OPINION
ANDERSON, G. BARRY, Justice.
The question presented here is whether a Minnesota statute that prohibits knowingly false reports of police misconduct violates the First Amendment because it allows the State to punish some people, but not others, depending on the viewpoint expressed about the police. A jury found Melissa Jean Crawley guilty of violating the challenged law, Minn.Stat. § 609.505, subd. 2 (2010), based on the fact that she informed a police officer that another officer forged her signature, knowing that the information conveyed was false. The court of appeals reversed her conviction after concluding that section 609.505, subdivision 2, is unconstitutional because it criminalizes false speech “critical” of the police but not false speech that favors the police. State v. Crawley, 789 N.W.2d 899, 910 (Minn.App.2010). Because we narrowly construe section 609.505, subdivision 2, to criminalize only defamatory speech not protected by the First Amendment, and because the statute falls within two of the exceptions to the constitutional prohibition against content discrimination in an unpro[98]*98tected category of speech, we reverse the court of appeals’ judgment that the statute is unconstitutional. Because Crawley’s conviction under section 609.505, subdivision 2, preceded our narrow construction of the statute, due process considerations entitle her to a new trial. We therefore reverse her conviction and remand for a new trial based on our narrowing construction of the statute.
On April 17, 2008, Melissa Jean Crawley went to the Winona County Law Enforcement Center, met with Winona Police Department Sergeant Christopher Nelson, and informed Nelson that a police officer had forged her signature on a medical release form at a Winona hospital. Nelson asked Crawley who she thought had forged her signature. Crawley noted that the form was signed “Melissa Crawley at 0600 hours,” and informed Nelson that “it has to be a police officer that did that. I don’t sign things and date them 0600 hours.” The release related to treatment Crawley received at the hospital for injuries sustained in an assault that Winona police were investigating, and Crawley informed Nelson she thought her signature was forged by “the police officer who requested the records, whoever was doing the investigation.”
Nelson investigated Crawley’s report. During his investigation, Nelson spoke to a nurse who told Nelson that she saw Craw-ley sign the release while Crawley was at the hospital. The State charged Crawley on April 30, 2008, with falsely reporting an act of police misconduct, Minn.Stat. § 609.505, subd. 2(a)(2), and falsely reporting a crime, Minn.Stat. § 609.505, subd. 1 (2010).1 Subdivision 2, the statutory provision at issue in this appeal, provides:
(a) Whoever informs, or causes information to be communicated to, a peace officer, whose responsibilities include investigating or reporting police misconduct, that a peace officer, as defined in section 626.84, subdivision 1, paragraph (c), has committed an act of police misconduct, knowing that the information is false, is guilty of a crime and may be sentenced as follows:
(1) up to the maximum provided for a misdemeanor if the false information does not allege a criminal act; or
(2) up to the maximum provided for a gross misdemeanor if the false information alleges a criminal act.
(b) The court shall order any person convicted of a violation of this subdivision to make full restitution of all reasonable expenses incurred in the investigation of the false allegation unless the court makes a specific written finding that restitution would be inappropriate under the circumstances. A restitution award may not exceed $8,000.
Crawley moved to dismiss the charge under subdivision 2(a)(2). In her motion to dismiss, Crawley relied wholly on Choker v. Crogan, 428 F.3d 1215 (9th Cir.2005). In Choker, the United States Court of Appeals for the Ninth Circuit held that a California statute, Cal. Penal. Code § 148.6, criminalizing knowingly false reports of police misconduct violated the First Amendment for targeting “only knowingly false speech critical of peace officer conduct during the course of a com[99]*99plaint investigation,” but not “[k]nowingly false speech supportive of peace officer conduct.” 428 F.3d at 1228. The Winona County District Court denied the motion to dismiss and, after a trial, a jury found Crawley guilty of both counts. The district court concluded subdivision 1 was a lesser included offense of subdivision 2(a)(2), convicted Crawley of subdivision 2(a)(2), and sentenced her to 15 days in jail.
Crawley appealed her conviction to the Minnesota Court of Appeals. In a divided decision, the court of appeals reversed Crawley’s conviction and remanded the case for sentencing on the subdivision 1 verdict. Crawley, 789 N.W.2d at 910. The majority of the court categorized the speech at issue as an “intentional lie” that was not protected by the First Amendment. Id. at 90S (emphasis omitted). The majority stated that “even though intentional falsehoods are subject to regulation, the government cannot pick and choose which falsehoods to prohibit so as to criminalize certain false statements but not others based on the content of the speech or viewpoint of the speaker.” Id. at 904. The majority concluded that subdivision 2 violates the First Amendment because it criminalizes “false critical information” but not “false exonerating information,” contrary to the prohibition on viewpoint discrimination announced in R.A.V. v. City of St. Paul, 505 U.S. 377, 395, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992). Crawley, 789 N.W.2d at 910. The court of appeals dissent, on the other hand, concluded that the statute targeted speech within the unprotected category of defamation, not simple “lies.” Id. (Harten, J., dissenting) (emphasis omitted). The dissent reasoned that, when viewed as a regulation of defamation, section 609.505, subdivision 2, comes within the exceptions to the R.A.V. prohibition on content discrimination. Crawley, 789 N.W.2d at 911-12 (Harten J., dissenting). The State sought further review,2 which we granted.3
[100]*100We begin by noting certain principles of First Amendment law that will frame our discussion of this case. Content-based restrictions of speech4 are presumptively invalid, R.A.V., 505 U.S. at 382, 112 S.Ct. 2538, and ordinarily subject to strict scrutiny, United States v. Playboy Entm’t Grp., Inc., 529 U.S. 803, 813, 120 S.Ct. 1878, 146 L.Ed.2d 865 (2000). But the Supreme Court “ha[s] long recognized that the government may regulate certain categories of expression consistent with the Constitution.” Virginia v. Black, 538 U.S. 343, 358, 123 S.Ct. 1536, 155 L.Ed.2d 535 (2003). As explained recently by the Supreme Court in United States v. Stevens:
From 1791 to the present, ... the First Amendment has permitted restrictions upon the content of speech in a few limited areas, and has never included a freedom to disregard these traditional limitations. These historic and traditional categoi’ies long familiar to the bar ... include[d] obscenity, defamation, fraud, incitement, and speech integral to criminal conduct....
