dissenting in part:
The majority interprets the sentences in Berkeley’s forced disclosure statement one at a time and holds that each is “literally true.” But consumers would not read those sentences in isolation the way the majority does. Taken as a whole, the most natural reading of the disclosure warns that carrying a cell phone in one’s pocket is unsafe. Yet Berkeley has not attempted to argue, let alone to prove, that message is true.
It is clear that the First Amendment prevents the government from requiring businesses to make false or misleading statements about their own products. See Video Software Dealers Ass’n v. Schwarzenegger, 556 F.3d 950, 967 (9th Cir. 2009), aff'd sub nom. Brown v. Entm’t Merchs. Ass’n, 564 U.S. 786, 131 S.Ct. 2729, 180 L.Ed.2d 708 (2011). Because — at least on the current record — that is what Berkeley’s ordinance would do, I believe the ordinance likely violates the First Amendment and therefore should have been pre*1125liminarily enjoined.1 See Klein v. City of San Clemente, 584 F.3d 1196, 1207-08 (9th Cir. 2009) (“Both this court and the Supreme Court have repeatedly held that ‘[t]he loss of First Amendment freedoms, for even minimal periods of time, unquestionably constitutes irreparable injury.’ ” (quoting Elrod v. Burns, 427 U.S. 347, 373, 96 S.Ct. 2673, 49 L.Ed.2d 547 (1976))).
I
Berkeley’s ordinance requires stores selling cell phones, to provide a disclosure stating:
To assure safety, the Federal Government requires that cell phones meet radio-frequency (RF) exposure guidelines. If you carry or use your phone in a pants or shirt pocket or tucked into a bra when the phone is ON and connected to a wireless network, you may exceed the federal guidelines for exposure to RF radiation. Refer to the instructions in your phone or use manual for information about how to use your phone safely.
Berkeley Mun. Code § 9.96.030(A) (2015).
The majority parses these sentences individually and concludes that each is “literally true.” In my view, this approach misses the forest for the trees. On its face, the disclosure begins and ends with references to safety, plainly conveying that the intervening language describes something unsafe. Indeed, the disclosure directs consumers to their user manuals for instructions on “how to use your phone safely.” The message of the disclosure as a whole is clear: carrying a phone “in a pants or shirt pocket or tucked into a bra” is not safe. Yet that implication is a problem for Berkeley because it has not offered any evidence that carrying a cell phone in a pocket is in fact unsafe. Instead, it has expressly denied that the required disclosure conveys that message. I disagree.
Berkeley insists the ordinance “rests exclusively upon existing FCC regulations.” But those regulations communicate something far different than does the ordinance. The FCC guidelines make clear that they are designed to incorporate a many-fold safety factor, such that exposure to radiation in excess of the guideline level is considered by the FCC to be safe:
Our current RF exposure guidelines ... include[e] a significant “safety” factor, whereby the exposure limits are set at a level on the order of 50 times below the level at which adverse biological effects have been observed in laboratory animals as a result of tissue heating resulting from RF exposure. This “safety” factor can well accommodate a variety of variables such as different physical characteristics and individual sensitivities— and even the potential for exposures to occur in excess of our limits without posing a health hazard to humans.
In re Reassessment of FCC Radiofrequency Exposure Limits and Policies, 28 FCC Red. 3498, 3582 (Mar. 29, 2013) (emphasis added). There is thus no evidence in the record that the message conveyed by the ordinance is true.2
*1126II
The First Amendment clearly does not permit the government to force businesses to make false or misleading statements about their products. In Video Software Dealers, we considered a challenge to a California law requiring that “violent” video games be labeled with a sticker that said “18” and preventing the sale or rental of violent video games to minors. 556 F.3d at 953-54. After striking down the law’s sale and rental prohibition, we concluded that continuing to require the label “18” “would arguably ... convey a false statement” that minors could not buy or rent the video game, and was therefore unconstitutional. Id. at 965-67. The same principle applies here: the First Amendment prohibits Berkeley from compelling retailers to communicate a misleading message. I would thus hold that CTIA is likely to succeed on the merits of its First Amendment challenge.
There are downsides to false, misleading, or unsubstantiated product warnings. Psychological and other social science research suggests that overuse may cause people to pay less attention to warnings generally: “[A]s the number of warnings grows and the prevalence of warnings about low level risks increases, people will increasingly ignore or disregard them.” J. Paul Frantz et al., Potential Problems Associated with Overusing Warnings, Proceedings of the Human Factors & Ergonomics Soc’y 43rd Ann. Meeting 916, 916 (1999). Relatedly, “[wjarnings about very minor risks or risks that are extremely remote have raised concerns about negative effects on the believability and credibility of warnings.... In essence, such warnings represent apparent false alarms as they appear to be ‘crying wolf.’ ” Id. at 918; see also David W. Stewart & Ingrid M. Martin, Intended and Unintended Consequences of Warning Messages: A Review and Synthesis of Empirical Research, 13 J. Pub. Pol’y & Marketing 1, 7 (1994). If Berkeley wants consumers to listen to its warnings, it should stay quiet until it is prepared to present evidence of a wolf.
. I agree with the majority’s preemption analysis so dissent only from sections IV.A. 1., IV.B., IV.C., and IV.D. of the majority opinion.
. Because even under Zauderer v. Office of Disciplinary Counsel of Supreme Court of Ohio, 471 U.S. 626, 105 S.Ct. 2265, 85 L.Ed.2d 652 (1985), any forced disclosure statement must be truthful, see id. at 651, 105 S.Ct. 2265, I do not think that any discussion of Zauderer is appropriate in this case. If nevertheless I were to consider the extent of Zauderer’s applicability, as the majority does, I would be inclined to conclude that Zauderer applies only when the government compels a truthful disclosure to counter a false or misleading advertisement. Given that the disclosure in Zauderer itself prevented an advertisement from being misleading, I have serious doubt that the Supreme Court intended the Zauderer test to apply in broader circumstances. See id. ("[W]e hold that an advertiser's rights are adequately protected as long as *1126disclosure requirements are reasonably related to the State’s interest in preventing deception of consumers.”). The majority’s contrary conclusion also seems to me to be in tension with our decision in Video Software Dealers, which treated Zauderer as applying only in the context of disclosures aimed at combatting otherwise misleading advertising. See 556 F.3d at 967 ("[T]he labeling requirement fails Zauderer's rational relationship test, which asks if the 'disclosure requirements are reasonably related to the State’s interest in preventing deception of customers.’ ” (quoting Zauderer, 471 U.S. at 651, 105 S.Ct. 2265)).