agency and our stakeholders. When that distrust becomes deep enough and broad enough, more companies will force us to make our penalty cases in court, a dynamic that benefits only their lawyers’ billable hours.

[1] Statement of Commissioner Joseph P. Mohorovic Regarding the Vote to Approve Provisionally a Civil Penalty Settlement with Office Depot (May 22, 2015), available at

[2] Statement of Commissioner Joseph P. Mohorovic Regarding the Commission’s Provisional Acceptance of a Settlement Agreement with Gree Electric (Mar. 24, 2016) (Gree Statement), available at

[3] Id.

[4] Settlement Agreement (Settlement), In the Matter of: Teavana Corporation.

[5] Pub. L. No. 92-573, 86 Stat. 1207 (Oct. 27, 1972) (codified as amended at 15 U.S.C. § 2051-89). The original reporting requirement, which has since been amended, is contained in Section 15 of the CPSA, 86 Stat. at 1221.

[6] 15 U.S.C. § 2064(b).

[7] 15 U.S.C. § 2064(b)(1)-(2).

[8] 15 U.S.C. § 2064(b)(3).

[9] 15 U.S.C. § 2064(b)(4).

[10] See 16 C.F.R. part 1115.

[11] As I noted in my most recent statement about this issue, I realize both parties may have incentives to include as few facts as necessary. See Gree Statement. I believe the Commission needs to provide clear direction as to what is necessary, keeping in mind that, more than resolving an individual case, penalties should provide color to the outline the statutes and rules sketch. Our staff have demonstrated time and again that, when the Commission provides the appropriate leadership, they will execute that vision quickly and capably.

[12] Settlement at 2.

[13] I realize that the term “reportable” is a bit inappropriate, as any injury of which a company learns is capable of being reported, and the real distinction is whether or not a company must report a particular injury. However, sometimes precision must give way to brevity.

[14] 16 C.F.R. § 1115.4.

[15] Id.

[16] Id.

[17] Id.

[18] The fact that the best I can provide is my own analysis of the SPH factors – without the benefit of the opinion of our expert staff – is one of the very few drawbacks of our award-winning Fast Track voluntary recall program. Under Fast Track, a company agrees to conduct a recall, and CPSC agrees not to perform a full SPH analysis to determine the extent of a potential hazard. As a result, the most we can say with confidence about any product involved in a Fast Track recall is that it may have presented an SPH. While that uncertainty can be inconvenient, the program enables us to remove potentially dangerous products from the market very quickly, which goes to the heart of our safety mission. That benefit certainly outweighs the loss of precision, but it does leave an important question unanswered.

[19] 16 C.F.R. § 1115.12(g)(1)(i).

[20] 16 C.F.R. § 1115.12(g)(1)(ii)

[21] 16 C.F.R. § 1115.12(g)(1)(iii)

[22] United States v. Michaels Stores, Inc., and Michaels Stores Procurement Co., Inc., No. 3:15-cv-1203 (N.D. Tex. Filed April 21, 2015).

[23] See Complaint at 5-7, Michaels, No. 3:15-cv-1203 (N.D. Tex.).

[24] 16 C.F.R. § 1112.6(b) (emphasis added).

[25] 16 C.F.R. § 1115.6(c).

[26] Contrast this with our recent settlement with Office Depot over a purported failure to report defective office chairs. I voted against that settlement because I thought the amount of the penalty too high and unsupported, but I acknowledged that Office Depot should have reported, in part because the back injuries involved could reasonably be considered serious.

[27] Settlement at 1.

[28] Chairman Elliot Kaye, Keynote Address at the Annual Symposium of the International Consumer Product Health & Safety Organization (Mar. 2, 2016), available at

[29] Panel Discussion at the Annual Symposium of the International Consumer Product Health & Safety Organization (Mar. 2, 2016) (remarks of Cheryl Falvey).

[30] It may have been useful, for the purposes of communicating our expectations to industry, to have reached a zero-dollar settlement, in which we outlined the facts and explained why they gave rise to a reporting obligation but, acknowledging the difficulty of those facts, did not impose a monetary penalty.

Report an unsafe product