- U.S. -, 130 S.Ct. 1577, 1584, 176 L.Ed.2d 435 (2010) (citations omitted) (internal quotation marks and alterations omitted). Categories of speech such as obscenity and defamation that may be restricted without violating the First Amendment are often called “unprotected speech,” R.A.V., 505 U.S. at 406, 112 S.Ct. 2538, and can, “consistently with the First Amendment, be regulated because of their constitutionally proscribable content.” Id. at 383, 112 S.Ct. 2538 (emphasis omitted). But these unprotected categories of speech are not “entirely invisible to the Constitution, so that they may be made the vehicles for content discrimination unrelated to their distinctively proscribable content.” Id. at 383-84, 112 S.Ct. 2538. Thus, the government is prohibited from discriminating on the basis of content within unprotected categories of speech unless one of the exceptions set forth in R.A.V. apply. Id. at 388-90, 112 S.Ct. 2538
Following this framework, in Part I of this opinion, we conclude that section 609.505, subdivision 2, is a content-based regulation of speech. In Part II, we construe section 609.505, subdivision 2, narrowly and conclude that the statute punishes only speech that meets the Minnesota definition of defamation, an unprotected category of speech. In Part III, we evaluate the statute under the constitutional rule that prohibits the State from drawing distinctions based on content within an unprotected category of speech. Finally, in Part IV, we conclude that under our narrowing construction, section 609.505, subdivision 2, is constitutional.
[101]*101I.
The constitutionality of a statute presents a question of law, which we review de novo. State v. Cox, 798 N.W.2d 517, 519 (Minn.2011); State v. Melde, 725 N.W.2d 99, 102 (Minn.2006). In this case, the court of appeals held that section 609.505, subdivision 2, criminalizes the “intentional lie.” State v. Crawley, 789 N.W.2d 899, 903 (Minn.App.2010) (emphasis omitted). The court of appeals determined that the intentional lie is one type of expression that is subject to regulation, since it “fails to ‘materially advance[ ] society’s interest in uninhibited, robust, and wide-open debate on public issues.’ ” Id. (quoting Gertz v. Robert Welch, Inc., 418 U.S. 323, 340, 94 S.Ct. 2997, 41 L.Ed.2d 789 (1974)). Nevertheless, the court of appeals concluded that the statute is facially unconstitutional because it exposes some speakers but not others to criminal sanction based on the speaker’s expressed viewpoint regarding the police. Id. at 905 (“The provision challenged in this case punishes only those known falsehoods that are critical of police conduct.”). Crawley asks us to affirm, arguing that the statute impermissibly discriminates against “a certain class of anti-government speech” while permitting an otherwise “similarly situated class of pro-government speech” to go unpunished. To address Crawley’s argument, we must first determine whether section 609.505 is a content-based regulation of speech.
Section 609.505, subdivision 2, criminalizes' knowingly false reports of police misconduct. Neither party disputes that the statute regulates speech. But while Craw-ley contends that the statute impermissi-bly discriminates on the basis of content, and even viewpoint,5 the State argues that the statute is a content-neutral time, place, and manner restriction. We conclude that section 609.505, subdivision 2, is a content-based regulation of speech because whether a person may be prosecuted under the statute depends entirely on what the person says. See Holder v. Humanitarian Law Project, — U.S. -, 130 S.Ct. 2705, 2712, 2723-24, 177 L.Ed.2d 355 (2010) (concluding a law that made it a crime to “knowingly providfe] material support or resources to a foreign terrorist organization” regulated speech based on its content because whether plaintiffs could speak without sanction “depends on what they say”).
II.
Because section 609.505, subdivision 2, is a content-based regulation of speech, we must determine if the statute, as written, criminalizes only unprotected speech.6 See United States v. Stevens, — [102]*102U.S. -, 130 S.Ct. 1577, 1584, 176 L.Ed.2d 435 (2010) (traditional categories of unprotected speech “long familiar to the bar” include obscenity, defamation, fraud, incitement, and speech integral to criminal conduct). If the statute, as written, criminalizes a substantial amount of protected speech in addition to unprotected speech, we must then determine if we can uphold the statute’s constitutionality by construing it narrowly to reach only unprotected speech.
