David S. Han[*]
One of the cornerstones of First Amendment doctrine is the general rule that content-based restrictions on all speech—apart from a few narrow categories of low-value speech—are evaluated under strict scrutiny. As many have observed, this rule has produced considerable strain within the doctrine because it applies the same onerous standard throughout the vast and varied expanse of all non-low-value speech, which includes not only the core, highest-value speech for which such stringent protection is clearly warranted, but also less valuable speech to which the application of strict scrutiny is often dissonant. Nevertheless, traditional accounts maintain that this blunt, highly prophylactic approach is necessary given the significant costs and risks associated with granting courts greater discretion to make value-based speech distinctions.
This Article challenges these accounts. I argue that courts should more explicitly recognize a broad conceptual category of what I call “middle-value speech”—that is, speech that falls within the hazy center of the speech-value spectrum between clearly high-value speech, like political speech or truthful news reporting, and clearly low-value speech, like true threats or incitement. The scope of such speech is vast, potentially encompassing speech as diverse as public disclosures of sensitive private data, sexually explicit speech, professional advice, search engine results, and false statements of fact. Yet current First Amendment doctrine broadly fails to recognize middle-value speech as a discrete conceptual category, and this failure has produced substantial costs in the form of doctrinal distortion and a lack of analytical transparency. These costs have grown precipitously—and will continue to grow—in conjunction with the First Amendment’s broad expansion beyond the familiar precincts of core ideological expression into increasingly eclectic varieties of speech.
I therefore propose an adjustment to the doctrinal framework. Rather than broadly presume that all speech outside of the low-value categories is subject to maximum First Amendment protection, courts should affirmatively designate and carve out the particular categories of high-value speech that merit such protection, in a manner similar to how courts have dealt with low-value speech. Once both low-value and high-value speech categories have been carved out, all remaining uncategorized speech is, by definition, middle-value speech, and courts should adopt intermediate scrutiny as the default rule applicable to all such speech. This approach would greatly reduce the doctrinal distortion and analytical opacity associated with the traditional default rule of strict scrutiny, and it would do so at a limited cost to doctrinal consistency and administrability.
In Young v. American Mini Theatres, Inc., Justice Stevens—writing for a plurality of the Court—famously observed that while “every schoolchild can understand why our duty to defend the right to speak” would apply to “political oratory or philosophical discussion,” “few of us would march our sons and daughters off to war to preserve the citizen’s right to see ‘Specified Sexual Activities’ exhibited in the theaters of our choice.” In other words, although “the First Amendment will not tolerate the total suppression of erotic materials that have some arguably artistic value,” society’s interest in protecting such speech “is of a wholly different, and lesser, magnitude than the interest in untrammeled political debate.”
Whether or not one agrees with Justice Stevens’s characterization of this particular type of speech, the natural intuition underlying his statement is difficult to deny, at least in the abstract: within the incredibly vast and varied expanse of speech covered by the First Amendment, some types of speech simply carry more constitutional value than others. The clearest manifestation of this intuition is the Court’s longstanding recognition of discrete low-value speech categories to which the First Amendment offers minimal protection, such as obscenity, true threats, and incitement. But this intuition also extends to the diverse and ever-expanding range of speech outside of these narrow low-value speech exceptions. It is one thing to say, for example, that political speech constitutes high-value speech, the regulation of which should be subject to the most exacting scrutiny. But what about search engine results? Or detailed factual instructions for illegal or dangerous conduct? Or professional advice given by a lawyer or accountant?
Justice Stevens’s argument for further subdivision of speech based on value, however, has gained little traction with the Court. Rather—save one notable exception—the longstanding default rule of First Amendment doctrine continues to apply, at least as a formal matter: outside of the low-value speech categories, content-based restrictions on speech are evaluated under strict scrutiny, which effectively dooms them to failure. This rule produces considerable strain within First Amendment jurisprudence simply because it applies the same onerous strict scrutiny standard to content-based regulations of all non-low-value speech, even though the vast expanse of such speech encompasses not only the core, highest-value speech for which such stringent protection is clearly warranted, but also—and to a rapidly increasing extent—less valuable speech, to which the application of strict scrutiny is often dissonant.
Academic accounts defending this rigid doctrinal framework, however, argue that the blunt over-inclusiveness of the strict scrutiny default rule represents the lesser evil as compared to granting courts more discretion to make value-based distinctions amongst different categories of speech. Scholars have argued, for example, that even though the current approach might often be ill-fitting, it is nevertheless necessary to adequately protect speech from the conscious or unconscious biases held by judges and juries, to correct for courts’ broad systemic tendency to undervalue the more abstract benefits of free speech in favor of more concrete regulatory interests, and to give clear guidance to speakers, legislators, and courts.
This Article challenges these accounts. I argue that the current state of First Amendment jurisprudence strongly suggests that the doctrinal framework should be modified—albeit cautiously—to grant courts greater freedom to make speech-based value judgments more openly and explicitly. Specifically, I argue that courts should more explicitly recognize a broad conceptual category of what I call “middle-value speech”—that is, speech that falls within the hazy center of the speech-value spectrum between clearly high-value speech, like political speech and truthful news reporting, and clearly low-value speech, like fraudulent speech and incitement. This category of middle-value speech is potentially vast. Areas of speech as diverse as computer code, sexually explicit speech, professional advice, panhandling, search engine results, false statements of fact, and public disclosures of sensitive private data can all be potentially characterized as middle-value in nature. Yet under current First Amendment doctrine, there is only minimal recognition of middle-value speech as a discrete conceptual category, and courts’ broad failure to recognize such a category has substantially undermined both the stability of the doctrine and its capacity to evolve amidst a rapidly changing communications culture.
My argument here is not based on a strong sense that the traditional justifications for the current doctrinal framework are fundamentally flawed. Nor do I argue that the Court was necessarily unwise to adopt this doctrinal framework when it did. Rather, my argument rests primarily on a pragmatic view of First Amendment jurisprudence as it stands today.
First, over forty years of experience with the strict scrutiny default rule has revealed courts’ consistent willingness to surreptitiously evade the formal doctrinal framework when confronted with middle-value speech cases where applying such scrutiny would seem dissonant or anomalous. In other words, although courts have not explicitly recognized various categories of speech as middle-value in nature, they have effectively treated them as such sub rosa, through doctrinal distortion. This doctrinal distortion not only severely limits the theoretical benefits of judicial constraint that underlie the current doctrinal framework, but it also works to destabilize the foundations of First Amendment doctrine as a whole while limiting analytical transparency in courts’ resolution of significant First Amendment questions.
Furthermore, the costs associated with the traditional doctrinal framework have multiplied—and will continue to multiply—given the rapid and continuing expansion of the First Amendment’s coverage well beyond core ideological speech. Today’s First Amendment is far more expansive than could ever have been imagined in the formative years of First Amendment jurisprudence, and the rapid technological and social changes associated with our ever-evolving communications culture ensure that difficult and unique questions will continue to be raised as courts are confronted with novel and increasingly varied categories of “speech.” Such variety and eclecticism magnifies the dissonance associated with a rigid and onerous strict scrutiny default rule, which mandates that even speech that is very different from core political speech nevertheless be treated similarly, even if such a result would seem absurd or anomalous.
Finally, courts’ demonstrated willingness to distort doctrine and the rapid expansion of First Amendment coverage highlight the strong need for increased transparency in courts’ First Amendment analyses. The current all-or-nothing doctrinal approach hinders courts—and society at large—from engaging in the sort of collective dialogue and discussion necessary to advance the common law development of First Amendment doctrine in step with rapidly evolving cultural and technological conditions. Difficult middle-value speech cases represent particularly valuable opportunities for courts to truly confront and grapple with the broad foundational questions of First Amendment doctrine: what exactly makes speech constitutionally valuable, and how such value should be weighed against the social harms caused by speech. Yet the current approach squanders this benefit of such cases, not only because courts have distorted doctrine to mask such judgments, but also because the bluntness of strict scrutiny allows courts to resolve such cases without undertaking any deeper foundational inquiry.
I therefore propose a modification to the doctrinal framework that would incrementally expand courts’ ability to openly and accurately capture the diversity of speech protected by the First Amendment. First, courts should adopt a categorical approach in identifying high-value speech, in the same manner currently used to identify low-value speech. In other words, rather than simply characterize all uncategorized speech as high-value speech subject to maximum First Amendment protection, courts should affirmatively identify and delineate the particular categories of speech that merit such protection under a strict scrutiny standard. Then—once both low-value and high-value speech have been carved out—all remaining uncategorized speech is, by definition, middle-value speech, and I propose that courts adopt intermediate scrutiny as the default rule applicable to all such residual speech, at least provided that the regulation in question is not viewpoint-based in nature. This approach would greatly reduce the doctrinal distortion and lack of analytical transparency associated with the traditional default rule of strict scrutiny, injecting a greater degree of theoretical clarity, candor, and flexibility into a rapidly evolving jurisprudence currently hamstrung by a rigid and opaque doctrinal regime.
It is worth clarifying a couple of important points up front. First, my proposal does not seek to eliminate or drastically alter the content-based/content-neutral inquiry at the center of the doctrine, as others have advocated; it merely seeks to tweak the present doctrine to better account for middle-value speech. Nor is my goal here to argue for a general broadening or narrowing of current First Amendment protection, or to advocate for any specific vision of what sorts of speech ought to be classified as high-value, middle-value, or low-value. The core of my proposal is simply that intermediate scrutiny be the default standard with respect to content-based regulation of residual, middle-value speech—that is, if a court finds that the speech does not fall into a delineated category of low- or high-value speech, it will simply apply intermediate scrutiny.
Second, my argument focuses specifically on considerations regarding the value and harms associated with speech—considerations that play a prominent role in First Amendment analyses. There are, however, other significant considerations that drive courts’ judgments—for example, the government’s motive in enacting a speech regulation or special factual contexts where the government’s regulatory interests may be heightened. This Article does not argue that considerations of speech value and harm are, or ought to be, the sole or primary drivers of First Amendment analysis—it merely recognizes that such considerations play a substantial role in the analysis, whether addressed explicitly or sub rosa.
This Article proceeds as follows. In Part I, I walk through relevant First Amendment theory and doctrine, focusing in particular on the default rule that content-based restrictions on speech are subject to strict scrutiny. In Part II, I introduce and describe the conceptual category of middle-value speech, and I outline the significant costs associated with the current doctrinal framework’s general failure to account for such speech. In Part III, I propose a new approach to remedy these shortcomings: courts should affirmatively designate categories of high-value speech—just as they currently do with low-value speech—and adopt intermediate scrutiny as the default rule applicable to any remaining uncategorized speech. In Part IV, I address a number of potential critiques to my approach, including arguments that it would be insufficiently protective of speech interests and that it would allow for excessive judicial discretion; I also address the practical question of how my proposed approach might realistically be implemented by courts. Part V concludes.
I. Theoretical and Doctrinal Foundations of the Current Strict Scrutiny Default Rule
A. Theoretical Foundations
At the theoretical core of the Free Speech Clause is the basic idea that speech is entitled to greater protection from government regulation than non-speech conduct. If this were not the case, of course, then the provision would be meaningless. Thus, First Amendment doctrine fundamentally rests on the broad theoretical rationales as to why speech is entitled to this special degree of protection; it is these rationales that ultimately drive determinations regarding the breadth and degree of First Amendment protection. Neither the sparse text of the Free Speech Clause nor the historical background surrounding its drafting, however, provide much useful guidance in elucidating these background rationales. Thus, at least as a practical matter, First Amendment doctrine is largely driven by courts’ underlying intuitions as to what exactly makes speech valuable, and these intuitions are the product of the particular theory or theories of speech protection adopted by the courts.
Although many theoretical bases for granting special constitutional protection to speech have been proposed, three particular rationales have dominated both the academic and judicial discourse. The first is the idea that unfettered speech has special value as a means of uncovering truth, an idea famously encapsulated by Justice Holmes’s statement that “the best test of truth is the power of the thought to get itself accepted in the competition of the market.” The second is the idea that unfettered speech is necessary for democratic self-governance; after all, if the citizens in a democracy are the ultimate sovereigns, they must have the freedom to openly debate and discuss matters of public concern to govern themselves effectively. The third is the idea that freedom of speech is an essential aspect of individual autonomy and personhood, and thus represents a good in itself, since “[o]ur ability to deliberate, to reach conclusions about our good, and to act on those conclusions is the foundation of our status as free and rational persons.”
