Nat’l Ass’n of Mfgrs v. SEC, No. 13-5252 (D.C. Cir. Aug. 18,
2105)
After the
AMI en banc decision, the panel
granted rehearing of National Association of Manufacturers v. SEC, 748 F.3d 359
(D.C. Cir. 2014).
The panel, over a
dissent, confirmed its initial ruling that the conflict mineral SEC disclosure
rule was unconstitutional, in the process saying some dumb things about what
constitutes commercial speech (the panel didn’t think product labels count) and
some very troubling things about legislative factfinding (apparently not
allowed in the face of controversy).
Basically, the panel majority strongly
disagrees with the
AMI en banc, so
there.
The AMI en banc
majority held that Zauderer covers
more than mandatory disclosures that cure misleading advertising, and also
covers disclosures that serve other governmental interests, such as allowing
consumers to choose American-made products.
The majority here began by responding to the dissent, which
pointed out that US law has a lot of disclosure requirements for securities
issuers, and First Amendment challenges to them really died in the 80s. But—SEC, get nervous—“Charles Dickens had a
few words about this form of argumentation: ‘“Whatever is is right”; an
aphorism that would be as final as it is lazy, did it not include the
troublesome consequence, that nothing that ever was, was wrong.’” And anyway,
even the SEC agrees that the conflict minerals disclosure regime is very
different from the “economic or investor protection benefits” that SEC rules
ordinarily strive to achieve.
Zauderer doesn’t
cover all commercial speech, only “advertising or product labeling at the point
of sale,” so Central Hudson
applied. The Supreme Court, after all,
didn’t apply Zauderer in Hurley v.
Irish-American Gay, Lesbian and Bisexual Group of Boston, 515 U.S. 557 (1995)
or United States v. United Foods, Inc., 533 U.S. 405 (2001), and corporations
generally have free speech rights. The
conflict minerals disclosures are supposed to be made on company websites and
reports to the SEC, so they aren’t advertising, even assuming they’re
commercial speech. [Like I said, get
nervous, SEC.]
The dissent takes this on very well, but I also find the
majority’s analysis here disingenuous; there is a large and contentious
literature about what constitutes commercial speech, but Hurley is not part of it, because no one thought that Hurley’s parade involved commercial
speech. The distinction Hurley made was commercial/noncommercial,
not advertising/commercial speech that is not advertising; “advertising” is
standard shorthand for commercial speech.
The majority noted the dissent’s objection to the anomalous
result that requiring producers to put the conflict minerals disclosure on
their product boxes—a much more onerous requirement—is judged by more relaxed
standards than the SEC reporting requirement, but said that was AMI’s fault for “stretching Zauderer to cover laws compelling
disclosures at the time of sale for reasons other than preventing consumer
deception.” And the disingenuousness
intensifies! Apparently Zauderer doesn’t apply when a commercial
entity engages in false or misleading commercial speech that isn’t
“advertising”? That is nonsensical. The panel majority doesn’t like AMI, I get it, but there are reasonable
ways to limit AMI and unreasonable
ones. Perhaps this is basically a dare
to the overall circuit to take this case en banc if the government so desires,
but the reasoning is just embarrassing.
Anyway, even if AMI
and Zauderer applied, the conflict
minerals disclosure would still violate the First Amendment, because it might
not work to end war in the Congo. Though
the court assumed that “ameliorat[ing] the humanitarian crisis in the DRC” was
a sufficient interest under AMI and Central Hudson, disclosure hadn’t been
shown to be effective at achieving that interest. Statements by two Senators, members of the
executive branch, and a United Nations resolution were insufficient, especially
given the cost of compliance, which was in the billions, and hundreds of
millions of dollars each year. (I do not understand what the cost of compliance
has to do with effectiveness, but let’s just call that a conflation of several Central Hudson steps; it’s hardly the
worst offense of this opinion.) The
prospect that companies will simply avoid mineral suppliers with a connection
to the DRC wouldn’t reduce the humanitarian crisis: “The idea must be that the
forced disclosure regime will decrease the revenue of armed groups in the DRC
and their loss of revenue will end or at least diminish the humanitarian crisis
there. But there is a major problem with this idea – it is entirely unproven
and rests on pure speculation.”
In commercial speech cases the government cannot rest on
“speculation or conjecture.” Congress
didn’t hold pre-enactment hearings on the likely impact of disclosure, and post-enactment
hearings contained testimony both pro and con.