To be a constitutional exercise of the police power of á state, a statute that punishes speech must not be overly broad.7 See Broadrick v. Oklahoma, 413 U.S. 601, 612, 93 S.Ct. 2908, 37 L.Ed.2d 830 (1973). In general, a statute can be said to be overly broad if it prohibits or chills a substantial amount of protected speech along with unprotected speech. See Ashcroft v. Free Speech Coal. 535 U.S. 234, 244, 122 S.Ct. 1389, 152 L.Ed.2d 403 (2002). But when possible, we uphold a law’s constitutionality by narrowly construing the law so as to limit its scope to conduct that falls outside First Amendment protection while clearly prohibiting its application to constitutionally protected expression. See In re Welfare of S.L.J., 263 N.W.2d 412, 419 (Minn.1978) (disorderly conduct statute limited to “fighting words” to preserve constitutionality); see also New York v. Ferher, 458 U.S. 747, 769, n. 24, 102 S.Ct. 3348, 73 L.Ed.2d 1113 (1982) (“If the invalid reach of the law is cured [by narrow judicial construction], there is no longer reason for proscribing the statute’s application to unprotected conduct.”).
We construe statutes de novo. Krummenacher v. City of Minnetonka, 783 N.W.2d 721, 726 (Minn.2010). Our primary objective is to ascertain and give effect to the legislature’s intent. Minn. Stat. § 645.16 (2010). We follow the directive that “[w]hen the words of a law in their application to an existing situation are clear and free from all ambiguity, the letter of the law shall not be disregarded under the pretext of pursuing the spirit.” Id. Ambiguity exists only where statutory language is subject to more than one reasonable interpretation. In re Welfare of J.B., 782 N.W.2d 535, 540 (Minn.2010).
A.
We begin with the language of the statute. Section 609.505, subdivision 2, provides in relevant part:
[103]*103(a) Whoever informs, or causes information to be communicated to, a peace officer, whose responsibilities include investigating or reporting police misconduct, that a peace officer, as defined in section 626.84, subdivision 1, paragraph (c), has committed an act of police misconduct, knowing that the information is false, is guilty of a crime and may be sentenced as follows:
(1) up to the maximum provided for a misdemeanor if the false information does not allege a criminal act; or
(2) up to the maximum provided for a gross misdemeanor if the false information alleges a criminal act.
Under the statute, in order for a person to be subject to criminal sanctions, he or she must “know[ ]” that the information he or she communicates to a peace officer is false. And the statute specifies the “information” that is actionable: “that a peace officer ... has committed an act of police misconduct.”
“[A]ct of police misconduct” has a clear, technical meaning when we turn to other statutes on the same subject, and to applicable rules. See Minn.Stat. § 645.08(1) (2010) (requiring that, when Minnesota statutes are construed, “technical words and phrases and such others as have acquired a special meaning ... are construed according to such special meaning or their definition”). First, Minnesota Rules 6700.2000-.2600 (2011) create procedures and guidelines for the investigation, processing, and resolution of misconduct allegations against licensed peace officers. The rules define “[mjisconduct” as “an act or omission by an employee or appointee of an agency licensed by the board which may result in disciplinary action by the agency or appointing authority.” Minn. R. 6700.2000, subp. 8. The model state professional conduct policy, developed by the state Board of Police Officer Standards and Training pursuant to Minn.Stat. § 626.8457, subd. 1, 2, (2010), outlines 35 rules of conduct, including a prohibition against committing a crime while on or off duty. See Board of Police Officer Standards and Training, Professional Conduct of Peace Officers Model Policy (2011). Considering these provisions, we construe “act of police misconduct” in Minn.Stat. § 609.505, subd. 2, to mean a specific act or omission that violates a policy or rule of professional conduct, adopted by a law enforcement agency, which would expose a peace officer to discipline.
The use of the identical phrase “a peace officer” in section 609.505, subdivision 2, can apply to situations in which the officer who is the subject of the false report is the same officer who is informed about the report. The statute can also apply to situations where the officer who is the subject of the false report is not the same officer who is informed about the report. A peace officer is identified generally by the reference to the statutory definition of “peace officer” at Minn.Stat. § 626.84, subd. l(c)(l)(2010).8 Although the two instances of the phrase “a peace officer” are followed by different descriptive clauses— “whose responsibilities include investigat[104]*104ing or reporting police misconduct” and “has committed an act of police misconduct” — there is no requirement that a peace officer being informed of the misconduct must be a different person than the peace officer who is accused of committing misconduct in order for the speech to be criminal.
We thus conclude that section 609.505, subdivision 2, as written, makes it a crime for a person to inform a peace officer, whose responsibilities include investigating or reporting police misconduct, that a peace officer, who may or may not be the same peace officer being so informed, committed an act of police misconduct, knowing that the information conveyed is false.
B.
It is clear that, as written, section 609.505, subdivision 2, is overly broad because it punishes a substantial amount of protected speech in addition to unprotected speech.9 As explained below, while the statute, as written, criminalizes defamatory speech, which is unprotected under the First Amendment, it also criminalizes a substantial amount of speech that is not defamatory and thus protected speech.