In developing the current First Amendment doctrinal framework, courts have not settled on any single rationale as the definitive theoretical basis for protecting speech. Rather, they have adopted a patchwork approach, recognizing these and other theoretical rationales in different contexts. And in a broad sense, these theoretical rationales represent the bedrock upon which we map the contours of First Amendment coverage and protection. If speech is constitutionally entitled to special protection because it is particularly valuable in some way, our intuitions as to when the value of speech justifies stringent constitutional protection are driven by these foundational theoretical justifications. And these justifications also provide an underlying normative basis for calculating the other side of the fundamental First Amendment calculus: evaluating the degree of social harm caused by the speech. For example, the democratic self-governance theory might dictate that we broadly exclude a person’s strong moral offense to her neighbor’s political views as a colorable speech-based harm—even though such offense would constitute a form of social disutility in the abstract—since that theory rests on the broad need for sovereign citizens to hear out all ideas, whether offensive or not.
Judgments of speech value and harm are certainly not the only relevant inquiries in First Amendment doctrine. Considerations such as the government’s motive in enacting a speech regulation often come into play, and indeed play a vital role in a wide variety of doctrinal contexts. All I mean to say here is that given the lack of guidance in the text or history of the Free Speech Clause, our broad sense of what constitutes the “right answer” in a given case is driven, to a significant extent, by the intuitions we hold regarding the value and harm associated with speech. And these intuitions, in turn, are rooted in the underlying theoretical rationales for protecting speech.
B. The Default Rule of Strict Scrutiny for Content-Based Restrictions on Speech
Perhaps the most important cornerstone of current First Amendment doctrine is the broad default rule that content-based restrictions on speech are evaluated under strict scrutiny. The origins of this rule can be traced back to two cases dealing directly with issues of political dissent: the Supreme Court’s 1970 decision in Schacht v. United States and its 1972 decision in Police Department of Chicago v. Mosley.
Schacht involved a statute that allowed “an actor in a theatrical or motion-picture production” to wear a military uniform only “if the portrayal does not tend to discredit that armed force.” Schacht was convicted under the statute for wearing his army uniform in a “street skit” that protested United States involvement in the Vietnam War. In striking down the statute, the Court focused on the fact that Schacht “was free to participate in any skit at the demonstration that praised the Army, but . . . he could be convicted of a federal offense if his portrayal attacked the Army instead of praising it.” The Court concluded that such a provision “cannot survive in a country which has the First Amendment,” although it did not formally establish any sort of broad rule governing content discrimination.
Two years later, in Mosley, the Court reviewed a Chicago ordinance that broadly prohibited picketing within 150 feet of a school during school hours but specifically carved out an exception for “peaceful picketing of any school involved in a labor dispute.” Mosley had been peacefully picketing outside of Jones High School during school hours, carrying a sign that read: “Jones High School practices black discrimination. Jones High School has a black quota.” Once the ordinance was enacted, however, he was forced to stop his protest. After observing that “[t]he central problem with Chicago’s ordinance is that it describes permissible picketing in terms of its subject matter,” the Court observed that “above all else, the First Amendment means that government has no power to restrict expression because of its message, its ideas, its subject matter, or its content.”
As later decisions would clarify, Mosley established the broad rule that content-based speech restrictions are subject to strict scrutiny—that is, they are upheld only if they are “narrowly tailored to promote a compelling Government interest” and represent the least restrictive means of promoting that interest. And in this particular context, strict scrutiny has broadly been applied by courts in its strictest, least-forgiving form; this reflects the Mosley Court’s absolutist rhetoric, which went so far as to suggest that the government has “no power” to restrict speech based on its content.
Although the rule originated in two cases dealing specifically with direct government regulation of political dissent, it was formulated—and has been applied—as a default rule that applies broadly to all speech. The Court has also, however, carved out a few narrow exceptions to the rule, specifically delineating certain categories of low-value speech that are effectively entitled to no First Amendment protection at all. These oft-recited categorical exceptions include obscenity, true threats, child pornography, incitement, and fighting words, and the government is generally free to regulate these categories of speech based on content under a broadly deferential standard of review. And on a very limited basis, the Court has identified some categories of speech to which the First Amendment offers some lesser degree of protection, such that content-based regulations are evaluated under a form of intermediate scrutiny rather than strict scrutiny; the most notable example of this is truthful commercial speech.
So at least as a matter of formal doctrine, the stringent strict scrutiny default rule generally applies to the vast expanse of uncategorized speech that does not fall into any of the narrowly designated subsets of low-value speech. Furthermore, the Court has, in recent years, expressly discouraged the development of any novel low-value speech categories beyond the limited categories that have already been historically recognized, which means that in the future, the strict scrutiny default rule will presumably apply in the vast majority of cases dealing with novel forms of speech.
This onerous default rule, along with the Court’s stringent limitations on recognizing additional low-value speech exceptions, reflects the massive gravitational pull that the highest-value ideological speech exerts on First Amendment doctrine. And this gravitational pull is unsurprising, given the historical development of the current doctrinal framework. The modern era of First Amendment jurisprudence began in the early twentieth century with a set of cases dealing with textbook examples of political dissent, and it is these early political speech cases that produced probably the two most influential and oft-quoted opinions in all of First Amendment jurisprudence: Justice Holmes’s dissent in Abrams v. United States and Justice Brandeis’s concurrence in Whitney v. California. Indeed, much of the fundamental structure of current First Amendment doctrine can be traced back to a handful of seminal cases dealing with political or otherwise ideological speech in some form.
The doctrinal primacy of the most valuable ideological speech is also reflected in the broad social understanding of the First Amendment, both within and outside of the courts. Mention the First Amendment to a person on the street, for example, and it will most likely evoke some conception of protecting dissent and debate regarding political, religious, or social issues. It is therefore natural that whenever people discuss the freedom of speech in general terms, or when courts construct First Amendment doctrine on a broad architectural level, they tend to do so while implicitly or explicitly conceptualizing speech in its highest-value forms. After all, these are the subsets of speech that best capture our fundamental reasons for protecting speech and thus raise our greatest concerns regarding government regulation.
The strict scrutiny default rule—which, as described above, arose in cases dealing with direct state regulation of political dissent—was thus presumably developed with the highest-value “core” speech specifically in mind. But the blunt presumption underlying the rule—that all speech (at least outside of a few narrow exceptions) is, in effect, as valuable as the highest-value ideological speech—fits poorly with the diverse and ever-expanding range of “speech” protected by the First Amendment. And as I describe in detail below, this wide and growing disconnect between the highly speech-protective default rule and the fact that not all uncategorized speech carries the same degree of First Amendment value has produced significant costs within the current doctrinal framework.
C. Theoretical Justifications for the Current Two-Tier Structure
As described above, the Court has settled on a largely binary, all-or-nothing structure with respect to questions of speech value. It groups the vast majority of speech by default into an expansive category of presumptively high-value speech, to which the onerous rule of strict scrutiny for content-based speech restrictions applies. It recognizes some narrow subsets of speech as low-value speech, the regulation of which is subject to a highly deferential standard of review, but compared to the broad expanse of the First Amendment’s coverage, these low-value speech categories are relatively limited in scope.
Only in the commercial speech context has the Court explicitly characterized speech as falling somewhere within the middle of the speech-value spectrum—what I refer to in this Article as “middle-value speech.” In Ohralik v. Ohio State Bar Ass’n, the Court observed that it “afforded commercial speech a limited measure of protection, commensurate with its subordinate position in the scale of First Amendment values, while allowing modes of regulation that might be impermissible in the realm of noncommercial expression.” Pursuant to this classification, the Court has held that content-based regulations of truthful commercial speech are subject only to a form of intermediate scrutiny rather than strict scrutiny.
In no other context, however, has the Court clearly designated a substantive category of speech as middle-value in nature, such that it is entitled to some intermediate degree of protection between that afforded to high-value speech and low-value speech. Rather—at least as a matter of formal doctrine—it has consistently resisted this sort of value-based differentiation of speech. Indeed, in United States v. Stevens, the Court emphatically rejected the government’s argument that “[w]hether a given category of speech enjoys First Amendment protection depends upon a categorical balancing of the value of the speech against its societal costs,” calling such value-based analysis “startling and dangerous.”
Needless to say, the current two-tier doctrinal structure does not accurately reflect the broad and varied range of speech that is covered by the First Amendment. Given the massive value placed on ideological speech—and, presumably, our heightened suspicions of illicit government motives if such speech is sought to be regulated—it makes sense to effectively bar the government from regulating such speech based on its content, as was the case in Schacht and Mosley. But much of the vast expanse of uncategorized speech is not easily characterized as this sort of particularly valuable speech. Surely, for example, speeches at political rallies carry more First Amendment value than nude dancing, and truthful reporting on issues of public concern is more valuable than false statements of fact. As many have recognized, the formal presumption that all speech is equally valuable simply does not square with the common-sense notion that the diverse and ever-expanding range of speech covered by the First Amendment—even outside of the designated low-value speech categories—carries varying degrees of constitutional value.
But the lack of doctrinal fit produced by the rigidity and over-inclusiveness of the current two-tier structure is not, by itself, a sufficient reason for rejecting or modifying it. There are significant considerations to doctrinal design beyond fit, and one’s decision as to the number of doctrinal categories to recognize directly reflects a judgment as to the best balance between a simple and administrable rule-like regime on the one hand and a more complex and open-ended standard-like regime on the other. For example, a single-category regime that treats all speech identically would be highly rigid and grossly over- or under-inclusive, but it would also be easier to administer, produce greater uniformity and consistency in decision–making, and give clearer notice to speakers and regulators as to what exactly would or would not be permissible. On the other hand, an infinite-category case-by-case balancing regime, although unpredictable and difficult to administer, would allow for fully tailored analyses that account for the precise facts and context of each individual case.
Many scholars have thus defended the current two-tier approach as reflecting the most suitable balance between these considerations, arguing that it is unwise and dangerous to afford courts discretion to make more nuanced value-based distinctions of speech. As Geoffrey Stone has argued, the preference for a blunt, widely over-inclusive framework is driven by the risk that “judges and jurors may be influenced by their own conscious or unconscious biases, which may undermine their ability to evaluate accurately and impartially the extent to which particular content-based restrictions actually impair the communication of specific, often disfavored, messages.” As a result, “the Court has appropriately embraced a ‘fortress model’ of jurisprudence that gives judges little room to maneuver and that intentionally overprotects speech, in order to minimize the potential harm from legislative and administrative abuse and judicial miscalculation.” In other words, the two-tier framework is prophylactic in nature: it adopts a rigid, overprotective doctrinal approach in order to account for the intentional and unintentional errors that might consistently favor regulatory interests over speech.
Such prophylaxis may be necessary for a number of reasons. For one, courts may be broadly reluctant to protect unpopular speech in the face of intense majoritarian pressures. As Vincent Blasi has noted, this danger is maximized particularly in those “pathological” times “when intolerance of unorthodox ideas is most prevalent and when governments are most able and most likely to stifle dissent systematically.” Thus, as Blasi argues, courts, in crafting doctrine, “should place a premium on confining the range of discretion left to future decisionmakers who will be called upon to make judgments when pathological pressures are most intense.” In other words, simple and overprotective doctrine is necessary to ensure that courts and other actors lack the wiggle room to erroneously underprotect speech amidst strong majoritarian pressures to do so.
Furthermore—even outside of these particularly dangerous moments of pathological pressure—courts, legislatures, or local bureaucrats might simply be predisposed to favor speech with which they agree. Or, perhaps, they may systematically overvalue the government’s regulatory interests over the speech interests at stake. After all, the government’s regulatory interests are usually concrete, focused, and immediate: for example, preserving national security, preventing severe emotional distress, or reducing harmful criminal activity. By contrast, the benefits of free speech—such as aiding in the pursuit of truth or supporting democratic self-governance—are far more abstract and dispersed in nature.