Post hoc evidence suggested that the law may have backfired: “miners are
being put out of work or are seeing even their meager wages substantially
reduced, thus exacerbating the humanitarian crisis and driving them into the
rebels’ camps as a last resort.” Other
sources support the disclosure, but its effectiveness was not “proven to the
degree required under the First Amendment to compel speech.”
[Part of the problem is the failure of the government to
defend an investor’s interest in refusing to participate directly in or benefit
directly from harm-generating activities, even if that refusal does not stop
the harm and only allows the investor to
walk away
from Omelas.
The best explanation of
this interest as a distinct one in legal terms is Douglas Kysar’s
Preferences
for Process: The Process/Product Distinction and the Regulation of Consumer
Choice.
Disclosure, which allows
investors (and potentially consumers) to make this choice to implicate or not
implicate themselves, directly furthers that exact interest.]
That was enough to doom the regulation, but the disclosure
was also not “purely factual and uncontroversial,” as required by Zauderer and AMI. You could read this
phrase as descriptive rather than definitional in Zauderer, but AMI said it
was a separate requirement for upholding the disclosure, and the panel was,
after all, bound by AMI. [OK, now the majority is just acting like a jerk. Brutus is an honorable man and all that.]
“Uncontroversial” must mean something different than “purely
factual.” It has to be controversial for
some reason other than a dispute about factual accuracy. We could understand this as a fact/opinion
divide,
[b]ut that line is often blurred,
and it is far from clear that all opinions are controversial. Is Einstein’s
General Theory of Relativity fact or opinion, and should it be regarded as
controversial? If the government required labels on all internal combustion
engines stating that “USE OF THIS PRODUCT CONTRIBUTES TO GLOBAL WARMING” would
that be fact or opinion? It is easy to convert many statements of opinion into
assertions of fact simply by removing the words “in my opinion” or removing “in
the opinion of many scientists” or removing “in the opinion of many experts.” It is also the case that
propositions once regarded as factual and uncontroversial may turn out to be
something quite different.
A footnote discussed changing scientific opinions on the
contribution of dietary cholesterol to blood cholesterol, and when the
assessment of factual correctness ought to be made, at enactment or at the time
of challenge/controversy. [Though it did
not discuss the extensive body of law that deals with whether starting a factual
statement with “in my opinion” means that the statement is one of opinion and
not fact. Spoiler: no.
So the minor premise is wrong too.
In my opinion.]
Anyway, the AMI en
banc viewed country of origin of disclosures for meat as “uncontroversial,” but
that was puzzling, rather than providing guidance. There was definitely a dispute about those
disclosures, since they were challenged at the WTO. [Again, disingenuous. AMI
didn’t give a great definition of “uncontroversial” by any means, but no one
disputed that meat required to be labeled as having been slaughtered in the US
was in fact slaughtered in the US—unlike the cholesterol example. Those origin labels were the paradigmatic
disclosures that were controversial “for reasons other than dispute over
factual accuracy.” I also note that we’re
not going to hear about biased disclosure regulations surrounding abortion in
this discussion, because abortion’s First Amendment is just different.]
The dissent’s alternative was to read “uncontroversial” as “accurate,”
which made the phrase redundant. “Is
there such a thing as a ‘purely factual’ proposition that is not ‘accurate’? [Well, yes.
“My car is red” is a purely factual proposition. It is not accurate, at least if I said
it.] Accurate information can also be
misleading, anyway, so it’s a bad line.
Nor could the statutory
definition of “conflict free” save
the law, because the government doesn’t get to force companies to use its
preferred language. [FDA, get more
nervous.] As NAM said, “companies could
be compelled to state that their products are not ‘environmentally sustainable’
or ‘fair trade’ if the government provided ‘factual’ definitions of those
slogans – even if the companies vehemently disagreed that their [products] were
‘unsustainable’ or ‘unfair.’” The majority continued:
A famous example of governmental
redefinition comes to mind:
WAR IS PEACE
FREEDOM IS SLAVERY
IGNORANCE IS STRENGTH
George Orwell, Nineteen
Eighty-Four.
[Professor Tushnet is impressed, and wonders where,
rhetorically, there is to go from here.] “Conflict free” is an ideological
statement, since gold doesn’t fight conflicts; the disclosure requires
companies “to tell consumers that its products are ethically tainted, even if
they only indirectly finance armed groups.”
Companies are allowed to disagree with that assessment, even by
remaining silent.
Judge Srinivasan dissented.
There are lots of “garden-variety” disclosure obligations for securities
issuers that no one [but the majority] thinks are a First Amendment
problem. The conflict minerals
disclosure “provides investors and consumers with useful information about the
geographic origins of a product’s source materials”—an interest specifically upheld as time-honored in AMI.