To establish a defamation claim in Minnesota, a plaintiff must prove four elements: he or she must show that the defamatory statement is “communicated to someone other than the plaintiff;” that “the statement is false;” that the statement tends to “harm the plaintiffs reputation” and to lower the plaintiff “in the estimation of the community,” Bahr v. Boise Cascade Corp., 766 N.W.2d 910, 919-20 (Minn.2009); and that the recipient of the false statement reasonably understands it to refer to a specific individual, see Glenn v. Daddy Rocks, Inc., 171 F.Supp.2d 943, 948 (D.Minn.2001) (stating, under Minnesota law, that the “of and concerning” element requires the plaintiff in a defamation lawsuit to prove that the statement at issue either explicitly referred to the plaintiff or that a reader “by fair implication” would understand that the statement referred to the plaintiff). “If the defamation affects the plaintiff in his business, trade, profession, office or calling, it is defamation per se and thus actionable without any proof of actual damages.” Bahr, 766 N.W.2d at 920 (citation omitted) (internal quotation marks and alteration omitted).
Because the statute, as written, does not require knowingly false accusations of police misconduct to be communicated to someone other than the plaintiff in order for the speech to be criminal, the statute fails to fulfill the first element of defamation: publication to a third person. The statute also fails to fulfill the fourth element because it does not require the statement to be “of and concerning” a specific individual. Because the statute does not [105]*105satisfy all of the elements of defamation, it punishes a substantial amount of protected speech and is therefore facially unconstitutional.
The United States Supreme Court has recognized that the overbreadth doctrine as “strong medicine” that has been employed “sparingly.” Broadrick, 413 U.S. at 613, 98 S.Ct. 2908. “Because of the wide-reaching effects of striking down a statute on its face,” the Supreme Court has employed the overbreadth doctrine “with hesitation, and then ‘only as a last resort.’ ” Ferber, 458 U.S. at 769, 102 S.Ct. 3348 (quoting Broadrick, 413 U.S. at 613, 93 S.Ct. 2908). “Facial overbreadth has not been invoked when a limiting construction has been or could be placed on the challenged statute.” Broadrick, 413 U.S. at 613, 93 S.Ct. 2908. In In re Welfare of R.A.V., we stated
[T]he complete invalidation of legislatively adopted laws [the overbreadth doctrine] permits is “strong medicine” that this court does not hastily prescribe. Where the overbreadth of the challenged law is both “real” and “substantial,” and where “the words of the [law] simply leave no room for a narrowing construction,” “so that in all its applications the [law] creates an unnecessary risk of chilling free speech,” this court will completely invalidate it. When possible, however, this court narrowly construes a law subject to facial overbreadth attack so as to limit its scope to conduct that falls outside first amendment protection while clearly prohibiting its application to constitutionally protected expression.
464 N.W.2d 507, 509 (Minn.1991) (citations omitted) (rev’d sub nom. R.A.V. v. City of St. Paul, 505 U.S. 377, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992)).
Although the statute does not satisfy all of the elements of defamation, we can uphold its constitutionality by construing it narrowly to refer only to defamation. The United States Supreme Court generally allows, and even encourages, state supreme courts to sustain the constitutionality of state statutes regulating speech by construing them narrowly to punish only unprotected speech. For example, in Chaplinsky, the Supreme Court upheld a New Hampshire statute as constitutional because the highest court of New Hampshire authoritatively construed the statute to reach only “fighting words.” 315 U.S. at 573, 62 S.Ct. 766. The statute at issue in Chaplinsky provided: “No person shall address any offensive, derisive or annoying word to any other person who is lawfully in any street or other public place, nor call him by any offensive or derisive name....” Id. at 569, 62 S.Ct. 766 (internal quotation marks omitted). The state court limited the statute’s reach to “fighting words” by holding that the statute only prohibited words that “have a direct tendency to cause acts of violence by the persons to whom, individually, the remark is addressed.” Id. at 573, 62 S.Ct. 766 (citation omitted). In upholding the constitutionality of the statute, the Supreme Court stated:
We are unable to say that the limited scope of the statute as thus construed contravenes the Constitutional right of free expression. It is a statute narrowly drawn and limited to define and punish specific conduct lying within the domain of state power, the use in a public place of words likely to cause a breach of the peace.
Id.
The Supreme Court has also deferred to our authoritative construction of statutes regulating speech. In S.L.J., we examined the constitutionality of the disorderly conduct statute, Minn.Stat. § 609.72, subd. [106]*1061(3) (1976).10 263 N.W.2d 412. We stated that, as written, section 609.72, subdivision 1(3), was both “overly broad and vague” because it did not satisfy the definition of “fighting words”:
Since the statute punishes words alone — “offensive, obscene, or abusive language” — it must be declared unconstitutional as a violation of the First and Fourteenth Amendments unless it only proscribes the use of “fighting words.” Section 609.72, subd. 1(3), however, punishes words that merely tend to “arouse alarm, anger, or resentment in others” rather than only words “which by their very utterance inflict injury or tend to incite an immediate breach of the peace.” Since the statute does not satisfy the definition of “fighting words,” it is unconstitutional on its face.
S.L.J., 263 N.W.2d at 418-19. But we went on to hold that although section 609.72, subdivision 1(3) “clearly contemplates punishment for speech that is protected under the First and Fourteenth Amendments, we can uphold its constitutionality by construing it narrowly to refer only to ‘fighting words.’ ” S.L.J., 263 N.W.2d at 419. We noted that the Supreme Court has, in fact, “encouraged state supreme courts to sustain the constitutionality of their offensive-speech statutes by construing them narrowly” to punish only unprotected speech. Id. at 419 n.5 (collecting cases).