Thus, a rigid and highly overprotective two-tier doctrinal structure—rather than one that grants courts and state officials additional discretion to classify speech based on its value—may be necessary to counteract these systemic biases against speech. As Frederick Schauer has observed, “[i]f . . . the first amendment seeks to protect that which may at first sight . . . seem worthless, then we must guard against the pressure to create subcategories that will leave to judges in a particular case the determination of either the truth or the social utility of a covered communicative act.”
Finally, the prophylactic, over-inclusive two-tier approach might be necessary simply to limit the chilling effects on speech that would result from a more complex and nuanced doctrinal structure. If the doctrine were to become more complex, then speakers would become increasingly uncertain as to whether their speech is protected, given that such complexity might strain the interpretive capacities of courts or afford them with greater opportunities to manipulate results. This uncertainty would discourage risk-averse speakers from partaking in constitutionally protected speech, whether because they are unsure of the speech’s protected status or because they are worried about the possibility of intentional or unintentional judicial error.
These arguments all serve as theoretically persuasive justifications for the current two-tier framework, suggesting that any costs in terms of a lack of doctrinal fit are outweighed by the gains of a system that is predictable, administrable, and less subject to dangerous judicial discretion. As I will argue, however, practical realities surrounding the current state of First Amendment jurisprudence have significantly weakened these justifications, strongly suggesting the need for doctrinal reformulation.
II. Middle-Value Speech and the Shortcomings of the Traditional Approach
A. The Concept of Middle-Value Speech
As discussed above, a framework that broadly presumes that all non-low-value speech is high-value speech subject to the most stringent protection against content-based regulations simply does not fit with the reality that the First Amendment covers a broad range of speech of varying constitutional value. In order to better articulate and analyze the substantial costs associated with this lack of fit, I here introduce and describe the idea of what I call “middle-value” speech. As I will argue below, many infirmities within the current doctrinal framework can be traced to courts’ broad failure to account for such speech.
On a purely conceptual level, middle-value speech is exactly as it sounds: it is speech that falls within the hazy middle of the speech-value spectrum, between clearly high-value speech (like, for example, political speech or truthful news reporting) on the one hand and clearly low-value speech (like, for example, fighting words or fraud) on the other. It is therefore useful to define middle-value speech in the negative: it is all of the speech that remains after one carves out all of the highest-value speech and all of the lowest-value speech.
If we broadly conceptualize middle-value speech in this manner—as a sort of residual catch-all—different types of middle-value speech need not share any overarching similarities. It is simply a grab-bag of all speech that, for one reason or another, cannot easily be categorized as high- or low-value. Its distinguishing characteristic is that it has some meaningful degree of First Amendment value—unlike, say, true threats or incitement—but not so much value that we are comfortable affording it the same stringent protection as the highest-value speech—like, say, ideological speech, artistic expression, or truthful news reporting.
This raises the question of how exactly one measures the constitutional value of speech. As I discussed above, given the absence of concrete guidance in the text or history of the Free Speech Clause, the constitutional value of speech is ultimately tied to the varying theoretical rationales as to why speech is entitled to special protection. Thus, the value of a particular subset of speech might be tied to its capacity to advance the pursuit of truth, or to its contribution to democratic self-governance, or to its centrality to individual autonomy.
As I noted, however, the Court has never settled on any single theoretical rationale for protecting speech, choosing instead to recognize different theoretical rationales in different contexts. And although scholars have long argued as to what particular rationale ought to stand as the sole or dominant basis underlying First Amendment doctrine, I do not delve into this debate for present purposes, simply because my discussion here does not rest on any particular normative preference as to how courts should value speech.
That is to say, although I will survey some possible subsets of middle-value speech below, my ultimate goal here is not to argue that any of these particular subsets of speech must or even ought to be classified as middle-value in nature. Nor is it to specifically endorse any particular theoretical rationale for protecting speech; to the extent that I discuss such rationales, such discussion simply reflects the varying theoretical considerations that the Court has set forth in delineating speech value. In the end, my argument requires only a basic recognition that some meaningful subset of speech covered by the First Amendment is not comfortably characterized as either the highest-value core speech or low-value speech—a claim that is, I think, difficult to dispute on its face, regardless of whether one believes that truth-seeking, democratic self-governance, autonomy, or any other consideration ought to predominate in speech-value determinations.
It is also worth clarifying that when I refer to speech as middle-value, I am referring specifically to the inherent constitutional value of the speech itself rather than the contextual, non-speech elements surrounding the speech that might call for greater deference to government regulation. Thus, for example, while the Court has implemented a more deferential standard of review for content-based restrictions of high-value speech within certain special settings—such as, for example, speech of government employees, or speech in prisons, public schools, or the military—these subsets of speech are not actually middle-value in substance. Their differential treatment reflects the heightened government interests associated with the particular circumstances surrounding the speech rather than a value-based judgment regarding the substance of the speech itself. Thus, although the Court has applied a more deferential standard of review to content-based restrictions of both truthful commercial speech and speech of government employees, only the former represents an actual substantive category of middle-value speech.
B. Potential Examples of Middle-Value Speech
If one defines middle-value speech as a sort of residual catch-all—that is, as all speech that is not comfortably characterized as either the highest-value speech or the lowest-value speech—then the potential variety of such speech is vast. As I noted above, truthful commercial speech is the only substantive subset of speech that the Court has formally recognized as middle-value in nature. But there are numerous additional subsets of speech that can likely—or at least plausibly—be classified as such.
I list some of these potential subsets of middle-value speech below. To be clear, this list is not meant to be exhaustive, nor is it simply a product of my own normative predilections as to what might or should count as middle-value speech. Rather, the subsets listed below have been gleaned from a survey of both case law and the academic literature; they represent areas of speech that have already been the subject of targeted scholarly and/or judicial analyses given the significant doctrinal tension created by their middle-value nature. These areas are generally marked by a substantial degree of doctrinal confusion or distortion, which can manifest itself in different ways, including disagreement as to whether they should even qualify as “speech” covered by the First Amendment in the first instance. They include:
Sexually explicit (but not obscene) speech
Professional speech made in the context of one’s occupation
False statements of fact
Search engine results
Automated, fact-based recommendations
Indeed, a significant proportion of purely factual speech might be classified as “middle-value” in nature, as such speech often raises distinct and more difficult analytical issues as compared to the core context of “normative, religious, ideological, and political disagreements” that underlies current First Amendment theory and doctrine. As Ashutosh Bhagwat has catalogued, this might include:
Factual instructions for illegal or dangerous activities
Public disclosures of personal data or highly offensive private facts
Scientific and technical details
Military and diplomatic secrets
Again, this list is not meant to be exhaustive, and there is strong reason to think that the scope of middle-value speech is substantially broader than the potential categories listed above.
So what, precisely, makes these subsets of speech potentially middle-value in nature? Again, this judgment is ultimately rooted in the intuition that they simply do less to advance the theoretical rationales for protecting speech than, say, core political speech. To use the Ohralik Court’s language, middle-value speech is afforded “a limited measure of protection, commensurate with its subordinate position in the scale of First Amendment values, while allowing modes of regulation that might be impermissible in the realm of [high-value] expression.” If, for example, one deems democratic self-governance to be the sole or dominant theoretical rationale driving the First Amendment’s protection of speech, then it seems quite natural to say that commercial speech, or nude dancing, or private lies about receiving military medals directly advance such values to a meaningfully lesser extent than political, religious, or other ideologically-oriented speech.
In theory, this fundamental idea of speech value is distinct from the idea of speech-based harm. But as a practical matter, judgments regarding the degree of social harm associated with the speech are often blended into the speech-classification analysis. As I have noted elsewhere, all of the designated categories of low-value speech—such as fraud, true threats, and obscenity—are at least partially defined in terms of the social harm they cause. These judgments of speech harm are premised on both normative considerations as to the appropriate weight to be afforded to different types of social harms and empirical judgments as to the actual or likely degree of harm associated with the speech in question. Thus, in practice, speech value judgments are often based not simply on innate characteristics of that speech, but also on the speech’s association with socially harmful consequences. For example, courts might explicitly or implicitly classify panhandling as middle-value speech not only because its inherently transactional nature makes it less central to democratic self-governance than core political speech, but also because it represents a particularly harmful type of personal harassment.
Again, however, my argument here does not rest on any broad agreement that any or all of the subsets of speech listed above should be regarded as middle-value in nature. Perhaps plausible arguments can be made that one or more of these subsets should, in fact, be deemed as valuable as core political speech. But I suspect that most will agree, at least in the abstract, that there is some meaningful difference in speech value between many of the subsets of speech listed above and the highest-value core speech, whether based on autonomy, truth-seeking, democratic self-governance, or any other theoretical considerations. And this difference in speech value produces tension in applying the onerous strict scrutiny default rule to all cases; at the very least, the subsets of speech listed above pose considerably more difficult questions regarding the appropriate extent of government regulation than, for example, core political speech.
C. The Present and Future Scope of Middle-Value Speech
If one were to recognize a conceptual category of middle-value speech, then the natural next question is how expansive this category might be. After all, if middle-value speech is just a minimal, sui generis sliver of speech, then perhaps we need not bother with any sort of deep reevaluation of the present doctrinal framework, even if we think that courts deal with such speech in problematic ways.
To get a sense of the scope of middle-value speech, we can start with the many discrete subsets of potential middle-value speech listed above—subsets that, as I explained, have already been implicitly or explicitly identified within the case law or contemporary scholarship as standing in tension with the strict scrutiny default rule. The breadth and variety of these areas strongly suggest that middle-value speech is not a highly anomalous, sui generis category, but rather a meaningful portion of the body of “speech” protected by the First Amendment.
Furthermore, an examination of the current frontiers of speech regulation and the long-term trajectory of modern First Amendment jurisprudence strongly suggests that the scope of middle-value speech is substantially broader than the potential categories listed above, and that it is likely to expand rapidly in the coming years. Much of contemporary First Amendment scholarship has focused on the constitutional status of, for example, search engine results, computer code, automated recommendations, pervasive video recording, or the widespread collection and dissemination of personal data—heretofore unexplored issues arising out of the many recent social and technological changes associated with our communications culture, such as the advent of mass data collection and analysis, the ubiquity of the internet and smartphones, the rise of social media, and the development of algorithm-based “smart” communications.
These potential subsets of speech are, on their face, very different in nature from the highest-value ideological speech most commonly associated with the First Amendment. They often implicate the preliminary question of whether the First Amendment even applies in the first instance, and in the event that they are classified as speech, a strong intuition remains that many of these subsets of speech are middle-value in nature—that is, that they carry some meaningfully lesser degree of constitutional value than, say, core political speech. For example, if one subscribes to a democratic self-governance theory of speech protection, it seems reasonable to argue that mass disclosures of sensitive private data broadly carry less constitutional value than speeches at political rallies or truthful reporting of newsworthy events. And as shifting social and technological conditions continue to drastically alter the landscape of our communications culture, courts will be increasingly forced to confront novel and difficult speech contexts that do not fit easily within the traditional conception of core protected speech upon which the doctrinal framework has been built.
This expansion of the breadth and variety of middle-value speech covered by the First Amendment is also consistent with the broad historical trajectory of First Amendment jurisprudence. The story of modern First Amendment doctrine has been one of steady expansion beyond the core categories of the highest-value speech that have disproportionately shaped the doctrine’s fundamental design. If, for example, attorneys in the 1940s were to suggest that nude dancing, crush videos, commercial advertising, or false statements of fact were entitled to protection under the First Amendment, they would likely be laughed out of the room. As Lillian BeVier has noted, “[b]efore the Court’s extension of First Amendment protection to commercial speech in 1976, the overwhelming majority of First Amendment cases involved attempts to regulate speech that was in one way or another speech about government.”
And this expansion continues to the present day. As Schauer has observed, recent cases have evinced an “accelerating attempt to widen the scope of First Amendment coverage to include actions and events traditionally thought to be far removed from any plausible conception of the purposes of a principle of free speech”—a trend perhaps driven by the unique magnetism and attractiveness of First Amendment arguments in a broad range of legal disputes. Thus, for example, the First Amendment has recently been invoked by companies arguing against mandated disclosures to the SEC, by tattoo parlors seeking to be shielded from health regulations, and by therapists seeking to escape state regulation of scientifically unproven therapeutic methods.