The term “DRC conflict free” is statutorily defined; if the issuer can’t
determine, after investigation, that a product is “DRC conflict free” under the
statutory definition, it must say so in a report disclosing that the product
has “not been found to be ‘DRC conflict free.’”
The requirement to make that
disclosure, in light of the anticipated reaction by investors and consumers,
aims to dissuade manufacturers from purchasing minerals that fund armed groups
in the DRC region. That goal is unique to this securities law; but the basic
mechanism—disclosure of factual information about a product in anticipation of
a consumer reaction—is regular fare for governmental disclosure mandates.
There was no First Amendment objection to the
source-investigation obligation. Nor was
there a challenge to the obligation to list products that fail to qualify as
“DRC conflict free” in a report for investors. They just objected to the
requirement to describe the listed products with the catchphrase “not been
found to be ‘DRC conflict free.’” But the prescribed shorthand phrase couldn’t
materially change the constitutional calculus.
This shorthand “comes amidst a set of mandated disclosures about the
measures undertaken to determine the source of minerals originating in the DRC
or adjoining countries.” So the meaning would be apparent in context, and the
SEC also allowed issuers to elaborate however they wanted, including the
statement that this is “a phrase we are obligated to use under federal
securities laws to describe products when we are unable to determine that they
contain no minerals that directly or indirectly finance or benefit armed groups
in the DRC or an adjoining country.” At that point, there would seem to be
nothing arguably confusing or misleading about the content of the Rule’s
mandated disclosure.
The basic rule is that, “when the government requires
disclosure of truthful, factual information about a product to consumers, a
company’s First Amendment interest in withholding that information from its
consumers is ‘minimal.’” That’s enough
to sustain this rule. Though the
disclosure “invites public scrutiny,” that’s also true of other requirements,
such as required calorie count or nutritional information. Even under Central Hudson, this requirement would survive, given that
commercial speech is valued for different reasons than non-commercial speech—it
helps consumers through providing them information.
Whether Zauderer
or Central Hudson applies depends on
whether a regulation adds information to the flow of truthful commercial
speech, or suppresses some truthful commercial speech. Under that standard, Zauderer obviously applies. The speech at issue is commercial: it
requires manufacturers to disclose information about product composition. The
fact that this disclosure appears on websites and annual reports filed with the
SEC doesn’t change its status as commercial speech; United States v. Philip
Morris USA, Inc., 566 F.3d 1095 (D.C. Cir. 2009) (per curiam), “treated
corrective statements about products required to be included on the company’s
website as commercial speech” in response to Philip Morris’ argument that such
disclosures couldn’t be commercial speech because they were unattached to
ads. Philip
Morris held that commercial speech “include[s] material representations
about the efficacy, safety, and quality of the advertiser’s product, and other
information asserted for the purpose of persuading the public to purchase” (or,
given the corrective disclosures at issue, not to purchase) “the product.”
The newly minted subclassing of Zauderer to only some instances of commercial speech contradicted Zauderer’s core rationale, which is that
First Amendment protection for commercial speech is justified only by its
informational value to consumers. Its
results were silly—“[a]fter all, if faced with the choice between an annual
website report and product packaging, a seller would predictably opt for the
former,” but the majority’s approach made it easier to impose a packaging
disclosure requirement than a website disclosure. As I noted above, this had nothing to do with
AMI, since the new rule applies to
anti-deception disclosures as well. Zauderer “unsurprisingly used the word ‘advertising’
numerous times in the relevant part of the opinion, but only because that was
the particular factual context in which the case arose. For what it’s worth,
the Court also used ‘commercial speech’ and ‘commercial speaker’ a number of
times in the same part of the opinion when explaining the rationale for the
relaxed First Amendment standard it set forth, and it also did so when framing
the question it addressed in that part of its opinion.” Nor did AMI
even stop to address whether “labels” were more like “advertising” than
like “non-advertising commercial speech,” because Zauderer applies to commercial speech. Hurley
isn’t a commercial speech case, and United
Foods merely described Zauderer’s
outcome.
Under Zauderer,
this disclosure was purely factual and uncontroversial—a standard that must be
assessed in light of Zauderer’s
rationale, which is the value of commercial speech in providing consumers with
useful information about products and services. That value is supported by the
disclosure of purely factual and accurate information; thus Zauderer requires that the factual
disclosure must be non-deceptive, and cannot prescribe “what shall be orthodox
in politics, nationalism, religion, or other matters of opinion.” The disclosure must be uncontroversially
factual: there could be no “disagree[ment] with the truth of the facts required
to be disclosed.” “[E]ven if the disclosure qualifies as ‘purely factual,’ it
would still fall outside of Zauderer
review if the accuracy of the particular information disclosed were subject to
dispute.” The meaning of “uncontroversial”
should be tethered to the core question of whether the disclosure is “factual.”