Then in In re R.A.V., 464 N.W.2d at 510, we relied on our previous construction in S.L.J. to limit the application of the statute at issue to “fighting words.” In In re R.A.V., prior to Supreme Court review, we considered a St. Paul ordinance that made it a misdemeanor to display a symbol, “including but not limited to, a burning cross or Nazi swastika, which one knows or has reasonable grounds to know arouses anger, alarm, or resentment in others on the basis of race, color, creed, religion, or gender.” 464 N.W.2d at 508 (citation omitted). Applying our previous construction in S.L.J., we limited the phrase “arouses anger, alarm, or resentment in others” to punish only unprotected “fighting words,” defined as “conduct that itself inflicts injury or tends to incite immediate violence.” In re R.A.V., 464 N.W.2d at 510. Although as written, the statute did not satisfy the elements of “fighting words,” we narrowly construed it to refer only to “fighting words.” Id. On appeal, the Supreme Court deferred to our authoritative construction of the ordinance: “In construing the St. Paul ordinance, we are bound by the construction given to it by the Minnesota court. Accordingly, we accept the Minnesota Supreme Court’s authoritative statement that the ordinance reaches only those expressions that constitute fighting words, within the meaning of Chaplinsky.” R.A.V., 505 U.S. at 381, 112 S.Ct. 2538 (internal citations omitted).11
[107]*107Based on prior decisions, including Chaplinsky, S.L.J., and R.A.V., we uphold the constitutionality of section 609.505, subdivision 2, by narrowly construing it to punish only “defamation.”12 Accordingly, we hold that to subject a person to criminal sanctions under section 609.505, subdivision 2, the State must prove that the person informed a police officer, whose responsibilities include investigating or reporting police misconduct, that another officer has committed an act of police misconduct, knowing that the information is false. In addition, in order to satisfy the “of and concerning” element of defamation, the State must prove that the officer receiving the information reasonably understands the information to refer to a specific individual.13
Under our narrowing construction, we conclude that the only speech reached by section 609.505, subdivision 2, is defamation.14 Because under our limit[108]*108ing construction we require the State to prove that a person, in order to be convicted under the statute, has informed a peace officer of an act of police misconduct by another officer, the first element of defamation — communication to a third party— is fulfilled. The statute also requires the communicator of the information to know that it is false, fulfilling the second element. Because an act of misconduct is an allegation that affects a peace officer “in his business, trade, profession, office or calling” the requirement for defamation per se is satisfied. See Bahr, 766 N.W.2d at 920 (citation omitted) (internal quotation marks omitted). Finally, requiring the State to prove that the officer receiving the information reasonably understands the information to refer to a specific individual satisfies the fourth element.
Moreover, the mental state required for a conviction under section 609.505, subdivision 2, exceeds the actual malice standard for defamation of a public official established by the Supreme Court. See New York Times Co. v. Sullivan, 376 U.S. 254, 279-80, 84 S.Ct. 710, 11 L.Ed.2d 686 (1964) (concluding that a public official may not recover damages for “a defamatory falsehood” absent proof that the statement was made with knowledge it was false, or with reckless disregard of whether the statement was true or false); Garrison v. Louisiana, 379 U.S. 64, 77-78, 85 S.Ct. 209, 13 L.Ed.2d 125 (1964) (applying New York Times standard to a criminal defamation statute). Under New York Times and Garrison, a person is exposed to liability for making a statement that he or she knew to be false, or for making a statement with reckless disregard for its truth. Under section 609.505, subdivision 2, the State must prove that a person knew the allegation that a peace officer committed an act of misconduct was false. Thus, section 609.505, subdivision 2, reaches only speech that is defamatory.15
C.
Crawley was convicted under section 609.505, subdivision 2, before our narrowing construction of the statute. She is therefore entitled to have a jury determine whether her statements to Winona Police Department Sergeant Christopher Nelson were “of and concerning” another peace officer. See State v. Vance, 734 N.W.2d 650, 657 (Minn.2007) (“[D]ue process ... ‘entitle[s] a criminal defendant to a jury determination that [she] is guilty of every element of the crime with which [s]he is charged, beyond a reasonable doubt.’ ”) (quoting Apprendi v. New Jersey, 530 U.S. 466, 476-77, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000)). We, therefore, reverse Craw-[109]*109ley’s conviction and remand for a new trial based on our narrowing construction of section 609.505, subdivision 2.
III.
Section 609.505, subdivision 2, does not pass constitutional muster based solely upon our construction that narrows the statute to defamation. Rather, we must evaluate the statute under Supreme Court precedent laying out the constitutional prohibition on content discrimination within unprotected categories of speech, and we turn now to that analysis.
Defamation and other categories of speech that may be restricted without violating the First Amendment are often referred to as “not within the area of constitutionally protected speech.” R.A.V. v. City of St. Paul, 505 U.S. 377, 383, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992) (quoting Roth v. United States, 354 U.S. 476, 483, 77 S.Ct. 1304, 1 L.Ed.2d 1498 (1957)). These unprotected areas of speech can, “consistently within the First Amendment, be regulated because of their constitutionally proscribable content.” Id. But these unprotected categories of speech are not “invisible to the Constitution, so that they may be made the vehicles for content discrimination unrelated to their distinctively proscribable content.” Id. at 383-84, 112 S.Ct. 2538. Thus, the government is prohibited from discriminating on the basis of content within unprotected categories of speech unless one of the exceptions set forth in R.A.V. apply. Id. at 384, 388-90, 112 S.Ct. 2538.