Much of this ostensible “speech” had historically been treated as “uncovered” by the First Amendment—that is, it resided completely off of the First Amendment’s radar, failing to trigger any sort of serious First Amendment analysis. The mere fact that these sorts of arguments are increasingly being made by litigants—and meaningfully considered by courts—indicates that the vast expanse of uncovered speech is shrinking as more and more “speech” is integrated into the First Amendment’s doctrinal framework. And, of course, much of this realm of heretofore uncovered “speech”—such as speech subject to regulation under securities laws, antitrust laws, labor laws, evidence law, and so forth—is a far cry from the highest-value speech that resides at the core of First Amendment doctrine. Thus, to the extent that uncovered speech continues to be invited into the First Amendment fold, such speech is most likely to be middle-value rather than high-value in nature.
All of this suggests a fundamental disconnect between a doctrinal system built around a broad assumption that nearly all speech is as valuable as political speech and the fact that a substantial—and rapidly expanding—proportion of the theoretical realm of “speech” protected by the First Amendment may well be middle-value in nature. Indeed, it may be that given the massive breadth of speech potentially subject to government regulation, it is the highest-value speech—like political speech, artistic expression, truthful news reporting, or academic debate—that represents the outlier, and that much of what we deem to be “speech” potentially subject to constitutional protection is speech that we broadly intuit to be middle-value in nature. And if, as I argue below, courts’ present treatment of middle-value speech incurs significant costs, these costs will only multiply as the proportion of middle-value speech covered by the First Amendment continues to grow, widening the already substantial gap between reality and the doctrinal presumption that all speech is high-value.
D. The Shortcomings of the Current Doctrinal Framework
As discussed above, traditional accounts hold that the current strict scrutiny default rule—despite its blunt and overprotective nature—is necessary to protect speech adequately, particularly given the risks posed by conscious and unconscious judicial biases. But although such arguments have their merits in the abstract, the current state of First Amendment doctrine strongly suggests that the benefits of the current approach are outweighed by its costs, and that some degree of doctrinal adjustment is therefore necessary.
On one side of the ledger, the theoretical benefits of the formally rigid, highly prophylactic binary approach have not been actually realized in many areas of First Amendment doctrine, simply because the substantial tension between an often ill-fitting strict scrutiny default rule and the wide variety of speech covered by the rule has consistently driven courts to distort doctrine to reach desired results rather than adhere to the formal rules. A simple, prophylactic rule meant to constrain judicial discretion has little value if, in practice, courts freely (and surreptitiously) distort formal doctrine to avoid applying the rule when it does not seem to fit. Thus, the actual benefits derived from the current doctrinal structure have proved to be more limited than theoretically envisioned.
On the other side of the ledger, the costs inflicted by the current doctrinal regime on the structural integrity of First Amendment doctrine and the doctrine’s capacity to adapt to a rapidly changing communications culture have been substantial. The doctrinal distortion produced by courts in many areas of First Amendment jurisprudence threatens to unsettle well-established doctrinal structures throughout. Furthermore, the current approach greatly limits the degree of meaningful analytical transparency in difficult middle-value speech cases where such transparency is most valuable, since it allows courts to hide behind formal doctrine without articulating, in clear and direct terms, what exactly makes the speech in question constitutionally valuable and how that value compares to the social harms associated with the speech. Not only are these costs substantial, but they are likely to increase significantly as the First Amendment’s coverage continues to expand into unexplored and novel realms of speech.
1. Doctrinal Distortion Arising from Imprecise Doctrinal Fit
In middle-value speech cases, applying the strict scrutiny default rule is intuitively dissonant because the outcome dictated by the rule—near-automatic invalidation—simply does not match our foundational intuitions regarding the value of the speech in question and/or the social harms associated with the speech. Such dissonance, however, is not necessarily a problem in doctrinal design that requires fixing; particularly in the First Amendment context, it is often worth trading off equitable flexibility for a more rigid but administrable doctrinal framework.
Thus, within the formal confines of the doctrinal framework, courts confronted with this dissonance in a given case can take one of two approaches. They can simply grit their teeth and apply the strict scrutiny default rule to invalidate the regulation, despite any strong intuitive judgment that such a result is incorrect. Or they might carve out a new categorical exception that takes the speech in question outside of the default strict scrutiny rule—that is, they can craft a new category of low-value or middle-value speech.
The plurality opinion and Justice Breyer’s concurrence in United States v. Alvarez illustrate these two approaches in action. In Alvarez, the Supreme Court struck down the Stolen Valor Act, a federal statute that criminalized lying about having received military medals. In doing so, both the plurality and Justice Breyer agreed that false statements of fact of this sort did not constitute low-value speech. But there is also undoubtedly a strong intuition that such false statements of fact are not as valuable as, say, truthful political speech, and that the same stringent standard of review applicable to the highest-value speech ought not to apply to lies about military honors.
The plurality dealt with this dissonance by simply applying the strict scrutiny default rule to invalidate the statute, regardless of any strong intuitive judgment that such an approach might seem anomalous or incorrect. On the other hand, Justice Breyer proposed modifying the doctrinal framework to account for this dissonance by explicitly carving out false statements of fact as a new category of partially protected speech, such that any content-based regulation would be subject to intermediate scrutiny rather than strict scrutiny. Both of these approaches account for the dissonance within the confines of the doctrinal framework, either by absorbing it in the former approach or by creating a formal exception in the latter.
If courts adhere to these two approaches in dissonant cases, then the doctrine is at least accomplishing its intended purpose: constraining judicial discretion, or—at the very least—requiring courts to explain themselves through formal doctrinal modification if they wish to craft exceptions to existing rules. Both of these approaches incur some costs—reaching results that seem anomalous in the first approach, or eroding the predictability and consistency associated with a simple bright-line rule in the second—and which approach is preferable rests on some judgment as to which of these costs is more severe. But as long as courts adhere to these two options, the formal doctrinal framework retains some degree of integrity and structure, as it accurately reflects courts’ actual decision–making processes.
There are strong indications, however, that courts are often disinclined to take either approach. On the one hand, courts generally do not like to reach results that seem incorrect, absurd, or otherwise anomalous. Take, for example, the recent Reed v. Town of Gilbert case, in which the Supreme Court struck down an Arizona sign ordinance. As the Court observed, the ordinance in question was clearly content-based on its face, as it accorded differential treatment to signs based on whether they were “ideological signs,” “political signs,” or “temporary directional signs.” Nevertheless, in the case below, the Ninth Circuit had classified the ordinance as content-neutral and upheld it under intermediate scrutiny—an approach likely dictated, to a significant extent, by a strong reluctance to apply the onerous strict scrutiny standard to a seemingly benign sign regulation.
On the other hand, the relative paucity of low-value speech exceptions suggests that courts are not generally inclined to craft broad exceptions to the strict scrutiny default rule. There are many possible reasons for this: perhaps lower courts are institutionally averse to crafting such exceptions without Supreme Court guidance; “perhaps it reflects courts’ general reluctance to craft numerous exceptions to bright-line rules; [or] perhaps courts simply wish to avoid any negative perception associated with adopting formal exceptions that limit First Amendment protections.”
In any case, even if courts were in fact inclined to create formal exceptions to the default rule, the Supreme Court’s recent adoption of a purely historical approach to low-value speech appears to greatly limit their ability to do so. In United States v. Stevens, the Court explicitly rejected the government’s argument that “[w]hether a given category of speech enjoys First Amendment protection depends upon a categorical balancing of the value of the speech against its societal costs,” even though the government’s proposed test merely reflected the Court’s longstanding characterization of low-value speech as speech “of such slight social value as a step to truth that any benefit that may be derived from [it] is clearly outweighed by the social interest in order and morality.” Rather, the Court held that the test for low-value speech is purely historical, based only on whether the subset of speech in question “ha[s] been historically unprotected.”
As I have argued at length elsewhere, the Court’s purely historical test is a poor means of delineating low-value speech; it offers little more than a veneer of objectivity, predictability, and constraint that ultimately works to obscure the underlying judgments regarding speech value and harm that actually drive the analysis. Nevertheless, the Court’s adoption of this test will likely deter courts from crafting novel categories of low-value speech. Although savvy courts could certainly find ways to characterize a wide variety of speech as historically unprotected, Stevens and its progeny send a clear message that the Court intends to curb any further expansion of the low-value speech doctrine. Thus, given that courts already appeared reluctant to carve out formal exceptions to the default strict scrutiny rule prior to Stevens, the Court’s imposition of a facially strict, history-based limitation on the expansion of low-value speech categories will discourage courts even further from doing so, even in cases where the default rule leads to an absurd or anomalous result.
If courts are broadly disinclined to reach intuitively incorrect results, and if they consider themselves handcuffed from crafting formal doctrinal exceptions, that leaves them with a third option: doctrinal distortion. That is, they might surreptitiously distort the doctrinal framework “to reach the ‘correct’ result in cases where the onerous strict scrutiny standard does not appear to fit.” They might, for example, water down the strict scrutiny standard to something resembling intermediate scrutiny, classify clearly content-based regulations as content-neutral regulations, characterize expression as non-speech conduct that falls outside of the scope of First Amendment coverage, or simply leave the specific standard of review applied in a case intentionally vague.
Probably the most notable example of such distortion is the Supreme Court’s use of the “secondary effects” doctrine in a series of cases dealing with zoning restrictions on sexually-oriented businesses. As discussed at the beginning of this Article, Justice Stevens, writing for a plurality of the Court in Young v. American Mini Theatres, Inc., had suggested that the government has greater latitude to regulate sexually explicit (but non-obscene) speech based on its content as compared to political speech. Nine years later, however, the Court adopted a completely different approach in City of Renton v. Playtime Theatres, Inc. There, the Court reviewed a zoning restriction that prohibited any “adult motion picture theater” from being located in close proximity to residential housing, parks, churches, or schools. Although the ordinance was clearly content-based on its face, the Court treated it as content-neutral, since the regulation “aimed not at the content of the films shown at ‘adult motion picture theatres,’ but rather at the secondary effects of such theaters on the surrounding community,” with goals such as preventing crime, preserving property values, and protecting retail trade. The Court therefore evaluated the ordinance under intermediate scrutiny rather than strict scrutiny and upheld it.
The secondary effects doctrine has been subject to nearly universal criticism. Stone, for example, called Renton “a disturbing, incoherent, and unsettling precedent” that “threatens to undermine the very foundation of the content–based/content-neutral distinction.” Other scholars have similarly described the extent to which Renton “warped” content-neutrality analysis by “turn[ing] the Court’s traditional focus on the language of statutes on its head.” Indeed, a number of the Justices themselves have acknowledged that characterizing such ordinances as content-neutral is “something of a fiction.” As Alan Brownstein observed, “[a]lthough the Court never explicitly affirms the view that sexually explicit expression is a generally less valuable form of speech . . . , no other explanation of Renton is plausible.”
As Bhagwat has noted, courts have similarly distorted doctrine in a wide variety of cases dealing specifically with the communication of detailed, purely factual information. He discusses, for example, the infamous “Nuremburg Files” case, in which the American Coalition of Life Activists (“ACLA”), an antiabortion group, posted on its website “Wanted” posters that targeted specific abortion doctors and included personal information such as photographs of the doctors and their home and work addresses. Although three doctors previously featured on such posters had been murdered, the website itself did not contain any direct threats of violence against the targeted individuals—indeed, the site “foreswore the use of violence and advocated lawful means of persuading plaintiffs to stop performing abortions or punishing them for continuing to do so.”
A group of doctors featured on the website sued the ACLA for civil damages and injunctive relief under a federal statute, and in an en banc opinion, the Ninth Circuit upheld a jury verdict in the plaintiffs’ favor. In doing so, the majority classified the speech in question as low-value “true threats” not subject to the default rule of strict scrutiny. But as Judge Kozinski observed in dissent, it seems inaccurate to classify the speech as such, even if it was designed to intimidate the doctors. As he noted, “[a] true threat warns of violence or other harm that the speaker controls,” yet there was no indication that the ACLA or its members were threatening violent action. Rather, the more doctrinally appropriate test would be the stringent Brandenburg standard, which clearly would not be met by the facts of the case.