Were it not so, AMI should have come
out the other way, as the panel majority recognized.
Under those principles, the requirement to identify whether
a product has “been found to be ‘DRC conflict free’” calls for disclosure of
“purely factual and uncontroversial” information, because “DRC conflict free”
is a defined term of art. It’s not
misleading, especially in its context, which is a description of the
manufacturer’s attempts to identify the source of the minerals it uses. The SEC, for example, approved this language:
Because we cannot determine the
origins of the minerals, we are not able to state that products containing such
minerals do not contain conflict minerals that directly or indirectly finance
or benefit armed groups in the Democratic Republic of the Congo or an adjoining
country. Therefore, under the federal
securities laws we must describe the products containing such minerals as
having not been found to be ‘DRC conflict free.’ Those products are listed
below.
That’s not a confession of an ethical taint. The fact that the issuer would prefer not to
say anything doesn’t distinguish this from many other disclosures, like calorie
counts, nutritional information, and disclosures about the presence of mercury.
“Such disclosures of course can elicit a reaction by consumers—that is often
the point, as with the country-of-origin rule upheld in AMI—but the disclosures still remain factual and truthful.”
Under this rule, the government can’t misleadingly redefine “peace”
to mean “war”—a consumer would have no reason to suppose that this redefinition
had occurred. Likewise, statements of
opinion such as “this product is environmentally unsustainable” are outside Zauderer, as compared to “this product releases
x units of ozone in y hours,” a pure fact.
There could be difficult questions at the margin, but that’s
standard. Also, “constitutional
protections outside of the First Amendment might constrain the government’s
ability to compel disclosures—for instance, if the disclosures facilitated
private discrimination. See Palmore v. Sidoti, 466 U.S. 429 (1984).” But that didn’t matter here.
The dissent would also have found that the disclosures
survived Central Hudson. The government’s interest isn’t just to
promote peace in the DRC. It’s to do so “by
reducing funding to armed groups in the DRC region from trade in conflict
minerals.” And, like country of origin labeling, the disclosure rule “operates
on the basis of assumptions about the reaction of investors to disclosures
about a product’s place of origin.” This
is a substantial government interest.
The disclosure directly advances that interest. AMI
held that “evidentiary parsing is hardly necessary when the government uses a
disclosure mandate to achieve a goal of informing consumers about a particular
product trait.” The requirement of due
diligence on product supply chains plus disclosure of the results of that due
diligence “encourages manufacturers voluntarily to reduce their reliance on
conflict minerals from the DRC and adjoining countries,” and disclosure further
enables consumers and investors to exert pressure on manufacturers to minimize
the use of conflict minerals from the DRC region. This was a sufficiently reasonable fit between
means and ends.
Moreover, deference to the political branches’ predictive
judgment was more warranted in the arena of foreign affairs. Nor did the cost of implementation affect the reasonability
of the means chosen. Recall that the
production audit doesn’t raise First Amendment questions; once that’s done,
obligating issuers to use a shorthand phrase and put it on their website/in SEC
reports isn’t unduly burdensome.
[Yes! Food manufacturers don’t
get to include the cost of determining the calorie count of a food in the costs
of disclosure—at least not if we don’t want the First Amendment to become
super-Lochner.]
Even if there were uncertainty about Congress’s predictive
judgments about the effect of the disclosure on the conflict in the DRC, the
court should defer to the political branches’ assessments, and Congress
determined that trade in conflict minerals was helping to finance conflict. In Holder
v. Humanitarian Law Project, 561 U.S. 1 (2010), the Court deferred to the
political branches’ foreign policy judgments even under strict scrutiny; the
more so here. Plus, constitutionality should not turn on a post hoc referendum
on a law’s effectiveness at a particular point in time. “Otherwise, a law’s
constitutionality might wax and wane depending on the precise time when its
validity is assessed.” The relevant
question is whether, at enactment, the disclosure regime was reasonably
designed to reduce the funding of armed groups in the DRC. [This is different
from assessing the truthfulness of
the disclosure over time, which is the cholesterol example.]
Moreover, the rule was having its desired effect even if its
larger effects were uncertain: companies in the US were now avoiding
DRC-sourced minerals, which was the direct aim. Unintended ripple effects
shouldn’t invalidate the law; those should be for the political branches to
judge.
