Civil liberties are not absolute—there are always limitations. Regarding the First Amendment, the Court has repeatedly established limitations on free speech in cases where the harm inflicted as a result of the speech is perceived to outweigh the merits of protection. Indeed, unprotected speech historically has been categorized as such precisely because of its potential to result in some type of tangible societal harm. Most Americans believe that conspiracy theories and misinformation have increased in prominence over time, resulting in a proliferation of dangerous behaviors. Legislators and policy experts have responded by proposing measures to curb the spread, profitability, or legality of conspiracy theories online, using societal harm as the central predicate for regulation. We argue that a new legal framework targeted at reducing the spread of conspiracy theories should only be considered if the following basic conditions are met: 1) “conspiracy theory” can be defined and ideas can be, with minimal error, classified as conspiracy theories; 2) the causal impact of conspiracy theories on unlawful and otherwise dangerous behavior can be empirically demonstrated. Using available social scientific evidence, we argue that neither condition can be met.
As articulated by Justice Brandeis in Whitney v. California (1927), a foundational assumption of First Amendment jurisprudence is that the best remedy for potentially harmful speech, including false and misleading speech, is “more speech, not enforced silence.” This extended Justice Holmes’ “free trade in ideas” model of speech in which the ultimate good is reached when people are free to exchange ideas in a marketplace without fear of government punishment (Nunziato 2018). However, the idea that an unregulated marketplace of ideas leads to the greatest public good has been increasingly challenged as our politics has become more contentious, polarized, and burdened with conspiracy theories that could potentially spread unimpeded through online networks (e.g., Sunstein 2021).
The January 6 Capitol riot provides the most striking example of this current state of affairs: Supporters of the sitting president, believing conspiracy theories about a stolen election (many of which were transmitted through social media), attacked the Capitol to disrupt the certification of the 2020 election. Of course, this is not an isolated incident—believers of conspiracy theories have been linked to numerous instances of societal harm. Supporters of the conspiracy theory-laden QAnon movement have engaged in harassment, kidnappings, domestic terrorism, and killings (Bump 2019). Those exhibiting beliefs in COVID-19 conspiracy theories—of which there are many—refuse social distancing, masking, and vaccination (Romer and Jamieson 2020), allowing the virus to spread unhindered. If conspiracy theories are causing people to engage in violent or otherwise harmful actions, doesn’t the government have the responsibility to prevent those harms by limiting the reach of conspiracy theories?
It is clear that conspiracy theories (and other similarly dubious ideas) are subject to existing jurisprudential doctrine regarding defamation, imminent lawless action, threats, and false statements (Han 2017, 178). Indeed, one could argue with relative ease that at least some conspiracy theories serve no purpose in contributing to the marketplace of ideas, promoting healthy democracy, or aiding in the search for the truth, and that any personal or societal harm stemming from such conspiracy theories outweighs the merits of protecting them. But as with all other forms of speech, circumstances matter, and under current legal frameworks, only particular conspiracy theories—those that fall into one of the categories of low-value speech listed above—will be denied constitutional protection. The result is that most conspiracy theories, even those that are intentional lies, will constitute protected speech.
Anxiety about the role that conspiracy theories have played in recent unlawful and normatively undesirable actions like those described above has prompted some legal scholars to argue that these theories should receive less protection under the First Amendment than they currently do (Sunstein 2021; Han 2017; Hay 2019; Waldman 2018; Schroeder 2019; Thorson and Stohler 2017). Their claim is that existing doctrine is antiquated and unsuited to ameliorating increasingly dire social ills in an era in which ideas can travel farther and faster than ever before.
This also appears to be the position of many policymakers (e.g., Klobuchar 2021). In recent years, the U.S. president and members of Congress have publicly browbeaten social media companies for promoting conspiracy theories (and other dubious ideas) on their platforms, calling for these companies to take “additional steps” and admonishing them for “killing people” (Bose and Culliford 2021). Congress has held hearings addressing the scope of conspiracy theories online, resulting in a number of proposals at the national and state levels to curb this type of potentially harmful speech vis-à-vis content moderation and legal penalties (Walker 2020; Riggleman 2020; Heilweil 2020). For example, Sen. Amy Klobuchar, D-MN, sponsored a bill that would remove the protections afforded by Section 230 of the Communications Decency Act if health misinformation, as defined by the Department of Health and Human Services, were algorithmically promoted by a platform (MacCarthy 2021).
In this paper, we argue that, from a normative perspective, laws restricting the dissemination of conspiracy theories should be permissible only if two conditions can be met: 1) “conspiracy theory” can be specifically defined, and ideas can be, with minimal error, classified as conspiracy theories; 2) the causal impact of conspiracy theories on unlawful and otherwise dangerous behavior can be empirically demonstrated. Satisfying the first condition prevents the limitation of speech based solely on the ideology of the ideas being expressed (i.e., viewpoint discrimination); satisfying the second condition ensures that there exists a reasonable societal interest in preventing the speech.
Drawing on an interdisciplinary body of literature about the basic nature, epistemology, and correlates of beliefs in conspiracy theories, we demonstrate that neither condition can be satisfied. Indeed, the concise definition of “conspiracy theory” is prevented by centuries-old epistemological quandaries, the accurate categorization of ideas as conspiracy theories is prohibited by a combination of definitional challenges and human psychology, and researchers’ ability to either explain or forecast unlawful, dangerous behaviors using the communication of or belief in conspiracy theories is extremely weak. Further, we challenge the premises underwriting the desire to construct a new legal framework for dealing with conspiracy theories. Specifically, we argue that conspiracy theories do not pose greater problems today than in the past, that social media and other new communication technologies have not ushered in an increase in conspiracy theorizing, and that the dangers that do spring from conspiracy theories are most realized when political leaders, rather than private citizens, traffic in them. Finally, we argue that conspiracy theories oftentimes possess the qualities of protected speech; namely, they can and, historically, have promoted democracy and a search for the truth. Not only does this evidence preclude the construction of a new legal framework designed to limit conspiratorial speech but it showcases how other proposals in this vein would capriciously censor ideas based on personal viewpoints, cause a severe chilling effect, ensnare more speech than stated intentions claim, and do little to stymie the harms to be prevented.
On Establishing a New Legal Framework
Defining conspiracy theory
Conspiracy theories span the length of recorded history. They are found in accounts of ancient Rome, medieval Europe, and the Renaissance and were prominent in the American colonies before the separation from Great Britain (Pagán 2020; Zwierlein 2020; Butter 2020). However, the term “conspiracy theory,” referring to a specific category of idea, is a relatively new invention, coined by journalists fusing the words “conspiracy” and “theory” to denote a special type of explanation for events and circumstances in the late 1800s (McKenzie-McHarg 2018). It was not until the latter half of the 20th century that the phrase fully entered in the popular vernacular (McKenzie-McHarg 2020). Despite the seemingly clinical meaning of each constitutive component of “conspiracy theory” (i.e., “conspiracy” and “theory”) the term quickly became understood as more than the mere sum of its parts, and a pejorative at that (Thalmann 2019). Indeed, “conspiracy theory,” as it is usually used today, is not merely a type of explanatory theory, but one deemed to be irrational, morally tainted, and false (Harambam and Aupers 2017; Dentith 2016). (For a detailed interrogation of the definition and application of “conspiracy theory,” see Uscinski and Enders 2022).