This sort of doctrinal distortion—or, to put it more charitably, doctrinal confusion—extends to other contexts dealing with factual information, such as detailed instructions for illegal or dangerous behavior or technical details. As Bhagwat observed, despite the fact that “modern doctrine would seem to be extremely hostile to attempted regulation of factual details, . . . [courts] have in fact been far more lenient than doctrine would seem to permit,” since “when faced with such regulations, courts have tended to twist or even ignore that doctrine.”
As a final example, consider the Court’s recent opinion in Williams-Yulee v. Florida Bar, in which the Court upheld a Florida Bar rule that prohibited candidates in judicial elections from personally soliciting campaign funds. Chief Justice Roberts—writing for only a plurality of the Court on this point—observed that the rule was clearly a content-based restriction on speech, and thus strict scrutiny was the appropriate standard to apply. Despite noting that it is a “rare case” in which a content-based speech restriction satisfies strict scrutiny, the Court held that the rule in question was, in fact, narrowly tailored to serve a compelling government interest and thus survived strict scrutiny. Yet as the dissenting Justices observed, the Court applied a watered-down version of strict scrutiny more akin to intermediate scrutiny than the “fatal in fact” version it usually applies in speech cases. The rule in question failed to draw distinctions based on obvious, analytically significant factors such as the identity of the people solicited and the method of solicitation, and thus did not come close to matching the far more rigorous tailoring required by the Court in other speech cases.
These examples of doctrinal distortion used to avoid the consequences of the strict scrutiny default rule are by no means exhaustive. Such distortion arises in a wide range of cases, and it takes many different forms. Furthermore, courts’ efforts to avoid anomalous results in applying the strict scrutiny default rule can influence the distinct, theoretically prior inquiry of whether certain actions constitute speech or non-speech conduct. As Schauer has noted, “if we take the ‘full protection within’ rule as the standard, there may be pressure to keep troublesome categories completely outside” the scope of First Amendment coverage. If, for example, courts are reluctant to treat detailed instructions on how to be a contract killer as fully protected speech, they might choose to classify such instructions as non-speech conduct in order to avoid applying the strict scrutiny default rule.
Consider occupational speech—that is, speech made in a professional capacity by people like doctors, lawyers, or financial advisors. Much of this speech is subject to stringent regulations—such as licensing requirements—that are clearly content-based in nature, yet courts have quite reasonably taken the position that the government should have some meaningful latitude to regulate occupational speech in this manner. Thus, to account for the substantial tension between this strong intuition and the onerous strict scrutiny default rule, courts have often characterized occupational speech as non-speech conduct falling completely outside of the scope of First Amendment coverage, despite strong arguments that such an approach is inconsistent with existing case law or prevailing First Amendment theory.
In all of these cases, courts have distorted doctrine to avoid the anomalous consequences of applying an onerous rule of strict scrutiny to cases where the rule does not fit. Their actions reflect an underlying intuition that sexually explicit speech, or occupational speech, or disclosures of personal information meant to intimidate carry meaningfully less constitutional value than the core categories of highest-value speech—like political speech or truthful news reporting—around which current First Amendment doctrine has largely been built. And these are not merely isolated examples; courts have grappled with this dissonance in cases dealing with, for example, charitable solicitations, panhandling, mass collection and dissemination of sensitive personal data, and profanity.
This doctrinal distortion undermines the theoretical benefits of creating an overprotective but administrable prophylactic rule. In areas where courts are not actually adhering to the rule, but instead distorting it sub rosa because it does not fit, the rule produces no effective judicial constraint, no consistency in application, and no predictability for litigants and lawmakers. And as I discuss in greater detail below, it allows courts to reach their desired results with no meaningful explanation for their actions.
Such distortion is a symptom of structural defects within the current doctrinal framework; it indicates that the doctrine is not operating as intended. And once a particular means of distortion is introduced, it greases the wheels for future courts to follow the same path when confronted with a similarly ill-fitting case. The Williams-Yulee Court’s use of a clearly watered-down strict scrutiny standard, for example, was heavily influenced by Burson v. Freeman, a 1992 case reviewing restrictions on campaign speech close to polling places on election days. In Burson, a plurality of the Court applied a similarly diluted version of strict scrutiny to uphold the content-based regulation, and the Williams-Yulee Court repeatedly and conspicuously relied upon that plurality opinion to justify its approach.
Furthermore, the costs created by such distortion potentially extend well beyond the particular precincts of First Amendment doctrine where they originally arise: doctrinal distortion in one area threatens to infect and destabilize the entire doctrinal framework. For example, the Court’s adoption of the secondary effects doctrine in the narrow context of adult-oriented businesses ultimately worked to destabilize the fundamental content-neutrality inquiry on a broader scale. Although the Court has never technically applied the secondary effects doctrine outside of this specific context, the suggestion in those cases that a facially content-based restriction can be deemed content-neutral based solely on motive analysis led to substantial disruption and confusion throughout the doctrine, which the Court only recently clarified in Reed. When doctrinal stability is undermined in this manner, the benefits of a simple, administrable rule-like regime begin to collapse; the doctrine becomes less predictable, more susceptible to abuse, and more opaque, which is a particularly troublesome state of affairs in the speech regulation context.
2. A Lack of Analytical Transparency
Courts’ broad failure to account for middle-value speech imposes another significant cost: it undermines the transparency of courts’ analyses, which hinders courts from developing the sort of collective dialogue and discussion necessary to advance the development of First Amendment doctrine. I discussed the broad importance of doctrinal transparency in First Amendment jurisprudence in detail in a previous article, and as I did there, I broadly define doctrinal transparency here as the extent to which doctrine encourages or forces courts to analyze speech cases openly, in a fashion that elicits direct discussion of foundational questions regarding the constitutional value of the speech in question and the social harms associated with it.
Doctrinal transparency is particularly important in First Amendment jurisprudence because it represents a vital means for courts—and, by extension, society in general—to evaluate and think through our fundamental intuitions as to why we value speech and how that value ought to compare to the different harms associated with speech. Hollow sloganeering often stands in for rigorous analysis in First Amendment cases; transparent doctrine avoids this by forcing courts to clearly articulate the underlying reasons behind their judgments of speech value and harm, which in turn sets the stage for rigorous deliberation and debate amongst courts and the public at large. It thus helps to create a positive feedback loop in which courts, by openly working through their judgments regarding speech value and speech harm, drive the public discussion, which may in turn shape the judgments made by courts in similar cases down the road.
Transparency is particularly valuable in advancing the evolution of these fundamental intuitions amidst a rapidly changing communications culture. The issues we face today are exceedingly different from those we faced even twenty years ago, and transparent doctrinal approaches give courts the tools to consider and debate, in a direct and open manner, the ways in which the First Amendment should apply to the broad range of novel questions now confronting them. Doctrinal transparency also encourages courts to review the efficacy of existing rules by reevaluating the extent to which they fit our foundational intuitions regarding speech value and harm as those intuitions continue to evolve. Such transparency ultimately prevents courts from losing sight of the normative superstructure underlying the doctrinal edifice they are creating: “the foundational reasons why we attribute special value to speech and our judgments as to how this value should be measured against different types and degrees of social harm.”
Of course, although transparency is particularly valuable in the First Amendment context, it is not the only consideration in crafting sound First Amendment doctrine. Doctrinal transparency is broadly associated with more open-ended, standard-like approaches, which may not be tenable in the First Amendment context given the risks of unchecked judicial discretion and chilling effects on speech. Some degree of opaque but administrable rules are necessary within speech doctrine to offer a measure of predictability and consistency. The key question in designing the doctrine is whether the optimal balance between opaque, rule-like approaches and transparent, standard-like approaches has been met.
The current doctrinal structure provides little in the way of meaningful analytical transparency. Particularly in the hardest middle-value speech cases—where direct, nuanced articulations of difficult speech value and harm judgments are presumably most valuable—the traditional strict scrutiny default rule allows courts to hide behind formal doctrine without being forced to openly grapple and engage with these difficult foundational questions. This is simply because under the present doctrine, the result is effectively foreordained in nearly all cases involving content-based regulations. If the speech in question is deemed to fall into one of the low-value speech exceptions, then the government is broadly free to regulate it; if it does not, then strict scrutiny applies, and the regulation will essentially always be struck down as a matter of course. Although courts might happen to embroider their analyses in these cases with deeper explanation, they generally need not do so, since such analyses carry little weight in the face of a foregone conclusion: except in the most extraordinary of circumstances, strict scrutiny means automatic invalidation when applied in the context of speech regulation.
Analytical transparency—and all of the benefits associated with it—is further undermined by the substantial degree of doctrinal distortion resulting from courts’ efforts to avoid the consequences of the strict scrutiny default rule in ill-fitting cases. Such distortion works to obscure the underlying value judgments actually driving the results reached by courts, as it clothes these judgments in formal “doctrinal” terms that allow courts to sidestep any meaningful and direct articulation of them. Take, for example, the Supreme Court’s decisions in Renton and Williams-Yulee, or the circuit courts’ decisions in Reed and the ACLA case. In all of these cases, the courts broadly characterize their decisions as the inevitable products of formal doctrine, when in fact they were likely driven largely by unarticulated value judgments—for example, foundational judgments as to the relative value of sexually explicit speech or the relative value and harm of posting personal information online as a means to intimidate.
Such distortion can further exacerbate doctrinal opacity by breeding analytical confusion, which limits the extent to which courts can directly and accurately grapple with the central underlying questions of speech value and harm. A court that, for example, inaccurately characterizes a content-based speech restriction as content-neutral ensures that the wrong analytical question will be asked: does the “content-neutral” law in question survive intermediate scrutiny? The analytical question that actually captures the issue posed in such a case is a very different one: even though the speech restriction is content-based, why should our fundamental judgments regarding speech value and harm dictate that the default strict scrutiny standard not apply? By allowing courts to decide cases by asking the wrong questions, such doctrinal distortion impedes courts—and society at large—from participating in meaningful dialogue regarding fundamental questions of speech value and harm, transforming what ought to be open and direct debate regarding such issues into squabbles over formal doctrine. It invites courts and critics to simply talk past each other and miss the underlying bases for disagreement, thereby limiting the potential for direct and forthright discussion.
Thus, both the inherent design of the current doctrinal framework and the significant doctrinal distortion produced by this design have resulted in a largely opaque doctrine—one that allows courts to broadly avoid meaningful articulation of the foundational value judgments that underlie the doctrine as a whole. Such opacity limits courts’ capacity to grapple with the difficult questions of speech value and harm posed in middle-value speech cases, particularly in cases raising novel issues. This cost is therefore likely to be particularly high given the consistent expansion of First Amendment coverage into novel realms of “speech” and the rapid social and technological changes within our communications culture.
III. A Revised Approach to Content-Based Speech Regulation
Thus, applying the traditional strict scrutiny default rule in middle-value speech cases has proven to be costly. On the one hand, the theoretical benefits of the traditional rule have been significantly undermined by courts’ demonstrated willingness to distort doctrine; on the other hand, the costs produced by the analytical opacity associated with the traditional rule are both substantial and likely to grow rapidly given the current trajectory of First Amendment jurisprudence. How, then, might we recalibrate the doctrinal framework to remedy these problems?
As an initial matter, any solution must be a balanced one; a radical shift to a fully discretionary and open-ended framework, for example, would be dangerous and unwise for the reasons outlined by many of the proponents of the current framework. An optimal approach should continue to recognize the broad need for judicial constraint, predictability, and doctrinal consistency while ensuring an incrementally greater degree of flexibility to limit doctrinal distortion and promote analytical transparency where warranted.
I therefore suggest an approach that incrementally shifts the First Amendment’s center of gravity away from the highest-value speech and closer to the middle of the speech-value spectrum. First, categories of high-value speech should be categorically carved out, in the same way that courts already carve out categories of low-value speech. In other words, rather than operate under an assumption that, as a default, all speech is high-value speech subject to the most stringent First Amendment protection, courts should affirmatively designate the categories of the highest-value speech to which strict scrutiny applies—for example, political speech, artistic expression, truthful news reporting, academic debate, and so forth. Once courts have carved out both the lowest and the highest-value speech, the uncategorized speech that remains is the residual category of middle-value speech—and for this remaining speech, I propose that intermediate scrutiny, rather than strict scrutiny, apply as the default standard, provided that the regulation in question is not viewpoint-based in nature.