Crawley argues that subdivision 2 is viewpoint-based, in addition to being content-based, because it impermissibly discriminates against “a certain class of anti-government speech” while permitting an otherwise “similarly situated class of pro-government speech” to go unpunished. We disagree and conclude that subdivision 2 is not viewpoint-based. Speech that is supportive of peace officer conduct fails to satisfy the elements of defamation because it does not tend to harm the plaintiffs reputation and to lower the plaintiff in the estimation of the community. If the statute were to reach pro-government speech as well as defamation, the statute would punish a substantial amount of protected speech and be facially unconstitutional as overly broad. Because speech that is supportive of peace officer conduct does not fall within the unprotected category of defamation, the statute does not discriminate on the basis of viewpoint.
We begin with the general rule of R.A.V., which is that content-based distinctions drawn within unprotected categories of speech are unconstitutional. 505 U.S. at 382, 112 S.Ct. 2538. But these content-based distinctions may survive constitutional attack if one or more specified exceptions from R.A.V. apply. Id. at 388-90, 112 S.Ct. 2538.
In R.A.V., the Supreme Court considered a St. Paul ordinance that made it a misdemeanor to display a symbol, “including, but not limited to, a burning cross or Nazi swastika, which one knows or has reasonable grounds to know arouses anger, alarm or resentment in others on the basis of race, color, creed, religion or gender.” 505 U.S. at 380, 112 S.Ct. 2538 (emphasis added) (citation omitted). The Supreme Court was bound by our construction of the ordinance, limiting the ordinance to reach only “fighting words,” R.A.V., 505 U.S. at 381, 112 S.Ct. 2538, but nonetheless reversed, concluding that the ordinance was facially unconstitutional. Id. at 396, 112 S.Ct. 2538.
[110]*110The focus of the Court’s reasoning was on the use of the words “on the basis of race, color, creed, religion or gender,” stating that it was “obvious that the symbols which will arouse ‘anger, alarm or resentment in others on the basis of race, color, creed, religion or gender’ are those symbols that communicate a message of hostility based on one of these characteristics.” Id. at 393, 112 S.Ct. 2538 (citation omitted). The Court explained the constitutional problem:
Displays containing abusive invective, no matter how vicious or severe, are permissible unless they are addressed to one of the specified disfavored topics. Those who wish to use “fighting words” in connection with other ideas — to express hostility, for example, on the basis of political affiliation, union membership, or homosexuality — are not covered. The First Amendment does not permit St. Paul to impose special prohibitions on those speakers who express views on disfavored subjects.
Id. at 391, 112 S.Ct. 2538. The Court underscored the nature of the constitutional failing of the St. Paul ordinance when it decided Virginia v. Black, 538 U.S. 343, 123 S.Ct. 1536, 155 L.Ed.2d 535 (2003). In Black, the Court upheld a Virginia statute that banned cross burning “with the intent of intimidating any person or group of persons.” 538 U.S. at 348, 362-63, 123 S.Ct. 1536 (citation omitted). In contrast with the ordinance in R.A.V., the Virginia law survived scrutiny because it squarely based its prohibition on cross-burning upon intimidation — which the Court held is a type of “true threat” not protected by the First Amendment — but not upon intimidation that results from or is based on any topic, subject, idea, or characteristic. 538 U.S. at 360, 362, 123 S.Ct. 1536. It was the inclusion of the “on the basis of’ factors that doomed the ordinance in R.A.V. R.A.V., 505 U.S. at 396, 112 S.Ct. 2538; see Black, 538 U.S. at 362, 123 S.Ct. 1536.
By prohibiting “only a particular type of threat” — cross burning — within the broader unprotected category of “true threats,” the Virginia cross-burning statute fell within the first of three exceptions to the general prohibition against content-based discrimination within unprotected categories of speech announced in R.A.V. Black, 538 U.S. at 362, 123 S.Ct. 1536; R.A.V., 505 U.S. at 388-90, 112 S.Ct. 2538. The first exception to the R.A.V. prohibition can be stated as follows: when the basis for the content discrimination of a subclass “consists entirely of the very reason the entire class of speech at issue is proscribable, no significant danger of idea or viewpoint discrimination exists.” R.A.V., 505 U.S. at 388, 112 S.Ct. 2538. The second exception to R.A.V. states that a “valid basis for according differential treatment to even a content-defined subclass of proscribable speech is that the subclass happens to be associated with particular ‘secondary effects’ of the speech, so that the regulation is ‘justified without reference to the content of the ... speech.’ ” R.A.V., 505 U.S. at 389, 112 S.Ct. 2538 (emphasis omitted) (quoting Renton v. Playtime Theatres, Inc., 475 U.S. 41, 48, 106 S.Ct. 925, 89 L.Ed.2d 29 (1986)). The third R.A.V. exception allows distinguishing a subclass even without identifying “any particular ‘neutral’ basis, so long as the nature of the content discrimination is such that there is no realistic possibility that official suppression of ideas is afoot.” Id. at 390, 112 S.Ct. 2538. Together, the exceptions allow content-based distinctions, within unprotected categories of speech, that pose little danger of government action eliminating or driving out ideas or viewpoints from public conversation. Id. at 388-90, 112 S.Ct. 2538.