Regardless of popular connotations, the community of scholars who study conspiracy theories has failed to agree upon a definition of “conspiracy theory” (Räikkä and Ritola 2020; Imhoff and Lamberty 2020). However, there does exist general agreement that the definition must address both formal characteristics of the idea in question (e.g., does the idea allege a conspiracy?) and an epistemological component about the relative truth, falsity, or more general state of evidence regarding the core claim (e.g., does this idea fail to meet some standard of truth?). On the formal end, for example, there are disagreements about whether or not the conspiratorial claim must implicate “powerful” people (Imhoff and Lamberty 2020) and whether the supposed plot must involve a malign intent (Pigden 1995). While answers to questions about the role of intent and relative power are potentially consequential in classifying ideas as conspiracy theories, they are also not unmanageable for a legal system with specific classes of “conspiracy” and methods for investigating intent. We suspect, therefore, that some consensus on what constitutes an alleged “conspiracy” can be achieved. The more difficult definitional hurdle to clear regards the epistemological component of “conspiracy theory.”
Should “conspiracy theory” refer only to false ideas, ideas that are likely false, or ideas that could be true, but that are currently unevidenced? In many ways, the answer to this question depends on the philosophical lens through which one understands science and (the pursuit of) truth. Regardless of one’s epistemological stance on the truth, evidence—including the highest quality scientific evidence—always changes. Changes to research designs, sampling variability, data quality, measurement, or assumptions can alter expert conclusions, as can the discovery of mistakes and fraud. On a grander scale, scientific paradigms can shift, rendering the theories and findings of previous regimes wholly invalid and inapplicable to the new system of ideas (Kuhn 1970). Thus, science—the act of producing evidence in support of or against a theory—is capable of producing only a probabilistic understanding of the truth at any given point in time, one that may occasionally shift quite radically.
Consider, for example, changing scientific consensuses over the course of the last century about cholesterol, fat and salt intake, and alcohol consumption: Conclusions that were at one time thought to be well-evidenced have been heavily revised, if not altogether abandoned, in light of new evidence produced by technological developments, better measurements, and more sophisticated research designs. If our understanding of truth is inherently tentative, then the truth value of conspiracy theories must also be tentative, regardless of the presence or absence of evidence. Thus, neither a lack of supporting evidence, nor a glut of unsupportive evidence, at a given point in time should be interpreted as the final word.
This exploration of the epistemological underpinnings of the “conspiracy theory” label amounts to more than an academic exercise in the philosophy of science. Real-world conspiracies—about which one may concoct a conspiracy theory—do occur. Richard Nixon and members of his administration did conspire to violate the Constitution and bedrock ground rules of American democracy (Olmsted 2000). The Hoover FBI did engage in a clandestine and illegal effort, referred to as COINTELPRO, to surveil, discredit, and harass individuals and groups it subjectively deemed subversive to the American government (Olmsted 2018). American history is replete with other instances of powerful actors working in the shadows, for their own benefit, in a way that violates the common good. And in each case, questions of and investigations into verifiable conspiracies began as unevidenced conspiracy theories, assuming that by conspiracy theory we mean an unevidenced or as of yet uninvestigated allegation.
Suppose we set aside concerns about paradigm shifts and the changing nature of scientific evidence and assume we should restrict our judgment about the relative truth value of a supposed conspiracy theory to a given point in time: According to the evidence available today, does there appear to be support for conspiracy theory “X”? Even in this scenario, separating fact from fiction can be extremely difficult. For example, philosopher Neil Levy proposed that conspiracies (N.B.: not conspiracy theories) are explanations for events and circumstances that cite a conspiracy as a causal factor and have been found to be likely by appropriate bodies of experts using open and available data and procedures (Levy 2007). Conversely then, a conspiracy theory is an idea that cites a conspiracy as the primary explanation for some event or circumstance, but that has yet to be deemed likely true by appropriate bodies of experts with open data and evidence. Such a standard attempts to limit subjectivity and novice judgment by appealing to “appropriate” experts and evidence—epistemic authority figures. However, the choice of which experts count, how many experts are needed, and what to do in case of expert disagreement are, again, left to subjective choice. Who counts as an “expert” is, therefore, often the subject of disagreement.
Consider, for example, the independent film, “Plandemic: The Hidden Agenda Behind Covid-19.” In May 2020, in the beginning of the COVID-19 pandemic, this film, which was shared millions of times, contained contested claims regarding the origin of COVID-19, the dangers of masks and vaccines, and an alleged conspiracy at the Centers for Disease Control and Prevention. The star expert of the film was Judy Mikovits, a biomedical research scientist with a Ph.D. in biochemistry from George Washington University. Should Mikovits be counted as an expert? She does, after all, possess appropriate credentials, including peer-reviewed work. Yet this film was widely criticized by government officials and subsequently banned by social media companies. This episode prompts further questions: How much evidence is enough evidence for a claim to be considered true, or to at least not face censorship? And what type of evidence counts as evidence? How many experts must agree? Must it be that 95 percent of seemingly “qualified” experts agree? 80 percent? 50 percent? Anthropogenic climate change is often considered an unassailable scientific fact given that 97 percent of climate experts agree climate change is unfolding (Cook et al. 2016), and social media companies are currently limiting posts that appear to question the prevailing consensus (Spangler 2021); but if only 70 percent of experts agreed that anthropogenic climate change is occurring, would it be any less true, and would we treat contradictory claims any differently? In short, even standards that explicitly rely on expert consensus and evidence fail to crisply separate conspiracies from conspiracy theories (e.g., Keeley 1999; Buenting and Taylor 2010; MacCarthy 2021).
To this point, experts can also be biased. Once government policy is involved, experts are rarely impartial, nonpartisan actors. When political elites are accused of engaging in conspiracies, can either co-partisans or out-party actors be trusted as unbiased experts who are primarily motivated by the search for truth? The general unwillingness of First Amendment jurisprudence to engage in viewpoint discrimination suggests that the answer to this question is a resounding “no.” Because issues of truth invariably become matters of partisan debate, determinations of what ideas constitute conspiracy theories (i.e., lacking truth) would be determined by highly biased actors with political agendas. Such a scenario would undermine the very essence of the First Amendment by permitting any speech that violated the sensibilities of powerbrokers to be censored or punished.
The questions and epistemological issues outlined above apply to other standards that scholars have proposed for separating conspiracies from conspiracy theories. For example, one of the most influential studies on the topic suggests that conspiracy theories be considered “unwarranted” when those theories become too large in scope (e.g., the number of conspirators and moving parts involved) to maintain secrecy over a long period of time (Keeley 1999; for an empirical demonstration, see Grimes 2016). The logic is that if a conspiracy is “too large,” then it will fail, be exposed, and be labeled a “conspiracy;” however, if it is “too large” and not exposed, then it is unlikely to be true. We do not dispute the basic logic of this idea, though it hardly helps in defining or classifying conspiracy theories. Indeed, the size standard only determines “warrant,” or how reasonable a conspiratorial claim might be, not truth. Application of this standard is also highly subjective—how many conspirators is “too many”?
This standard and others like it ultimately succumb to the centuries-old epistemological quandaries outlined above, failing to separate true from false and providing only subjective schemes for determining which ideas should be worthy of merit. Without a cogent, tightly parameterized definition of “conspiracy theory,” it is unclear how a new First Amendment legal framework designed to limit the spread of conspiracy theories can be developed and evenhandedly mobilized to classify the wide variety of elements in the marketplace of ideas.
Other issues in categorizing ideas as “conspiracy theories”
We have outlined above a number of reasons why defining “conspiracy theory” is difficult, why scholars have failed to agree upon a definition, and how, because of these issues, categorizing ideas as conspiracy theories is an inherently thorny endeavor. Next, we consider an additional hurdle: human psychology. Repeated polling of beliefs in different conspiracy theories, at various time points, has revealed that most people harbor conspiratorial ideas about the government, scientists, big business, aliens, and religious, racial or political out-groups, or some other topic (Enders et al. 2021b). However, few recognize their own beliefs as conspiracy theories (Douglas, van Prooijen and Sutton 2021); to most people, only others believe in conspiracy theories (Walker 2018).