A. Adopting a “Defining in” Approach to High-Value Speech
A natural reaction to my proposal may be that it overcomplicates things. A simpler approach, perhaps, would be to retain the strict scrutiny default rule but allow courts to more aggressively identify and carve out categorical middle-value speech exceptions subject to a lesser degree of scrutiny. Such an approach would effectively extend what the Court has already done with truthful commercial speech; courts could delineate panhandling, or computer code, or public disclosures of sensitive private information as middle-value speech exceptions such that content-based regulations would be evaluated under intermediate scrutiny. Like the Court’s current approach to low-value speech, strict scrutiny would remain the default rule, but courts could carve out categorical exceptions for middle-value speech under appropriate circumstances.
Both approaches may well be effective in alleviating the problems associated with the current doctrinal framework, but they attempt to do so through fundamentally different means. The current doctrinal framework crafts the category of high-value speech by “defining out”—that is, it presumes that all speech is high-value, and it earmarks certain subsets of speech for different treatment by carving them out from this default position. The approach described above adheres to this “defining out” approach, as it merely pushes courts to be more solicitous in “defining out” categorical middle-value speech exceptions from the default high-value classification. My proposed approach, however, adopts a “defining in” approach to high-value speech. That is, speech is by default categorized as middle-value rather than high-value in nature, and if courts want to instead recognize certain subsets of speech as high-value, they must specifically designate the speech as such to take it outside of the default position.
One might argue that a “defining out” approach to high-value speech is preferable in light of our sense of what biases are most likely to infect the analysis or what values ought to trump under conditions of uncertainty. As Schauer observed, “[w]hen we use presumptions and allocate the burden of proof, we attempt to ensure that decisions under uncertainty will be biased away from restriction of those values we hold to be of greatest importance.” If we assume that courts are broadly inclined to underprotect speech under conditions of uncertainty—and there are good reasons to think this might be the case—we might prefer a “defining out” regime that, by default, classifies speech in the highest-value category in order to counteract this bias.
As an initial matter, the basic assumption underlying this argument is an empirical one, and it is the same assumption that underlies the traditional defenses of the current doctrinal framework: that courts are so strongly predisposed to be underprotective of speech interests that a highly prophylactic doctrinal approach is necessary to adequately safeguard such interests. As I discuss in greater detail below, however, some cautious scrutiny of this assumption might be in order. While few will dispute the basic point that some degree of doctrinal prophylaxis is necessary to adequately protect speech, it may be worth considering whether the current doctrine broadly overestimates this risk of systemic underprotection of speech, thus creating a doctrinal regime that is excessively prophylactic in nature.
Putting this issue to the side, however, there are two countervailing practical considerations that ultimately lead me to prefer the “defining in” approach. First, a preference for one approach over the other should be driven by the ease with which courts and other actors can identify and define the relevant categories in question, and a category is useful only insofar as it identifies a particular subset of speech entitled to a particular degree of protection based on the theoretical rationales underlying the First Amendment. Particularly at this stage of the First Amendment’s doctrinal development, it is far easier to identify and carve out distinct speech categories at the extremes of the speech-value spectrum (that is, the highest-value speech and the lowest-value speech) as opposed to identifying distinct categories that reside in the hazy middle of the spectrum. Our shared intuitions regarding speech value and harm—which reflect decades (if not centuries) of theoretical and jurisprudential development of the idea of free speech—readily yield the clearest examples of the highest-value speech, such as political speech, truthful news reporting, and artistic expression. While the precise boundaries of these categories are certainly fuzzy and subject to debate, both the core categories themselves and paradigmatic instances within these categories are relatively easy to identify and broadly reflected in both judicial and popular understandings of the First Amendment.
It is a far more difficult undertaking, however, to identify discrete subsets of speech that fall within the murky middle of the speech-value spectrum, simply because such speech, by definition, does not implicate the theoretical rationales behind speech protection in the same obvious and direct manner. This merely reflects the basic truth that core applications of a legal rule or principle are generally more concrete and easier to identify than borderline or “penumbral” applications. Furthermore, as a matter of historical development and doctrinal design, First Amendment doctrine has naturally focused primarily on the highest-value “core” speech and the low-value speech exceptions, as these are the instances that most starkly reflect both our theoretical rationales for protecting speech and the practical necessity of limiting speech protection in certain circumstances.
Thus, not only are middle-value speech categories more difficult to carve out by nature, but they have not been worked through and discussed as thoroughly as high-value or low-value speech categories. It therefore makes practical sense to set middle-value speech and intermediate scrutiny—rather than high-value speech and strict scrutiny—as the default residual category and standard of review, since middle-value speech categories are, by nature, more eclectic and harder to define than categories of the highest-value speech. And courts would presumably have an easier time articulating clear constitutional values and principles in delineating specific high-value speech categories rather than middle-value speech categories, which would have positive effects on both the transparency of First Amendment doctrine and the doctrine’s capacity to evolve in light of rapidly changing cultural and technological conditions.
Second, although worries about systemic underprotection of speech might theoretically favor an approach that defines speech “out” of a default high-value speech category, there are strong indications that the current default rule is simply too strong, and that merely exhorting courts to more actively carve out middle-value speech exceptions may not bring about meaningful change. As I discussed above, the doctrinal distortion associated with the current doctrinal framework is rooted, to a significant extent, in courts’ broad reluctance to carve out categorical exceptions to the default strict scrutiny rule when warranted by the circumstances—a reluctance that appeared to be ingrained even before the Court’s recent decisions explicitly limiting courts’ capacity to create such exceptions. Thus, as a practical matter, even if any formal doctrinal obstacles to carving out middle-value speech exceptions were removed, courts might still be highly reluctant to do so, resulting in little meaningful change.
On the other hand, if the doctrinal framework were adjusted so that intermediate scrutiny—rather than strict scrutiny—represents the default standard unless speech is otherwise entitled to special treatment, courts would presumably be more willing to apply it in difficult middle-value speech cases. And as I discuss in detail below, to the extent we may be concerned with systemic underprotection of speech given this greater degree of judicial discretion and the potential inertial pull of intermediate scrutiny as a default standard, courts can include within the doctrinal framework a series of prophylactic rules or principles designed to limit the risk of such underprotection; for example, they might openly adopt the broad principle that in all close cases of speech categorization, speech should be categorized as high-value rather than middle-value. In other words, courts, in designing the doctrine, can incorporate measures that would mitigate ex ante the risk that the inertial pull of an intermediate scrutiny default rule will be too strong.
B. The Intermediate Scrutiny Standard
Again, under my proposal, if the regulated speech in question cannot be categorized as either low-value or high-value in nature, then content-based regulations of such speech should be evaluated under intermediate scrutiny rather than strict scrutiny by default.
To be clear, I am not arguing that intermediate scrutiny must apply in all middle-value speech cases dealing with content-based regulations. As I noted above, there are many factors that play a role in evaluating speech regulations beyond the value of the regulated speech in question, and these factors might dictate a more stringent standard of review in some cases. Most notably, when the government adopts viewpoint-based distinctions in regulating middle-value speech, it makes sense to apply strict scrutiny rather than intermediate scrutiny, since such regulatory approaches represent direct government distortion of the marketplace of ideas and thus implicate most clearly the significant dangers of government abuse. Indeed, as the Court held in R.A.V., viewpoint-based restrictions of even the lowest-value speech can be subject to the most stringent First Amendment scrutiny, so it follows naturally that strict scrutiny similarly ought to apply to such restrictions of middle-value speech. Apart from cases dealing with viewpoint discrimination, however, I broadly propose that intermediate scrutiny be the default standard applicable to content-based regulations of middle-value speech.
The next question is what this default intermediate scrutiny standard ought to look like. As Bhagwat has catalogued in an insightful and comprehensive article, intermediate scrutiny has emerged in a wide variety of areas within First Amendment doctrine, such as in evaluating content-neutral regulations, symbolic speech, and speech of government employees; as such, the precise formulation of the standard has varied based on the context. But the most relevant formulation, which I will adopt for present purposes, is the Central Hudson test for evaluating content-based regulations of truthful commercial speech—again, the only substantive category of speech that the Court has explicitly recognized as middle-value in nature. Under this test, for the regulation to survive, (1) the government interest must be “substantial,” (2) the regulation must “directly advance the governmental interest asserted,” and (3) it must do so in a way that is “not more extensive than is necessary to serve that interest.”
As many have noted, the essence of all intermediate scrutiny tests like the Central Hudson test is balancing. On one side of the balance is the constitutional value of the speech, and this value judgment is made at the outset, through the selection of the appropriate standard of review. Under my proposal, the first inquiry is whether the speech in question falls into a category of either high-value or low-value speech; if so, then the appropriately stringent or deferential standard of review is applied. If the speech cannot be carved out in this manner, it is, by default, classified as middle-value speech to which intermediate scrutiny applies. Adopting this standard of review to middle-value speech is, in effect, an acknowledgement of some meaningful degree of constitutional value—but a value that is less substantial than that of the highest-value speech, such that it is subject to open balancing against the government’s regulatory interests.
Once the middle-value nature of the speech has been established, the three prongs of the Central Hudson test effectively balance that value against the government’s regulatory interests in light of the particular manner by which the government chooses to regulate. The first prong represents a judgment as to whether the government’s regulatory interests are of a sufficient magnitude to outweigh the constitutional value of the speech in question. The second prong works to ensure that the asserted interests are genuine ones that the government is pursuing in good faith, and it acts as a check on the government’s judgments, empirical or otherwise, regarding the efficacy of the regulation. Finally, the third prong works to ensure that the government’s particular regulatory approach is proportional in nature—that the regulation is tailored such that it takes adequate account of the speech-based interests on the other side of the equation.
This intermediate scrutiny test is, of course, highly open-ended and indeterminate in nature, which may raise significant concerns as to administrability, predictability, and the possibility of judicial abuse. I address this potential critique in greater detail below, but broadly speaking, some degree of open-endedness and indeterminacy is to be expected given the inherent difficulty of resolving middle-value speech cases. And if, as current First Amendment jurisprudence suggests, the alternative to this sort of open-ended approach is doctrinal distortion with similar value judgments being made sub rosa, then an approach that at least openly recognizes the sorts of value-based balancing judgments driving courts’ decisions certainly represents a superior state of affairs.
C. The Benefits of the Proposed Approach
My proposed adjustment would greatly mitigate the substantial doctrinal distortion and confusion associated with the current doctrinal framework. As discussed above, under the current strict scrutiny default rule, courts faced with an ill-fitting middle-value speech case are often inclined to distort the doctrine in order to reach the “correct” result, given their general reluctance to reach anomalous results and the doctrinal and practical obstacles to crafting formal exceptions to the rule. If, however, strict scrutiny is limited solely to delineated categories of high-value speech—and intermediate scrutiny applies to all residual middle-value speech—then courts faced with a difficult middle-value speech case would have a far more palatable set of options.
Under this framework, courts would know that the path of least resistance leads to something resembling a true balancing analysis rather than a severe (and potentially ill-fitting) rule of automatic invalidation. This would reduce their incentives to distort doctrine and leave them free to work through difficult middle-value speech cases openly and directly. Thus, for example, the Renton Court need not have relied on the suspect secondary effects doctrine in order to avoid the strict scrutiny default rule. Rather, under my approach, the Court need only find that the speech in question did not fall into a designated category of high-value speech, and on this basis it could apply intermediate scrutiny directly. Similarly, in the Nuremburg Files case, the court could have reached its result by openly articulating, under the intermediate scrutiny standard, the broad judgments of speech value and harm driving its analysis rather than by shoehorning the speech into the ill-fitting category of true threats.
Indeed, in all of the examples discussed above, courts need not have resorted to any sort of doctrinal distortion, at least insofar as such distortion flowed from a strong intuition that the speech in question was, in fact, middle-value rather than high-value in nature. Rather, they would be free to apply the more open-ended intermediate scrutiny standard, which would give them the flexibility to openly grapple with the difficult aspects of these cases rather than surreptitiously manipulate doctrine to arrive at the desired result.