[111]*111Turning to this case, because we construe section 609.505, subdivision 2, to reach defamation, R.A.V. controls our analysis here. Subdivision 2 criminalizes a content-based subclass of defamation because it applies to defamation per se that alleges an act of misconduct implicating a peace officer, made to a specific sort of peace officer: one “whose responsibilities include investigating or reporting police misconduct.” Rather than criminalizing defamation generally, subdivision 2 prohibits a subset of defamatory speech with certain content — that a peace officer committed an act of misconduct — made to a certain audience — a peace officer whose responsibilities include investigating and reporting police misconduct. Because the statute addresses a subset of defamatory speech, in order to decide whether section 609.505, subdivision 2, is constitutional, we must examine and apply the exceptions announced in R.A.V. 505 U.S. at 388-90, 112 S.Ct. 2538. We proceed now to those three exceptions.
1.
The first exception identified in R.A.V. is the exception that the Court applied in Black: when the basis for the content discrimination of a subclass “consists entirely of the very reason the entire class of speech at issue is proscribable, no significant danger of idea or viewpoint discrimination exists.” 505 U.S. at 388, 112 S.Ct. 2538. As examples of this exception, in R.A.V. the Court described a subclass of obscenity that is prohibited because of its exceptionally prurient nature; threats against the President, which carry “special force” when compared with “true threats” against other persons; and the decision of a state to “choose to regulate price advertising in one industry but not in others, because the risk of fraud (one of the characteristics of commercial speech that justifies depriving it of full First Amendment protection) is in its view greater there.” Id. at 388-89, 112 S.Ct. 2538 (emphasis added) (citation omitted). As contrasting examples of subclasses that would be improper, the Court pointed to an ordinance that prohibited obscene material that contained certain political messages; a prohibition on threats against the President that “mention his policy on aid to inner cities”; and a ban on commercial messages that applies only to commercial advertising “that depicts men in a demeaning fashion.” Id.
The Court further explained and applied the first R.A.V. exception in Black, concluding that the Virginia cross-burning statute at issue validly banned a “particularly virulent form of intimidation.” 538 U.S. at 363, 123 S.Ct. 1536 (“Instead of prohibiting all intimidating messages, Virginia may choose to regulate this subset of intimidating messages in light of cross burning’s long and pernicious history as a signal of impending violence.”). In contrast, the St. Paul ordinance in R.A.V. that banned the display of a burning cross or other symbols failed because of its “on the basis of’ language, through which the ordinance banned “fighting words ... that communicate messages of racial, gender, or religious intolerance.” 505 U.S. at 393-94, 112 S.Ct. 2538 (emphasis added).
Turning to the issues in this case, we conclude that section 609.505 subdivision 2, fails to meet the first R.A.V. exception.
The Supreme Court describes “[t]he legitimate state interest” underlying defamation as “the compensation of individuals for the harm inflicted upon them by defamatory falsehood.” Gertz v. Robert Welch, Inc., 418 U.S. 323, 341, 94 S.Ct. 2997, 41 L.Ed.2d 789 (1974). Here the basis for the content discrimination of the subclass does not “consist [ ] entirely of the very reason the entire class of speech at issue is proscribable.” R.A.V., 505 U.S. at 388, 112 S.Ct. 2538. As discussed in [112]*112Part 111(B)(2) of this opinion, the secondary effects of the statute, such as the expending of public resources to investigate false reports of misconduct and the diversion of personnel and resources from legitimate reports of crime or misconduct, provide the primary justification for the statute.
2.
The second exception to R.A.V. states that a “valid basis for according differential treatment to even a content-defined subclass of proscribable speech is that the subclass happens to be associated with particular ‘secondary effects’ of the speech, so that the regulation is ‘justified without reference to the content of the ... speech.’ ” R.A.V., 505 U.S. at 389, 112 S.Ct. 2538 (quoting Renton, 475 U.S. at 48, 106 S.Ct. 925).
We note, first, that secondary-effects jurisprudence is an independent and complex area of First Amendment law that, as in Renton, most often applies to municipal efforts to regulate, through measures such as zoning ordinances, constitutionally-protected speech that occurs at legal businesses, such as theaters that exhibit pornographic movies or stage nude dancing.16 See, e.g., Barnes v. Glen Theatre, Inc., 501 U.S. 560, 111 S.Ct. 2456, 115 L.Ed.2d 504 (1991); Renton, 475 U.S. 41, 106 S.Ct. 925.
Knowingly false accusations of misconduct against a peace officer have substan[113]*113tial secondary effects — they may trigger the expenditure of public resources to conduct investigations of the accusations. Minn. R. 6700.2200 requires the chief law enforcement officer to “establish written procedures for the investigation and resolution of allegations of misconduct against” licensed peace officers. These procedures must “minimally specify” the “misconduct which may result in disciplinary action” and “the process by which complaints will be investigated.” Minn. R. 6700.2200. Section 609.505, subdivision 2(b), which was enacted in the same law as subdivision 2(a), indicates that the statute targets recouping the cost of investigations that arise from knowingly false reports of acts of police misconduct. Act of June 2, 2005, ch. 136, § 30, 2005 Minn. Laws 901, 1138. Subdivision 2(b) requires a court to order “any person convicted of a violation of [subdivision 2] to make full restitution of all reasonable expenses incurred in the investigation of the false allegation unless the court makes a specific written finding that restitution would be inappropriate under the circumstances. A restitution award may not exceed $3,000.”17
In addition, other matters requiring police time and attention may suffer as a result of investigations of knowingly false reports of police misconduct. The public resources dedicated to law enforcement agencies are already inadequate in many communities throughout the State to maintain an ideal level of order and safety. The public resources, including wasted police time, expended in investigating knowingly false reports of police misconduct would further exacerbate this problem. Thus, the secondary effects of knowingly false accusations of peace officer misconduct — the expenditure of public resources to conduct investigations and the diversion of those resources away from other matters — -justify the regulation “ ‘without reference to the content of the ... speech.’ ” R.A.V., 505 U.S. at 389, 112 S.Ct. 2538 (quoting Renton, 475 U.S. at 48, 106 S.Ct. 925). We thus conclude that subdivision 2 is valid under the second R.A.V. exception.