This is so for several reasons. First and foremost, the “conspiracy theory” label has become a pejorative, used to reference bizarre ideas with little truth value held by tinfoil-hatted loons or deadly extremists (Thalmann 2019). To recognize one’s beliefs as “conspiracy theories” would be to denigrate said beliefs and associate oneself with irrational and potentially dangerous conspiracy theorists. Thus, people rarely refer to their own ideas as such (Wood and Douglas 2013). Relatedly, the “conspiracy theory” label has become a useful weapon in the culture war, a new way for polarized partisans to malign the out-party. Out-party ideas that are rejected by the in-party as false, ideologically or morally incongruent with their own, or otherwise objectionable are frequently labeled “conspiracy theories.” Likewise, conspiracy theories, the ideas, are developed about the out-party in an effort to stall agendas, influence election outcomes, and generally erode the image of competitors. Thus, use of the label “conspiracy theory” and the categorization of ideas as “conspiracy theories” are corrupted by innate human tendencies regarding social pressure, the impact of partisan motivated reasoning and social identity, and the out-group homogeneity bias fueled by polarization.
The combination of political biases and loose definitions of what counts as a conspiracy theory is on full display when political elites and journalists attempt to make distinctions about what counts as a conspiracy theory, which conspiracy theories are sanctioned, who counts as a conspiracy theorist, and which groups are more likely to engage in conspiracies against the rest of us. Just like average Americans, journalists show a tendency to castigate those conspiracy theories they personally find most bizarre but rarely employ the term to describe ideas they believe or find plausible. For example, while the contention that Hillary Rodham Clinton conspired with Russia is oftentimes referred to as a “conspiracy theory,” the allegation that Donald Trump conspired with Russia to win the 2016 presidential election is infrequently labeled as a conspiracy theory to this day (Hay 2017), even though both literally allude to an (alleged) conspiracy, and a lengthy investigation of the latter theory generated little supportive evidence. Even mainstream news outlets are not immune to conspiracy theorizing, especially in the absence of clear journalistic norms about which alleged conspiracies should be investigated or covered. Consider, for example, the 1996 “Dark Alliance” reports by journalist Gary Webb, who alleged a conspiracy by the CIA to distribute drugs into African American neighborhoods (Bratich 2004). This story was “[d]erided by some as conspiracy theory and heralded by others as investigative reporting at its finest;” to this day, the underlying truth remains disputed (Devereaux 2014).
Like journalists, politicians have historically used the term conspiracy theory unevenly. Political leaders are often quick to label unflattering accusations hurled in their direction as conspiracy theories. For example, former New Jersey Gov. Chris Christie’s administration dismissed accusations of wrongdoing in what became known as “Bridgegate” as “conspiracy theories;” yet evidence substantiating many of the accusations was subsequently produced (Benen 2014). If political leaders, including nonpartisan judges, were allowed to determine which ideas are conspiracy theories and therefore “unprotected” by the First Amendment, then the definition of conspiracy theory and its application would likely become ostensibly tethered to the ideas one does not hold, especially about out-groups. This would certainly run counter to the First Amendment’s core functions of promoting democracy and a search for truth. Moreover, the ambiguity and inconsistency of the practice would surely foster perceptions of bias and further erode trust in political institutions. We see this on a smaller scale with respect to public opinion on fact checking, the labeling of supposed misinformation, and censorship of particular ideas on social media platforms—the most conspiratorial elements of the mass public notice the subjective nature of these actions, which provides fuel to the fire of conspiracism, distrust, and perhaps even extremist action and causes them to move underground to other platforms where there is less oversight and ability to monitor and track potential threats (Soave 2021).
Do conspiracy theories and misinformation cause harm?
In addition to being able to define and properly categorize ideas as conspiracy theories, we argue there should exist sufficient evidence for the causal impact of conspiracy theories on unlawful and dangerous behaviors to justify the creation of a new First Amendment legal framework designed around conspiracy theories. Beliefs in some conspiracy theories are, indeed, associated with a range of undesirable attitudes and personality traits including (but not limited to): narcissism, Machiavellianism, psychopathy, authoritarianism, anomie, a predisposition toward physical conflict, and support for political violence (Enders et al. 2021b). Conspiracy theories are also associated with the January 6 Capitol attack, as well as a host of idiosyncratic events involving pizza shops, kidnappings, killings, and train derailments.While these associations are undeniable, the assumption that conspiracy theories uniquely motivate behaviors or alter attitudes in a causal fashion currently lacks robust evidence (Uscinski et al. 2022b).
Decades’ worth of public opinion and psychological research demonstrates that specific beliefs are the downstream product of higher order, abstract values and principles, as well as other motivations and characteristics, such as personality and social identities (Leeper and Slothuus 2014; Converse 2006; Kinder 1998; Lazarsfeld, Berelson, and Gaudet 1948). In this light, specific beliefs are best thought of as markers for more stable, foundational traits, worldviews, ideologies, and identities instead of causal first movers or even causal factors responsible for other downstream attitudes and behaviors (Campbell et al. 1980; Douglas et al. 2019; Zaller 1992; Uscinski et al. 2022b). For example, while the economy may be an important consideration in the minds of American citizens as elections approach, their view of the health of the economy and who is responsible for it is colored by partisan identities and political ideology—the latter orientations guide the formation of specific beliefs (e.g., about the economy) and actions (e.g., vote choice). The same process is operational for beliefs in conspiracy theories—while believing in conspiracy theories may be a marker for particular worldviews and orientations of concern, they are not best thought of as the linchpin on which unlawful or dangerous behaviors depend (Uscinski et al. 2022b).
There are two important implications of this well-evidenced process of opinion formation. While conspiracy beliefs are associated with a bevy of anti-social personality traits and normatively troubling attitudes, they certainly do not cause them. Rather, these characteristics encourage both conspiratorial thinking and illegal and dangerous behaviors, perhaps even independent of each other. In other words, conspiracy theories are a symptom, rather than a cause of the factors that could lead to undesirable behavior, and the communication of conspiracy theories (i.e., speech) is not determinative of who will act on those theories (Uscinski, Klofstad, and Atkinson 2016). We showcase both the popularly assumed causal pathway (A) and the model suggested by social scientific evidence (B) in Figure 1. In Model A, exposure to conspiracy theories, perhaps vis-à-vis social media, leads one to adopt conspiracy theories, which subsequently encourage illegal, dangerous behavior. In Model B, particular political, social, and psychological predispositions and motivations encourage conspiracy theories and illegal, dangerous behavior. Note that there is both a direct link between predispositions and behavior and a plausible indirect link through beliefs in conspiracy theories; conspiracy beliefs are unlikely to cause behaviors on their own (Uscinski et al. 2022b).
The second implication regards the role of the internet and social media. Popular narratives hold that social media is rife with conspiracy theories (and other dubious ideas), and anyone could be exposed to those ideas and subsequently lured “down the rabbit hole” of additional conspiracy theories, misinformation, and lies (Collins 2020). However, if we understand conspiracy beliefs to be the product of deep-seated dispositions and worldviews, this narrative crumbles. Indeed, communication researchers and public opinion experts alike agree that information is sorted, processed, and either accepted or rejected vis-à-vis the process of selective exposure (avoidance)—the process of intentionally seeking out (avoiding) information that is congenial (uncongenial) to one’s preexisting worldviews, ideology, or values (Stroud 2010). For example, Democrats and liberals intentionally avoid Fox News, which is widely understood to provide a right-leaning point of view, just as Republicans and conservatives avoid MSNBC, which provides left-leaning content (Iyengar and Hahn 2009). The process is the same for conspiracy theories: Those who possess the types of worldviews congenial to conspiracy theories tend to seek them out, or at least not actively avoid them; those whose worldviews are not hospitable to conspiracy theories avoid them and ignore them when they are exposed (Enders et al. 2021a; Bessi et al. 2015).