Furthermore, my approach would greatly increase the degree of analytical transparency—and all of the benefits associated with such transparency—at a limited cost. First, and most obviously, it would promote such transparency simply because it would limit the degree of doctrinal distortion, as I described above. When doctrine is distorted, transparency necessarily suffers because courts’ actions and explanations no longer bear a direct relationship to formal doctrine. This gives courts more opportunities to hide the ball in their analyses, which makes it more difficult to identify and critique the underlying foundational value judgments actually driving their decisions. By substantially reducing the need for doctrinal distortion, my approach would help to ensure that courts’ decisions accurately reflect the formal law that applies to them.
In addition, the intermediate scrutiny standard is itself highly transparent, since unlike strict scrutiny or rational basis review, it is a balancing mode of analysis—the only standard of review in which the outcome of the case is not effectively foreordained. Thus, if courts apply intermediate scrutiny as a default in middle-value speech cases, this effectively forces them to confront and openly grapple with foundational questions of speech value and harm—first by articulating exactly why the speech in question does not fall into a designated low-value or high-value exception, then by walking through the three open-ended prongs of the Central Hudson test. In other words, courts would not be able to hide behind formal doctrine; they would have to articulate their underlying judgments as to what makes the speech in question valuable and directly analyze how this value measures up against the government’s countervailing regulatory interests.
This sort of broad discussion of foundational values is particularly valuable in middle-value speech cases, which are by definition the most difficult cases. When clearly high-value speech is in play, it is clear that effectively no content-based regulation will pass muster; when clearly low-value speech is in play, the opposite is true. Middle-value speech cases are the difficult cases in the middle; they are thus the cases that that truly force courts to consider, in a nuanced manner, what makes the speech constitutionally valuable and how that ought to be balanced against the harm associated with the speech. Intermediate scrutiny allows for an open, balancing mode of analysis to deal with these difficult cases.
My proposed approach would further enhance analytical transparency insofar as it would force courts to openly articulate why certain speech ought to be entitled to the most stringent degree of First Amendment protection. Under the current approach, courts are free to subject all content-based restrictions of uncategorized speech to strict scrutiny without much thought or analysis. Under a default rule of intermediate scrutiny, however, courts who wish to apply strict scrutiny in, say, a case dealing with a novel subset of speech must explain why that speech falls within a categorical high-value speech exception to the default rule. Thus, my proposal would ensure that courts cannot apply strict scrutiny blindly; they would have to articulate the basis for its application, just as they would have to articulate their basis for recognizing a new category of low-value speech.
All of this would ultimately produce far more candid, direct, and nuanced analyses of foundational questions of First Amendment value and harm. And as I discussed above, such analytical transparency is particularly vital in spurring the common-law development of First Amendment doctrine amidst rapidly changing cultural and technological conditions. Furthermore, on the other side of the ledger, these substantial gains in doctrinal integrity and analytical transparency would come at a limited cost to case-by-case predictability and consistency. Under my approach, the standards for evaluating content-based regulations of clearly high-value or low-value speech would remain unchanged. My proposed adjustments would ultimately affect only middle-value speech cases—cases that have largely served to undermine rather than promote predictability and consistency within the current doctrine given the substantial degree of doctrinal distortion and confusion associated with them.
IV. Potential Critiques and Practical Considerations
In this Part, I address some of the most likely substantive critiques of my proposed approach, which would presumably center around a perceived reduction in the protection of speech interests or the dangers associated with a heightened degree of judicial discretion. I also address the more practical critique that my approach is simply too radical a shift to be realistically considered by courts.
A. Insufficient Formal Protection of Speech Interests
One might argue that my approach would be insufficiently protective of speech interests on a purely formal level, since setting the default standard to intermediate scrutiny rather than strict scrutiny would necessarily mean that a large swath of speech regulations would now be subject to this more deferential standard of review. Any such argument, however, misunderstands the nature and scope of my proposal. I do not argue here that the middle-value speech classification should necessarily be expansive in nature. The nub of my proposal is merely that intermediate scrutiny, rather than strict scrutiny, serve as the default standard of review for content-based regulations of any residual middle-value speech. The reach of this default intermediate scrutiny standard will ultimately rest on how broadly one delineates the categories of high-value speech: one might define these categories very broadly, leaving only a narrow swath of residual middle-value speech, or one might define them narrowly, leaving a broad expanse of middle-value speech.
Thus, under my approach, courts could certainly construct an aggressively speech-protective doctrine while leaving intermediate scrutiny as the default standard applicable to any uncategorized speech. And the broad benefits of my approach would remain regardless of how large or small the residual category of middle-value speech ends up being. If courts are forced to articulate the basis for classifying speech as high-value—rather than simply assume speech to be high-value by default—then any decision to classify speech as such will, by nature, more transparently reflect the fundamental value judgments and intuitions underlying it. And even if the size of the residual middle-value speech category is relatively small, its position within the doctrinal framework as the default category would make it easier for courts to formally apply intermediate scrutiny—rather than distort doctrine—in those cases where applying the onerous strict scrutiny standard would be most dissonant.
My proposed approach therefore does not necessarily translate to a broad reduction in speech protection across the board. By formulating highly expansive categories of high-value speech, courts could craft a First Amendment doctrine that offers a similar or greater degree of overall speech protection compared to the current doctrine. But they would do so in a far more open and transparent manner, without having to resort to the doctrinal distortion that has infected the present doctrine.
B. Categorical Vagueness and the Risk of Watering Down Protection of High-Value Speech
Another concern might be that establishing intermediate scrutiny as the default rule would lead to slippery slope concerns given the difficulties in cabining the scope of the rule’s application. That is, courts might be tempted to expand the application of intermediate scrutiny very broadly, eventually eroding even the significant protection currently afforded to the highest-value speech. There are strong, foundational reasons why we have concluded that, for example, content-based restrictions on political speech ought to be prohibited in all but the most extreme circumstances. Perhaps shifting the default rule away from strict scrutiny would trigger a broad erosion of speech protection that ultimately unsettles even these most firmly held intuitions regarding the freedom of speech.
In more concrete terms, some degree of vagueness is inevitable in delineating high-value speech categories, and this vagueness creates a risk that courts will, over time, expand the boundaries of middle-value speech until it threatens to engulf even what we today consider core instances of high-value speech. It is relatively straightforward and uncontroversial to say, for example, that political speech and truthful news reporting are categories of high-value speech subject to the most stringent protection. But even if the core instances within each of these categories are clear, the boundaries will be fuzzy and vague—speech, after all, comes in infinite variations and contexts. And if it is in fact true that courts will (either intentionally or unintentionally) regularly favor state regulatory interests or personal predilections over speech interests, then the broad discretion afforded to courts by this categorical vagueness may well lead to systematic and ever-expanding dilution of the substantial protection currently afforded to the highest-value speech.
This argument suggests that my proposed cure would be worse than the disease, and that it might be wiser to forego creating discrete high-value speech categories. As Schauer put it, when a First Amendment category “is so inherently and extremely indeterminate and so linguistically ill-defined, a serious risk exists that the category will in practice be misapplied, and a powerful argument therefore arises against the creation of the category.”
These are legitimate concerns, and they represent a significant source of caution in considering any broad adjustments to the present doctrine. Nevertheless, in implementing my approach, some doctrinal measures can be taken to at least mitigate this risk of dilution. For example, the Supreme Court could articulate specific approaches to category-setting that would limit the degree of discretion afforded to courts: perhaps a strong principle that high-value speech categories will be construed expansively, or that all borderline cases are to be treated as high-value speech cases, or that the creation of new high-value speech categories is both expected and encouraged. If there are indeed systemic risks that courts will tend to erode the protection of speech if afforded greater discretion to make speech-value judgments, there are ways to implement my proposed approach that work to cabin the scope of this discretion.
Furthermore, there are strong reasons to believe that the incremental risks associated with categorical vagueness are outweighed by the benefits of reducing doctrinal distortion and increasing analytical transparency, especially if precautionary doctrinal measures such as those suggested above are implemented. In evaluating the costs and benefits of my proposed approach, the appropriate point of comparison should not be the current approach as it would work in a theoretical vacuum; rather, my approach should be compared to what the current approach has produced in actual practice. And as I described in detail above, the current approach has produced substantial doctrinal distortion, which threatens to destabilize the entire doctrinal structure, and analytical opacity, which limits courts’ ability to critically evaluate the many novel issues of speech regulation arising from our rapidly evolving communications culture. Of course, this is ultimately an exercise of comparing incommensurables: the risks associated with a more open-ended framework on the one hand, and the shortcomings of the present doctrinal approach on the other. But at the very least, it does not seem a foregone conclusion that the current state of affairs is any better than what I propose.
Finally, these slippery–slope concerns are ultimately driven by the same core assumption underlying much of traditional First Amendment theory and doctrine: that courts are so strongly predisposed to be underprotective of speech interests that an extremely prophylactic doctrinal approach is necessary to adequately safeguard such interests. As I suggested earlier, however, it is perhaps worth reevaluating the scope of this basic assumption. Few will disagree, I think, that there are strong reasons to fear conscious or unconscious biases against speech when speech interests are weighed against regulatory interests, and that First Amendment doctrine should take this risk of underprotection into account. But there is the additional question as to whether the current doctrine accurately reflects the actual degree of this risk.
That is to say, even if we all agree that prophylactic doctrinal measures are necessary, we must nevertheless question whether the particular measures adopted either underestimate or overestimate the actual risk of underprotection. After all, overestimation of this risk incurs significant costs just as much as underestimation. If, for example, the Supreme Court were to adopt the (presumably) empirically inaccurate assumption that every single court, in every single case, would always uphold speech restrictions if given the discretion to do so, then it would presumably institute rules that are excessively prophylactic in nature, which would handcuff legislatures from reasonably regulating speech in cases where such regulation is clearly justified.
The actual extent of the risk that courts will systematically underprotect speech is ultimately an unanswerable empirical question, so any judgments on the subject are necessarily speculative. But there are perhaps reasons to believe that the current, highly prophylactic approach overestimates these risks within the present cultural context. As Schauer has observed, the First Amendment carries massive cultural and political force in contemporary society, with a “magnetism” unmatched by other constitutional protections. He states:
To an extent unmatched in a world that often views America’s obsession with free speech as reflecting an insensitive neglect of other important conflicting values, the First Amendment, freedom of speech, and freedom of the press provide considerable rhetorical power and argumentative authority. The individual or group on the side of free speech often seems to believe, and often correctly, that it has secured the upper hand in public debate. The First Amendment not only attracts attention, but also strikes fear in the hearts of many who do not want to be seen as opposing the freedoms it enshrines.
The enormous political and cultural force associated with the First Amendment’s protection of free speech seems undeniable on its face. There is perhaps no constitutional provision that is as universally embraced and praised as the Free Speech Clause, such that—as Schauer colorfully noted—“the First Amendment’s magnetism leads strategic actors to embrace it as easily as politicians embrace motherhood, the flag, and apple pie.”
There is thus perhaps reason to believe that any systemic risk that courts (or the general public) will favor government regulation over conflicting speech interests is less severe today than it may have been in, say, the early-to-mid–twentieth century. As Neil Richards has observed, the sorts of slippery–slope arguments that have often been marshalled to support the current doctrine’s highly prophylactic approach have not actually been borne out in practice. The scope of the traditional low-value speech categories, for example, has shrunk rather than expanded over time; indeed, it is unclear whether the traditional low-value category of fighting words remains viable. Furthermore, the long-term trajectory of commercial speech doctrine has been a broad shift from less protection to greater protection. Truthful commercial speech has evolved from unprotected speech in Valentine v. Chrestensen to partially protected speech under the Central Hudson framework, and the Court’s recent opinion in Sorrell v. IMS Health Inc. suggests that full constitutional protection of such speech may be imminent. Thus, as Richards concluded, “the principal theoretical and practical difficulty” in defining the limits of First Amendment protection might actually be ensuring the adequate protection of non-speech interests “under a juridical regime in which free speech always wins.”
I want to tread lightly here, however, as I am fully cognizant of the risks of making these sorts of unprovable judgments. It is certainly dangerous to assume that either the past trajectory of doctrinal development or currently prevailing cultural conditions will necessarily continue into the future. Cultural attitudes towards free speech—and any doctrine associated with them—might shift quickly in times of “pathological” stress, such as wartime. I therefore do not suggest here that it is unreasonable to take doctrinal approaches that seem overly protective of speech given currently prevailing cultural conditions. What I do want to emphasize, however, is that the degree of such prophylaxis can be excessive, and it is worth considering the extent to which the severe slippery–slope concerns that have driven the theoretical development of First Amendment doctrine—the same sorts of concerns that might be raised against my proposed approach—actually hold true.