3.
The third R.A.V. exception is a general exception that allows distinguishing a subclass even without identifying “any particular ‘neutral’ basis, so long as the nature of the content discrimination is such that there is no realistic possibility that official suppression of ideas is afoot.” Id. at 390, 112 S.Ct. 2538. The Court’s example for the third exception was a prohibition on “only those obscene motion pictures with blue-eyed actresses.” Id. This exception implicates, here, the question of whether a false statement of fact is, or may be, an “idea.”
The Supreme Court has made a distinction between ideas and false statements of fact:
Under the First Amendment there is no such thing as a false idea. However pernicious an opinion may seem, we depend for its correction not on the conscience of judges and juries but on the competition of other ideas. But there is no constitutional value in false statements of fact. Neither the intentional lie nor the careless error materially advances society’s interest in “uninhibited, robust, [114]*114and wide-open” debate on public issues.18
Gertz, 418 U.S. at 339-40, 94 S.Ct. 2997 (quoting New York Times Co. v. Sullivan, 376 U.S. 254, 270, 84 S.Ct. 710, 11 L.Ed.2d 686 (1964)). Nevertheless, false statements of fact are “inevitable in free debate” and “a rule of strict liability that compels a publisher or broadcaster to guarantee the accuracy of his factual assertions may lead to intolerable self-censorship.” Id. at 340, 94 S.Ct. 2997. But false statements of fact that can be characterized as defamation can be proscribed without running afoul of the First Amendment.19 See New York Times Co., 376 U.S. at 279-80, 84 S.Ct. 710 (concluding that a public official may not recover damages for “a defamatory falsehood” absent proof that the statement was made with knowledge it was false, or with reckless disregard of whether the statement was true or false); Garrison, 379 U.S. at 77-78, 85 S.Ct. 209 (applying New York Times standard to a criminal defamation statute).
Turning to the instant case, section 609.505, subdivision 2, criminalizes knowingly false reports of police misconduct, which are false statements of fact. As discussed in Part I of this opinion, we construe “an act of police misconduct” to mean a specific act or omission which violates a policy or rule of professional conduct, adopted by a law enforcement agency, that would expose a peace officer to discipline. In order for a person to be convicted under section 609.505, subdivision 2, he or she must knowingly report a specific act or omission by a peace officer that may subject the peace officer to disciplinary action. For instance, under the statute, a knowingly false accusation that a peace officer engaged in a specific act of illegal racial profiling may be punishable, whereas a statement that “a peace officer is a scoundrel” would never be criminal. Accordingly, we hold that section 609.505, subdivision 2, does not pose a threat of “official suppression of ideas.” R.A.V., 505 U.S. at 390, 112 S.Ct. 2538. We thus conclude that our construction of section 609.505, subdivision 2, meets the third R.A.V. exception.
4.
We note here that reliance on Chaker v. Crogan, 428 F.3d 1215 (9th Cir.2005), cited by Crawley and the court of appeals, is misplaced. In Chaker, the United States Court of Appeals for the Ninth Circuit decided that a California statute prohibiting false reports of police misconduct im-permissibly discriminated on viewpoint because, while the government could prohibit false speech during misconduct investigations, it could not prohibit only false speech that implicated officers. 428 F.3d at 1227 (finding Cal.Penal Code § 148.6 unconstitutional under R.A.V.). Chaker is unpersuasive for several reasons. Most importantly, while the Chaker court applied the R.A.V. rule, it did not mention— let alone analyze — the exceptions announced by the Supreme Court. See Chaker, 428 F.3d at 1227-28. Moreover, in Chaker, the Ninth Circuit defined the unprotected speech at issue to be “know[115]*115ingly false speech” — a category that has been since questioned by federal appellate courts in more recent decisions. Id. at 1228; see Alvarez, 132 S.Ct. at 2545 (plurality opinion) (“[FJalsity alone may not suffice to bring the speech outside the First Amendment.”); 281 Care Comm. v. Arneson, 638 F.3d 621, 633-34 (8th Cir.2011) (declining to recognize “knowingly false campaign speech” as categorically unprotected in absence of Supreme Court precedent).
IV.
Because we construe section 609.505, subdivision 2, narrowly to reach only defamatory speech not protected by the First Amendment, and because the statute falls within two of the R.A.V. exceptions to the constitutional prohibition against content discrimination within a category of unprotected speech, we conclude that section 609.505, subdivision 2 is constitutional. Accordingly, we reverse the judgment of the court of appeals that the statute is unconstitutional. But because Crawley was convicted under section 609.505, subdivision 2, before our narrowing construction of the statute, we reverse her conviction and remand for a new trial.
Reversed and remanded.
Dissenting, STRAS, ANDERSON, PAUL H. and MEYER, JJ.