Figure 1: Theoretical models of the relationship between (exposure to) conspiracy theories and illegal, dangerous behavior. A) model assumed by journalists and politicians; B) model suggested by empirical social scientific evidence.
Thus, relatively nonconspiratorial people are generally not unwittingly lured down rabbit holes because incidental exposure—whereby idiosyncratic interaction with information (e.g., a crazy uncle’s post on Facebook) influences the formation of beliefs—is not a supported explanation of how information impacts beliefs or actions. If individuals are hardwired to reject information that is not congenial with preexisting beliefs, worldviews, and values, then online conspiracy theories (or on any medium, for that matter) can hardly be understood to pose a societal danger. Indeed, the strength of the relationship between beliefs in specific conspiracy theories and exposure to conspiracy theories via social media use is heavily conditional—if one is not predisposed to interpreting the world through a lens of conspiracy, they are unaffected by exposure to conspiracy theories (Klofstad et al. 2019; Enders et al. 2021a; Uscinski et al. 2016).
This pattern is found in recent studies of social media behavior, as well—interaction with conspiratorial content is much more likely to be the result of intentional subscriptions to such content than “rabbit hole” processes whereby one is repeatedly exposed to unwanted conspiratorial content due to an algorithm (Chen et al. 2022).More broadly, people with extremist views and toxic personalities are likely to seek out compatible content because they are toxic, extremist people, not because they were led blindly by a recommendation system (Chen et al. 2021; Kim et al. 2021). Therefore, the availability of conspiracy theories in social or political discourse—online or offline—has little inherent correspondence with either beliefs in conspiracy theories or the undesirable actions associated with them, as demonstrated by recent work examining beliefs in dozens of conspiracy theories before and during the social media era (Uscinski et al. 2022a; e.g., Mancosu and Vassallo 2022; Romer and Jamieson 2020). This also challenges the popular idea that conspiracy theories spread unimpeded through online networks: Evidence suggests that conspiracy theories spread to the like-minded (Guess, Nyhan and Reifler 2018; Guess et al. 2020), frequently prompted by political leaders (Benkler et al. 2020), rather than through herding behavior whereby ideas spread and amplify like a snowball rolling downhill (Uscinski, DeWitt, and Atkinson 2018). These findings hold for misinformation more generally, with scholars finding that online misinformation has “minimal” effects on beliefs or behaviors (Kalla and Broockman 2018; Mitchelstein, Matassi, and Boczkowski 2020; Bail et al. 2019).
Of course, one could still wonder whether conspiracy theories are worth inhibiting precisely because of the orientations and worldviews they are associated with: Even if the beliefs are not the causal first movers, they reinforce the types of worldviews and values that do lead to unlawful and dangerous behaviors. In other words, conspiracy beliefs can be markers for both intent and effect. As reasonable as this supposition may seem, it, too, is critically flawed. The first reason why involves the treatment of conspiracy theories as monolithic. Not all conspiracy theories are created alike, and the correlates of conspiracy beliefs depend heavily on the conspiracy theories under consideration. For example, while beliefs in Holocaust denial conspiracy theories are related to “dark triad” personality traits (e.g., psychopathy) and support for political violence, conspiracy beliefs regarding climate change, Donald Trump’s relationship with Russia, and the birthplace of Barack Obama are not; instead, these beliefs are most strongly related to run-of-the-mill partisan identities and political ideologies (Enders et al. 2021b). If the intent and potential effects of conspiracy theories require evaluation on a case-by-case basis, it strikes us that the current First Amendment legal framework is already well calibrated for such a task.
Finally, we must emphasize a critical lesson from decade’s worth of research on terrorism and political extremism inside and outside of academia: Forecasting, with any accuracy, the types of extremist activities that provide the greatest cause for worry using specific beliefs and characteristics is ostensibly impossible. It is easy to see conspiracy beliefs and related orientations once wrongdoing has unfolded and a perpetrator has been caught (akin to the way law enforcement and media scour the social media accounts of school shooters after tragedy has unfolded). However, most Americans harbor conspiratorial sentiments of some type; at the same time, most Americans do not engage in damaging behaviors, irrespective of their beliefs. Explanation is not a synonym for prediction.
The events of January 6 provide a clear example. Polls fielded after the 2020 presidential election estimated that between 50 percent-80 percent of Republicans—the equivalent of tens of millions of Americans—believed that the election was illegitimate or had been fraudulently stolen (Bump 2020; Arceneaux and Truex 2022). But only a tiny fraction of those believers were present for the “Save America” rally, and an even smaller fraction participated in the breach of the Capitol. Likewise, while a majority of Democrats believed that President Trump was engaged in a crooked scheme with Russia to steal the 2016 presidential election (Enders et al. 2021b), only one believer in such an idea terrorized congressional Republicans during baseball practice (Lynch 2017).
In sum, we cannot establish a clear link between conspiracy theories and unlawful or dangerous behaviors, exposure to conspiracy theories does not inherently promote conspiracy beliefs, let alone the actions supposedly encouraged by such beliefs, and extremist behaviors are so idiosyncratic and situational that even a combination of specific beliefs and other, more stable characteristics are not sufficiently predictive to result in useful forecasting. Under these conditions, we argue that creation of a new First Amendment legal framework to better address conspiracy theories and related ideas (e.g., misinformation, “fake news”) is unnecessary and untenable.
From the Perspective of Existing Legal Frameworks
The two-level speech theory upon which most First Amendment legal doctrines rest, at least informally, holds that speech should be either protected or unprotected depending on the extent to which it does (or does not) present value to the First Amendment’s core functions of promoting democracy and the discovery of truth (Bork 1971; Mill 1869) or results in some social or personal harm (Chen and Marceau 2015). We have argued above that conspiracy theories and misinformation do not regularly, directly cause personal or societal harm—this is one count on which the conditions necessary for constructing a new legal framework for evaluating conspiratorial speech are not met. However, we also argue here—in the context of existing legal doctrine—that conspiracy theories oftentimes carry democratic and truth value.
As discussed above, determining which ideas are “conspiracy theories,” rather than verifiable “conspiracies,” is no easy task—this depends on decisions regarding who counts as an expert, what counts as evidence, and, ultimately, what is deemed “true.” It is important to recognize that these decisions are made in the course of investigating criminal allegations of conspiracy (such as conspiracy to commit murder). In other words, all verifiable conspiracies start as conspiracy theories, just as all scientific theories start as testable hypotheses; some hypotheses may be more reasonable than others, just as some conspiracy theories may seem more reasonable than others, but the critical similarity is that they both begin as speculative propositions. If all conspiracies start as conspiracy theories, and real-world conspiracies do unfold, it is apparent how a legal framework designed to depress speech about conspiracy theories will also hamper the discovery of true conspiracies. Failing to protect most conspiracy theories under the First Amendment could run afoul of the amendment’s core functions by hampering the search for the truth. Likewise, conspiracy theories oftentimes possess real truth value on their own—conspiracy theories are not inherently false, but rather unsupported (at a given point in time) by evidence.