C. Excessive Judicial Discretion in Intermediate Scrutiny Analysis
Similarly, one might argue that even if the scope of the rule’s application could be effectively cabined, intermediate scrutiny analysis itself is simply too open-ended and discretionary in nature, and it will produce excessive unpredictability, inconsistency, and potential for judicial abuse. As stated above, intermediate scrutiny is, at its heart, a balancing mode of analysis. And as Bhagwat has argued, “intermediate scrutiny doctrine, as articulated by the Supreme Court, does not provide any guidance on how such assessments should be made, thereby eliminating any hope that the Court can assert control over (and consistency among) appellate courts applying its precedents.” This open-endedness, he argues, has led to a failure amongst lower courts to systematically account for relevant factors such as the particular social value and regulatory needs associated with different types of speech, causing courts to “systematically overprotect speech in some contexts and underprotect it in others.”
The amorphous nature of the intermediate scrutiny standard might thus raise the same broad concerns generally associated with balancing approaches. In the speech context, such an open, discretionary standard might lead to unpredictable and inconsistent results, as well as intentional or unintentional errors in judicial judgment, all of which may produce substantial chilling effects on protected speech. These are certainly legitimate and important concerns to be considered any time some form of balancing is proposed within the First Amendment context. But there are, I think, strong reasons to believe that these concerns can be mitigated within my proposed framework.
As an initial matter, the magnitude of these concerns would be directly related to the scope of application of the intermediate scrutiny standard—that is, the breadth of speech deemed to be middle-value in nature. Thus, to the extent we are concerned with granting courts the sort of discretion inherent to intermediate scrutiny analysis, we might choose to define the categories of high-value speech extremely broadly, leaving only a very narrow swath of residual middle-value speech. Under such a doctrinal structure, courts would apply the open-ended intermediate scrutiny standard in only a limited subset of middle-value speech cases, which would mitigate any concerns associated with excessive judicial discretion. And as I noted above, such a structure would not blunt the efficacy of my proposal, since the default nature of the intermediate scrutiny standard would still provide courts a more palatable doctrinal option in cases where applying strict scrutiny would be highly anomalous or dissonant.
Furthermore, the sort of intermediate scrutiny analysis I envision need not—and should not—be a form of ad hoc, case-by-case balancing. One of the great lessons gleaned from the development of modern First Amendment doctrine is the practical necessity of a broader frame of reference when balancing speech value against harm, most notably reflected in the Court’s adoption of categorical balancing rather than ad hoc balancing in evaluating low-value speech. Such an approach produces a more manageable and administrable doctrinal regime that mitigates the unpredictability, inconsistency, and potential for abuse associated with ad hoc balancing.
Thus, while a broad intermediate-scrutiny standard might apply to all middle-value speech in the abstract, this standard should evolve to apply in a more specific, tailored manner in different contexts. That is, different variations and more concrete standards should emerge based on the type of speech involved, the particular method of regulation used, and so forth. This echoes Bhagwat’s overarching “call for disaggregation” in the intermediate scrutiny standard as applied in speech cases. As Bhagwat observed, the more that courts apply a broad, one-size-fits-all intermediate scrutiny standard to a wide variety of different First Amendment problems, the higher the likelihood that the standard devolves into a vague and open-ended “doctrinal mush.”
In the context of middle-value speech, some degree of doctrinal partitioning within the broad intermediate scrutiny framework is sensible, given the varied and eclectic nature of the speech that can fall within this broad classification. For example, truthful commercial speech, public disclosures of embarrassing private facts, and false statements of fact might all sensibly be deemed categories of middle-value speech, but the underlying bases for classifying each category as such are distinct. And as Bhagwat noted, this segmentation within the broad intermediate scrutiny standard would have salutary effects on doctrinal transparency, as it would lead courts to “articulate standards regarding what kinds of speech, and what kinds of regulatory interests, should be accorded more or less weight (or indeed, any weight at all) in each of the different areas of law.” In the end, this is simply the sort of common law development that is de rigueur in developing practical doctrine within First Amendment jurisprudence—which, after all, is ultimately built on a sparse textual provision with a fuzzy historical record. Although the starting point for evaluating middle-value speech may be the general intermediate scrutiny standard, more tailored standards applicable to specific subsets of such speech can and should emerge as courts accumulate experience and data in evaluating a wide range of cases.
That being said, even if implemented in the way I propose, my approach may well sacrifice some predictability and consistency in the doctrine, at least on a formal level. But that, of course, is the basic tradeoff between rule-like approaches and standard-like approaches. The more standard-like balancing mode of intermediate scrutiny, by its very nature, offers more case-by-case equity and flexibility at the cost of predictability and consistency. And my strong sense is that the gains associated with the greater flexibility and openness of this standard would offset any losses associated with diminished administrability.
As I discussed above, courts are already in the practice of ignoring the formally applicable strict scrutiny default rule in difficult middle-value speech cases, choosing instead to distort doctrine to reach their desired results. Thus, there is little value to be lost in introducing intermediate scrutiny to these cases. My proposal merely seeks to bring the value-based judgments underlying these existing doctrinal distortions out into the open, which is certainly a superior state of affairs to the current one.
To be sure, my proposals in this section assume some capacity on the part of the Supreme Court to institute meaningful clarity and uniformity within the doctrine: to articulate clear and direct judgments of speech value and harm in different circumstances, to translate such judgments into practically administrable doctrine, and to police lower courts that are making these same sorts of determinations. In practice, however, this might ultimately look messy and unwieldy. But even if that is the case, messy, unpredictable, and transparent is certainly better than messy, unpredictable, and opaque, which is the current state of affairs.
D. Loosening Constraints on Local Officials
Perhaps my proposal would undermine efforts to constrain not just courts and legislatures, but also what Seth Kreimer calls “village tyrants”—the “local officials and street level bureaucrats” who make speech-related enforcement decisions against “village Hampdens,” or local dissidents. Kreimer observed that these sorts of local enforcement cases represent a significant proportion of instances in which the content-neutrality rules are actually applied, yet they rarely reach court at all, let alone the Supreme Court. As a result, he argues that the current, highly prophylactic two-tier approach is necessary to adequately protect local dissidents, since such an approach “is well adapted to counterbalance predictable cognitive biases that warp judgment when potential village tyrant confronts aspiring village Hampden,” and “[i]t is easy enough to write an administrative manual that prohibits treating speakers differently because of what they say.”
My proposed approach, however, would keep any such constraints largely intact. As an initial matter, the actual degree of formal constraint that the current doctrine imposes on village tyrants might be limited, given the substantial doctrinal distortion associated with the current framework. But in any case, the onerous strict scrutiny standard would still apply to all high-value speech, which would include the broad categories of speech most likely to be at the center of local disputes between village tyrants and village Hampdens, such as political speech, truthful news reporting, or artistic expression. Strict scrutiny would also still apply to all viewpoint-based restrictions of middle-value speech. Thus, in these vital contexts, village tyrants would be constrained by the same strong and clear rules.
This is not to say that such constraints will necessarily remain just as robust under my approach; village tyrants may well perceive incrementally greater opportunities to suppress speech given the vagaries associated with categorizing high-value speech. But these increased opportunities for speech suppression would presumably arise not in the most vital cases dealing with clearly high-value speech, but rather in borderline cases dealing with middle-value speech or speech at the fringes of the high-value speech categories. Furthermore, as I discussed above, this sort of discretion produced by vagueness can be mitigated ex ante by adopting doctrinal presumptions, and any risks associated with any incremental discretion afforded to courts or local officials under my proposal are likely to be outweighed by the benefits of a better-fitting, less distorted, and more transparent doctrine.
E. The Practical Likelihood that Courts Will Adopt the Approach
On its face, my approach might appear somewhat radical. After all, the default rule of strict scrutiny has been entrenched for well over forty years, and courts have applied the rule—or at least paid lip service to it—on countless occasions. So even if one were to agree with my proposed approach, how realistic is it to expect that courts would adopt it?
My proposal, however, is far less radical than it may appear. The broad idea that there are gradations of speech value beyond the binary low-value/high-value framework is ingrained throughout First Amendment jurisprudence. The Court’s treatment of truthful commercial speech as a form of middle-value speech subject to intermediate scrutiny is the most obvious example of this. But beyond that, the Court has often singled out political speech as uniquely valuable speech that resides at the core of First Amendment protection; similarly, it has explicitly recognized, in various contexts, that speech on matters of public concern carries greater constitutional value than speech on matters of private concern. Thus, my proposal can be seen as a means of capturing and formalizing basic intuitions regarding gradations of speech value that the Court has already recognized to a significant extent.
Furthermore, as I discussed above, courts have often found ways to distort the existing doctrinal framework to avoid the ramifications of the strict scrutiny default rule. So on a practical level, my approach does not represent any sort of radical change in the doctrine. It merely involves openly recognizing and formalizing something that courts have already been doing implicitly. In other words, courts have already effectively recognized middle-value speech cases and subjected them to special treatment distinct from cases dealing with clearly high-value speech; they have just done so sub rosa.
Indeed, there appears to be a substantial appetite amongst some members of the current Court to revise the traditional default rule of strict scrutiny. Justice Breyer has long been highly vocal in his displeasure with a rigid strict scrutiny framework—for example, in his opinion in Reed, he argued that “content discrimination, while helping courts to identify unconstitutional suppression of expression, cannot and should not always trigger strict scrutiny.” And in recent cases, both Justice Ginsburg and Justice Kagan have similarly evinced an inclination to soften the traditional rule: in Williams-Yulee, Justice Ginsburg argued that strict scrutiny should not apply to the clearly content-based Florida Bar rule in question, and Justice Kagan rejected the application of strict scrutiny in Reed, observing that “[w]e can administer our content-regulation doctrine with a dose of common sense, so as to leave standing laws that in no way implicate its intended function.”
Thus, my proposal is not a radical departure from what courts have already been doing or from the Court’s existing approach to questions of speech value in other doctrinal contexts, and it reflects the same frustrations with the current strict scrutiny default rule shared by a number of Justices currently on the Court. So while it might perhaps be unrealistic to expect the Court to, say, suddenly adopt this exact proposal in one fell swoop, it seems at least plausible to envision a gradual recognition by a majority of the Court, perhaps over the span of many cases, that the traditional strict scrutiny default rule is simply too ill-fitting to function as the backbone of a workable First Amendment doctrinal framework. And should the Court ever arrive at this moment of recognition, my proposed approach would represent a relatively simple and broadly effective means for the Court to address the problems associated with the current doctrine.
My proposal here is not rooted in any deep-seated sense that the theoretical underpinnings of the existing doctrinal framework are flawed or unpersuasive. The strict scrutiny default rule is easy to administer and highly speech-protective, and as a theoretical matter, this approach makes sense within the First Amendment context. Theory, however, often translates imperfectly to reality. The benefits of predictability, consistency, and judicial constraint associated with the current rule extend only insofar as courts actually adhere to it, and as I have discussed, courts—when confronted with the significant strain produced by a lack of fit between the value of speech and the onerous strict scrutiny rule—have often chosen to distort doctrine to avoid application of the default rule.
Nor does my proposal reflect some underlying belief that the current framework was necessarily infirm from its inception. The strict scrutiny default rule made sense within the historical context in which it was developed—an era dominated by issues surrounding government regulation of the highest-value ideological speech, which stands at the core of the varying theoretical rationales underlying the First Amendment’s protection of free speech. But the present-day scope of the First Amendment’s coverage is far more expansive and eclectic than what could have been imagined in the formative years of modern First Amendment jurisprudence, particularly as rapid technological and cultural changes raise novel issues that are far afield from the regulation of core ideological speech.
My proposed approach simply reflects the practical realities of the First Amendment we actually have today. It reflects the lessons we have learned from courts’ actual practices in applying (and misapplying) the existing doctrinal framework, and it accounts for the present reality of a far more expansive First Amendment jurisprudence—one that must be better equipped to navigate the significant uncharted realms of potential middle-value speech that courts will inevitably encounter in the near future.