We also argue that conspiracy theories have the potential to promote democracy, especially by encouraging governmental transparency. For example, conspiracy theories about the assassination of President Kennedy motivated activists to call for the release of classified investigation material; similar demands in the aftermath of the 9/11 terror attacks surely contributed to the formation of the 9/11 Commission (Olmsted 2009). In both instances, details about who perpetrated devastating attacks, their motives, and the failures of the government to prevent the attacks came to light, presumably influencing policy and procedure for the better.
More abstractly, those who exhibit no suspicion or distrust of policymakers and other leaders are, in many ways, failing in their civic duties in the same way as those who interpret all salient events and circumstances as the product of real-world conspiracies (Moore 2018). Citizens of democracies should operate with a healthy suspicion of the motives and clandestine actions of their leaders who certainly stray from their elected duties and ethical commitments from time to time (Locke 1967). To signal to an already distrustful mass public exhibiting low levels of civic engagement that there is no place for conspiracy theories about powerful actors and institutions in a democratic society would be to further erode trust and dissuade individuals from their civic responsibilities.
Nonjudicial Solutions to the Conspiracy Theory Problem
Even though conspiracy theories do not appear to be on the rise, and the direct, causal relationship between such ideas and unlawful and dangerous behaviors is suspect, the societal problem posed by conspiracy theories likely outweighs any societal or truth value in at least some circumstances. Instead of devising a new legal framework for curbing conspiracy theories generally, plenty of nonjudicial options exist.
Conspiracy theories are part and parcel of the human condition—they are a product of innate political and social motivations and psychological biases. It stands to reason then, as scholars of history and philosophy know well, that conspiracy theories are hardly new types of speech. Yet only recently have policymakers, journalists, and researchers voiced concern about the role of conspiracy theories in political discourse and society, more generally. Why? We suspect the meteoric rise in concern in the U.S. can be traced to a few focusing events, namely the COVID-19 pandemic and the Donald Trump presidency.
As for the former event, we note that pandemics such as the COVID-19 pandemic are, historically, once in a century (or longer) phenomena—therefore, altering existing institutions and norms based on the possibility of an idiosyncratic event strikes us as reactionary and myopic. Moreover, conspiracy theories regarding COVID-19 flourished precisely because the roots of conspiracy theories are in the human condition. The pandemic activated the fears and anxieties of an already suspicious and distrustful mass public, leading to the construction and spread of conspiracy theories about the motives of the government, the pharmaceutical industry, Big Tech, news media, and doctors, each of which were held in relatively low regard well before the onset of the pandemic (Gallup 2021).
As for the latter event, the tumultuous Trump presidency, we believe more can be done. There is, of course, nothing inherently conspiracy theory-provoking about a democratic election, even though those on the losing side sometimes turn to conspiracy theories in an effort to explain their loss (Uscinski and Parent 2014). Trump was, however, unique in his willingness to break longstanding norms in the U.S. that kept political elites from consistently espousing conspiracy theories, or that prevented candidates who did traffic in such ideas from gaining political power. Trump’s willingness to buck norms, coupled with a hotly polarized political climate where lies, rumors, and conspiracy theories about political opponents were slowly becoming new weapons of the culture war, explain not merely why Trump trafficked in conspiracy theories, but why this tactic was effective.
One method for reducing the impact of conspiracy theories involves researchers, journalists, and policymakers taking more seriously the worldviews, orientations, and conditions that make conspiracy theories attractive. People do not believe in, and certainly are not motivated to action by, ideas that they casually encounter vis-à-vis online incidental exposure. Rather, deep-seated distrust of the political establishment, scientists, and doctors, anti-social personality traits, and group prejudices drive individuals to concoct and seek out conspiratorial explanations for the world around them. Conspiracy theories might not cause distrust as much as distrust causes conspiracy theories. Thus, banning or criminalizing conspiracy theories will do little address the societal challenges typically—albeit erroneously—attributed to them. One need only look to the current political climate of any number of Western democracies with more limited free speech traditions than the U.S. for evidence; conspiracy theories are just as widely believed in political systems where such ideas can be and censored or prosecuted (Walter and Drochon 2020).
Another nonjudicial method for curtailing the negative social outcomes commonly associated with conspiracy theories is to restore norms regarding the proper tone and tenor of democratic discourse. These norms successfully kept the most outlandish conspiracy theories from being directly injected into political culture for decades. Public opinion is, in many ways, a “top-down” phenomenon—it is a reaction to and reflection of elite rhetoric and behavior (Zaller 1992). If political elites inside and outside of government—but especially elected partisan leaders—restored a norm that conspiratorial rhetoric would not be tolerated, conspiracy theories would fade from political discourse. We must ask more of our leaders, not criminalize the types of speech that can frequently be employed by citizens to hold the worst behaviors of the powerful at bay and promote the discovery of wrongdoing.
We argued that two basic conditions must be met in order to justify the creation of a new First Amendment legal framework designed to curb the spread of conspiracy theories: 1) “conspiracy theory” can be specifically defined and ideas can be, with minimal error, classified as conspiracy theories; 2) the causal impact of conspiracy theories on unlawful and otherwise dangerous behavior can be empirically demonstrated. Based on the state of scientific evidence and philosophic reasoning about conspiracy theories, we demonstrated that neither condition is satisfied. Scholars have repeatedly failed to agree on a definition of “conspiracy theory,” in large part because centuries-old questions about the nature of truth and falsity remain unresolved. We also argued that, epistemological issues aside, the systematic categorization of ideas as “conspiracy theories” is plagued by human nature and an ever-changing body of evidence. Finally, we questioned the presumed causal relationship between conspiracy theories, on the hand, and unlawful and otherwise dangerous behaviors, on the other. While conspiracy theories are oftentimes associated with undesirable outcomes, rarely are they uniquely causal or predictive. Although our arguments here focus specifically on conspiracy theories, many of the same lines of interrogation apply to supposed misinformation, disinformation, and fake news in general.
We conclude by questioning the very premise upon which most calls to policy or judicial action regarding the spread of conspiracy theories rests: that conspiracy theories are “spreading” farther than ever before, driving otherwise productive citizens down the post-truth rabbit hole. The simple fact of the matter is that not a single scientific study has demonstrated that beliefs in conspiracy theories and misinformation have, generally, increased over time. In fact, studies tend to show stability or decreases in belief of conspiracy theories (Romer and Jamieson 2020; Mancosu and Vassallo 2022), including our data, some of which is depicted in Figure 2. The left-hand panel depicts the proportion of U.S. adultswho either “agree” or “strongly agree” with the following conspiracy theories, over time: “Coronavirus was purposely created and released by powerful people as part of a conspiracy” and “The threat of coronavirus has been exaggerated by political groups who want to damage President Trump.” The right-hand panel depicts the average score for the “QAnon movement” on a feeling thermometer ranging from 0 (very negative feelings) to 100 (very positive feelings), over time. In neither case do we observe increases in conspiracy beliefs over time, despite the centrality of COVID-19 and QAnon to popular narratives about recent rise in conspiracism.
Figure 2. Changes in proportion of U.S. adults who believe in COVID-19 conspiracy theories over time (left), and changes in the average rating of the “QAnon movement” on a 0–100 feeling thermometer over time (right). Vertical lines represent 95 percent confidence intervals.
While the internet and social media, in particular, offer a mechanism by which conspiracy theories can be communicated with greater ease and speed than in the past, the search for and acceptance of conspiratorial ideas critically rests on consumers harboring conspiracy-congenial worldviews, orientations, and values. In other words, incidental exposure to conspiracy theories is insufficient for changing minds. Moreover, as many of the correlates of beliefs in conspiracy theories (e.g., educational attainment, scientific literacy, and media literacy) have improved with time, along with journalistic standards (e.g., a departure from yellow journalism toward a norm of objectivity) and certain political norms, it is likely that American society is less conspiratorial today than in the days of witch hunts, illuminati panics, Freemason freak-outs, and Red Scares (Walker 2013).
Whereas most popular accounts of the role of conspiracy theories—and misinformation and fake news, for that matter—in promoting undesirable behavior assume a causal association, decade’s worth of social scientific research is more supportive of a view of conspiracy theories as symptoms, of a deep-seated distrust of politicians, of scientists and other experts, and of a political system they do not perceive to serve them well. If conspiracy theories are attractive explanations for already distrusting citizens, criminalizing speech categorized as conspiratorial or false will only exacerbate the very orientations that encouraged the speech in the first place. The solution to the social problem of conspiracy theories lies not in the (in)ability to communicate seemingly unreasonable ideas, but in the institutions, processes, and conditions (e.g., polarization, social and economic inequality) that make such ideas attractive in the first place.
Even though we have limited our discussion to conspiracy theories in particular, we suspect that interest in curtailing the spread of conspiracy theories accompanies similar interests regarding misinformation, disinformation, fake news, and the like—all manner of false claims and lies. These other types of false and misleading ideas are subject to many of the same concerns described above: They are difficult to define and evidence for causal relationships with undesirable behaviors is weak. We also believe that creation of a new, more general First Amendment framework for dealing with lies could be quite dangerous. Because the truth is difficult to define and uncover, any such framework would have to be designed with an expectation of tolerating degrees of truth/falsity or evidence. This has the potential to ensnare a wide variety of speech about gods, angels, demons, divinity, miracles, prophets, ghosts, the afterlife, creationism, Bigfoot, and the Loch Ness monster, many of which are either relatively harmless or provide some societal value but cannot likely meet any scientific burden of proof. For this reason and those outlined above, we urge extreme caution in proceeding down the road of increased censorship, even though conspiracy theories may occasionally be associated with socially undesirable outcomes.
© 2022, Adam M. Enders & Joseph Uscinski.
Cite as: Adam M. Enders & Joseph Uscinski, What’s the Harm?, 22-15 Knight First Amend. Inst. (Dec. 16, 2022), https://knightcolumbia.org/content/whats-the-harm [https://perma.cc/UMU5-TCL7].
Arceneaux, Kevin, and Rory Truex. 2022. “Donald Trump and the Lie.” Perspectives on Politics. 1-17.
Armaly, Miles T., David T. Buckley, and Adam M. Enders. 2022. “Christian Nationalism and Political Violence: Victimhood, Racial Identity, Conspiracy, and Support for the Capitol Attacks.” Political Behavior 44: 937-60.
Bail, Christopher A., Brian Guay, Emily Maloney, Aidan Combs, D. Sunshine Hillygus, Friedolin Merhout, Deen Freelon, and Alexander Volfovsky. 2019. “Assessing the Russian Internet Research Agency’s Impact on the Political Attitudes and Behaviors of American Twitter Users in Late 2017.” Proceedings of the National Academy of Sciences 117(1): 243-50.
Benen, Steve. 2014. “Reflecting on ‘Conspiracy Theories.’” MSNBC. https://www.msnbc.com/rachel-maddow-show/reflecting-conspiracy-theories-msna247541[https://perma.cc/G8FU-UTHY] (2/23/2022).
Benkler, Yochai, Casey Tilton, Bruce Etling, Hal Roberts, Justin Clark, Robert Faris, Jonas Kaiser, and Carolyn Schmitt. 2020. “Mail-in Voter Fraud: Anatomy of a Disinformation Campaign.” Berkman Klein Center for Internet and Society at Harvard University. https://dash.harvard.edu/bitstream/handle/1/37365484/Benkler-etal-Mail-in-Voter-Fraud-Anatomy-of-a-Disinformation-Campaign.pdf?sequence=1&isAllowed=y [https://perma.cc/FW6E-HQA7] (11/10/2022).
Bessi, Alessandro, Mauro Coletto, George Alexandru Davidescu, Antonio Scala, Guido Caldarelli, and Walter Quattrociocchi. 2015. “Science vs Conspiracy: Collective Narratives in the Age of Misinformation.” PLoS ONE 10(2): e0118093.
Bork, Robert H. 1971. “Neutral Principles and Some First Amendment Problems.” Indiana Law Journal 47: 1-35.
Bose, Nandita, and Elizabeth Culliford. 2021. “Biden Says Facebook, Others ‘Killing People’ by Carrying COVID Misinformation.” Reuters. https://www.reuters.com/business/healthcare-pharmaceuticals/white-house-says-facebooks-steps-stop-vaccine-misinformation-are-inadequate-2021-07-16/ (11/13/2021).
Bratich, Jack Zeljko. 2004. “Trust No One (on the Internet): The CIA-Crack-Contra Conspiracy Theory and Professional Journalism.” Television & New Media 5: 109-39.
Buenting, Joel, and Jason Taylor. 2010. “Conspiracy Theories and Fortuitous Data.” Philosophy of the Social Sciences 40: 567-78.
Bump, Philip. 2020. “More Than 8 in 10 Trump Voters Think Biden’s Win Is Not Legitimate.” The Washington Post. https://www.washingtonpost.com/politics/2020/11/11/more-than-8-in-10-trump-voters-think-bidens-win-is-not-legitimate/ [https://perma.cc/9V4T-YSXP] (6/20/2022).
Bump, Philip. 2019. “The Murder of an Alleged Gangster on Staten Island Loops in an Unexpected Figure: QAnon.” The Washington Post. https://www.washingtonpost.com/politics/2019/03/28/murder-an-alleged-gangster-staten-island-loops-an-unexpected-figure-qanon/ [https://perma.cc/L3T6-GBCM] (7/18/2021).
Butter, Michael. 2020. “Conspiracy Theories in American History.” In Routledge Handbook of Conspiracy Theories, eds. Michael Butter and Peter Knight. New York: Routledge. 648-59.
Campbell, Angus, Philip Converse, Warren Miller, and Donald E. Stokes. 1980. The American Voter. Unabridged Edition. Chicago: The University of Chicago Press.
Chen, Alan K, and Justin Marceau. 2015. “High Value Lies, Ugly Truths, and the First Amendment.” Vanderbilt Law Review 68: 1435-1507.
Chen, Annie Y., Brendan Nyhan, Jason Reifler, Ronald E. Robertson, and Christo Wilson. 2021. “Exposure to Alternative & Extremist Content on YouTube.” Anti-Defamation League. https://www.adl.org/resources/reports/exposure-to-alternative-extremist-content-on-youtube [https://perma.cc/398H-WCGE] (11/10/2022).
Chen, Annie Y., Brendan Nyhan, Jason Reifler, Ronald E. Robertson, and Christo Wilson. 2022. “Subscriptions and External Links Help Drive Resentful Users to Alternative and Extremist YouTube Videos.” https://cpb-us-e1.wpmucdn.com/sites.dartmouth.edu/dist/5/2293/files/2022/04/YouTube.pdf [https://perma.cc/NWH7-KFJ5] (11/10/2022).
Collins, Ben. 2020. “How QAnon Rode the Pandemic to New Heights — and Fueled the Viral Anti-Mask Phenomenon.” NBC News. https://www.nbcnews.com/tech/tech-news/how-qanon-rode-pandemic-new-heights-fueled-viral-anti-mask-n1236695 (8/15/2020).
Converse, Philip E. 2006. “The Nature of Belief Systems in Mass Publics (1964).” Critical Review 18: 1-74.
Cook, John, Naomi Oreskes, Peter T. Doran, William R. L. Anderegg, Bart Verheggen, Ed W. Maibach, J. Stuart Carlton, Stephan Lewandowsky, Andrew G. Skuce, Sarah A. Green, Dana Nuccitelli, Peter Jacobs, Mark Richardson, Bärbel Winkler, Rob Painting, and Ken Rice. 2016. “Consensus on Consensus: A Synthesis of Consensus Estimates on Human-Caused Global Warming.” Environmental Research Letters 11: 1-8.
Dentith, Matthew R. X. 2016. “When Inferring to a Conspiracy Might Be the Best Explanation.” Social Epistemology 1-20.
Devereaux, Ryan. 2014. “How the CIA Watched Over the Destruction of Gary Webb.” The Intercept. https://theintercept.com/2014/09/25/managing-nightmare-cia-media-destruction-gary-webb/ [https://perma.cc/7EDC-L227] (2/23/2022).
Douglas, Karen M., Joseph E. Uscinski, Robbie M. Sutton, Aleksandra Cichocka, Turkay Nefes, Chee Siang Ang, and Farzin Deravi. 2019. “Understanding Conspiracy Theories.” Advances in Political Psychology 40: 3-35.
Douglas, Karen M., Jan-Willem van Prooijen, and Robbie M. Sutton. 2021. “Is the Label ‘Conspiracy Theory’ a Cause or a Consequence of Disbelief in Alternative Narratives?” British Journal of Psychology 113(3): 575-90.
Dow, Benjamin J., Amber L. Johnson, Cynthia S. Wang, Jennifer Whitson, and Tanya Menon. 2021. “The Covid-19 Pandemic and the Search for Structure: Social Media and Conspiracy Theories.” Social and Personality Psychology Compass 15(9): 1-22.
Enders, Adam M., Joseph E. Uscinski, Michelle I. Seelig, Casey A. Klofstad, Stefan Wuchty, John R. Funchion, Manohar N. Murthi, Kamal Premaratne, and Justin Stoler. 2021a. “The Relationship between Social Media Use and Beliefs in Conspiracy Theories and Misinformation.” Political Behavior.
Enders, Adam M., Joseph E. Uscinski, Casey A. Klofstad, Michelle I. Seelig, Stefan Wuchty, Manohar N. Murthi, Kamal Premaratne, and John R. Funchion. 2021b. “Do Conspiracy Beliefs Form a Belief System? Examining the Structure and Organization of Conspiracy Beliefs.” Journal of Social and Political Psychology 9(1): 255-71.
Gallup. 2021. “Confidence in Institutions.” Gallup. https://news.gallup.com/poll/1597/confidence-institutions.aspx [https://perma.cc/L6ZZ-A7SG] (2/26/2022).
Grimes, David Robert. 2016. “On the Viability of Conspiratorial Beliefs.” PLoS ONE 11(1): e0147905.
Guess, Andrew M., Brendan Nyhan, Zachary O’Keeffe, and Jason Reifler. 2020. “The Sources and Correlates of Exposure to Vaccine-Related (Mis)Information Online.” Vaccine 38(49): 7799-7805.
Guess, Andrew M., Brendan Nyhan, and Jason Reifler. 2018. “Selective Exposure to Misinformation: Evidence from the Consumption of Fake News During the 2016 U.S. Presidential Campaign.” European Research Council. https://about.fb.com/wp-content/uploads/2018/01/fake-news-2016.pdf [https://perma.cc/DW6U-XRUE].
Han, David S. 2017. “Conspiracy Theories and the Marketplace of Facts.” First Amendment Law Review 16(2): 178-99.
Harambam, Jaron, and Stef Aupers. 2017. “‘I Am Not a Conspiracy Theorist’: Relational Identifications in the Dutch Conspiracy Milieu.” Cultural Sociology 11: 113-29.
Hay, Mark. 2017. “Untangling the 'Clinton Gave the Russians Uranium' Conspiracy Theory.” Vice, November 1, https://www.vice.com/en/article/a377wk/untangling-the-clinton-gave-the-russians-uranium-conspiracy-theory [https://perma.cc/L7QJ-PGS7]
Hay, Samantha. 2019. “‘Alternative Facts’ and Hate: Regulating Conspiracy Theories That Take the Form of Hateful Falsity.” Southern California Interdisciplinary Law Journal 29: 659-81.
Heilweil, Rebecca. 2020. “The Member of Congress Who’s Trying to Stop QAnon.” Vox. https://www.vox.com/recode/21508115/tom-malinowski-facebook-social-media-algorithms [https://perma.cc/Y2W9-QYLN] (11/19/2021).
Imhoff, Roland, and Pia Lamberty. 2020. “Conspiracy Beliefs as Psycho-Political Reactions to Perceived Power.” In Routledge Handbook of Conspiracy Theories, eds. Michael Butter and Peter Knight. New York: Routledge. 192-205.
Iyengar, Shanto, and Kyu S. Hahn. 2009. “Red Media, Blue Media: Evidence of Ideological Selectivity in Media Use.” Journal of Communication 59(1): 19-39.
Jolley, Daniel, and Karen M. Douglas. 2014. “The Effects of Anti-Vaccine Conspiracy Theories on Vaccination Intentions.” PLoS ONE 9(2): e89177.
Jolley, Daniel, Silvia Mari, and Karen M. Douglas. 2020. “Consequences of Conspiracy Theories.” In Routledge Handbook of Conspiracy Theories, eds. Michael Butter and Peter Knight. New York: Routledge. 231-41.
Kalla, Joshua L., and David E. Broockman. 2018. “The Minimal Persuasive Effects of Campaign Contact in General Elections: Evidence from 49 Field Experiments.” American Political Science Review 112(1): 148-66.
Keeley, Brian L. 1999. “Of Conspiracy Theories.” Journal of Philosophy 96(3): 109-26.
Kim, Jin Woo, Andrew Guess, Brendan Nyhan, and Jason Reifler. 2021. “The Distorting Prism of Social Media: How Self-Selection and Exposure to Incivility Fuel Online Comment Toxicity.” Journal of Communication 71(6): 922-46.
Kinder, Donald R. 1998. “Opinion and Action in the Realm of Politics.” In The Handbook of Social Psychology, eds. Daniel Todd Gilbert, Susan T. Fiske, and Gardner Lindzey. New York: McGraw-Hill. 778–867.
Klobuchar, Amy. 2021. Antitrust: Taking on Monopoly Power from the Gilded Age to the Digital Age. New York: Vintage.
Klofstad, Casey A., Joseph E. Uscinski, Jennifer M. Connolly, and Jonathan P. West. 2019. “What Drives People to Believe in Zika Conspiracy Theories?” Palgrave Communications 5(36):1-8.
Kuhn, Thomas S. 1970. The Structure of Scientific Revolutions. 2nd ed. Chicago: Chicago University of Chicago Press.
Lazarsfeld, Paul F., Bernard Berelson, and Hazel Gaudet. 1948. The People's Choice: How the Voter Makes Up His Mind in a Presidential Election. 2nd ed. New York: Columbia University Press.
Leeper, Thomas J., and Rune Slothuus. 2014. “Political Parties, Motivated Reasoning, and Public Opinion Formation.” Political Psychology 35(S1): 129-56.
Levy, Neil. 2007. “Radically Socialized Knowledge and Conspiracy Theories.” Episteme 4(1): 181-92.
Locke, John. 1967. Two Treatises of Government, ed. Peter Lasett. Cambridge: Cambridge University Press.
Lynch, Conor. 2017. “Can We Lose the Liberal Jingoism? Loose Talk About ‘Treason' Is Only Harming the Resistance.” Salon. https://www.salon.com/2017/06/24/can-we-lose-the-liberal-jingoism-loose-talk-about-treason-is-only-harming-the-resistance/ [https://perma.cc/L3XA-QNDG] (6/20/2022).
MacCarthy, Mark. 2021. “Senator Amy Klobuchar Seeks to Quell Health Misinformation on Social Media.” Brookings. https://www.brookings.edu/blog/techtank/2021/07/27/senator-amy-klobuchar-seeks-to-quell-health-misinformation-on-social-media/ [https://perma.cc/F723-XMPE] (2/20/2022).
Mancosu, Moreno, and Salvatore Vassallo. 2022. “The Life Cycle of Conspiracy Theories: Evidence from a Long-Term Panel Survey on Conspiracy Beliefs in Italy.” Italian Political Science Review/Rivista Italiana di Scienza Politica 52(1): 1-17.
McKenzie-McHarg, Andrew. 2020. “Conceptual History and Conspiracy Theory.” In Routledge Handbook of Conspiracy Theories, eds. Michael Butter and Peter Knight. New York: Routledge. 16-27.
McKenzie-McHarg, Andrew. 2018. "Conspiracy Theory: The Nineteenth-Century Prehistory of a Twentieth-Century Concept." In Conspiracy Theories and the People Who Believe Them, ed. Joseph E. Uscinski. New York: Oxford University Press. 62-81.
Mill, John Stuart. 1869. On Liberty. London: Longmans, Green, Reader, and Dyer.
Mitchelstein, Eugenia, Mora Matassi, and Pablo J. Boczkowski. 2020. “Minimal Effects, Maximum Panic: Social Media and Democracy in Latin America.” Social Media + Society 6(4): 1-11.
Moore, Alfred. 2018. “On the Democratic Problem of Conspiracy Theory Politics.” In Conspiracy Theories and the People Who Believe Them, ed. Joseph E. Uscinski. New York: Oxford University Press. 111-21.
Nunziato, Dawn Carla. 2018. “The Marketplace of Ideas Online.” Notre Dame Law Review 94(4): 1519-83.
Olmsted, Kathryn S. 2000. Challenging the Secret Government: The Post-Watergate Investigations of the CIA and FBI. Chapel Hill: The University of North Carolina Press.
Olmsted, Kathryn S. 2018. "Conspiracy Theories in U.S. History." In Conspiracy Theories and the People Who Believe Them, ed. Joseph E. Uscinski. New York: Oxford University Press. 285-97.
Olmsted, Kathryn S. 2009. Real Enemies: Conspiracy Theories and American Democracy, World War I to 9/11. New York: Oxford University Press.
Pagán, Victoria Emma. 2020. “Conspiracy Theories in the Roman Empire.” In Routledge Handbook of Conspiracy Theories, eds. Michael Butter and Peter Knight. New York: Routledge. 531-41.
Pigden, Charles. 1995. “Popper Revisited, or What Is Wrong with Conspiracy Theories?” Philosophy of the Social Sciences 25(1): 3-34.
Pummerer, Lotte. 2022. “Belief in Conspiracy Theories and Non-Normative Behavior.” Current Opinion in Psychology 47: 101394.
Räikkä, Juha, and Juho Ritola. 2020. “Philosophy and Conspiracy Theories.” In Routledge Handbook of Conspiracy Theories, eds. Michael Butter and Peter Knight. New York: Routledge. 56-66.
Riggleman, Denver. 2020. “The Spread of Conspiracy Theories Like QAnon Threatens Our Politics: GOP Congressman.” USA Today. https://www.usatoday.com/story/opinion/2020/11/14/conspiracy-theories-pose-threat-our-politics-and-civil-discourse-column/6241514002/ [https://perma.cc/Q2M3-NDJ2] (5/24/2021).
Romer, Daniel, and Kathleen Hall Jamieson. 2020. “Conspiracy Theories as Barriers to Controlling the Spread of COVID-19 in the U.S.” Social Science & Medicine 263: 113356.
Schroeder, Jared. 2019. “Saving the Marketplace from Market Failure: Reorienting Marketplace Theory in the Era of AI Communications.” William & Mary Bill of Rights Journal 28(3): 689-735.
Soave, Robby. 2021. “Why Purging Social Media of Extremist Speech Might Not Make Us Safer.” Reason. https://reason.com/2021/01/12/parler-twitter-social-media-speech-ban-terrorists-capitol-riot-trump/ [https://perma.cc/U3LM-HSVD] (2/23/2022).
Spangler, Todd. 2021. “John Stossel Sues Facebook Alleging Defamation over Fact-Check Label, Seeks at Least $2 Million.” Variety. https://variety.com/2021/digital/news/john-stossel-sues-facebook-defamation-fact-check-1235072338/ [https://perma.cc/V3ZJ-LSR2] (2/22/2022).
Stroud, Natalie Jomini. 2010. “Polarization and Partisan Selective Exposure.” Journal of Communication 60(3): 556-76.
Sunstein, Cass R. 2021. Liars: Falsehoods and Free Speech in an Age of Deception. New York: Oxford University Press.
Thalmann, Katharina. 2019. The Stigmatization of Conspiracy Theory Since the 1950s:“Plot to Make Us Look Foolish.” New York: Routledge.
Thorson, Emily A., and Stephan Stohler. 2017. “Maladies in the Misinformation Marketplace.” First Amendment Law Review 16(3): 422-53.
Uscinski, Joseph E., Darin DeWitt, and Matthew Atkinson. 2018. “A Web of Conspiracy? Internet and Conspiracy Theory.” In The Brill Handbook of Conspiracy Theory and Contemporary Religion, eds. Asbjorn Dyrendal, David Robinson, and Egil Asprem. Boston: Brill. 106-30.
Uscinski, Joseph E., Adam Enders, Casey Klofstad, Michelle Seelig, Hugo Drochon, Kamal Premaratne, and Manohar Murthi. 2022a. “Have Beliefs in Conspiracy Theories Increased Over Time?” PLoS ONE17(7): e0270429.
Uscinski, Joseph E., Adam M. Enders, Casey Klofstad, and Justin Stoler. 2022b. “Cause and Effect: On the Antecedents and Consequences of Conspiracy Theory Beliefs.” Current Opinion in Psychology 47: 101364.
Uscinski, Joseph E., Casey Klofstad, and Matthew D. Atkinson. 2016. “What Drives Conspiratorial Beliefs? The Role of Informational Cues and Predispositions.” Political Research Quarterly 69(1): 57-71.
Uscinski, Joseph E., and Joseph M. Parent. 2014. American Conspiracy Theories. New York: Oxford University Press.
van der Linden, Sander. 2015. “The Conspiracy-Effect: Exposure to Conspiracy Theories (About Global Warming) Decreases Pro-Social Behavior and Science Acceptance.” Personality and Individual Differences 87: 171-73.
Waldman, Ari Ezra. 2018. “The Marketplace of Fake News.” University of Pennsylvania Journal of Constitutional Law 20(4): 845-70.
Walker, Jesse. 2013. The United States of Paranoia: A Conspiracy Theory. New York: Harper Collins.
Walker, Jesse. 2018. “What We Mean When We Say ‘Conspiracy Theory.’” In Conspiracy Theories and the People Who Believe Them, ed. Joseph E. Uscinski. New York: Oxford University Press. 53-61.
Walker, Jesse. 2020. “You Can't Always Trust What You Hear Online, and Congress Has Some Ideas About Fixing That.” Reason. https://reason.com/2020/10/19/you-cant-always-trust-what-you-hear-online-and-congress-has-some-ideas-about-how-to-fix-that/ [https://perma.cc/CZ9Q-VY3K] (10/21/2020).
Walter, Annemarie S., and Hugo Drochon. 2020. “Conspiracy Thinking in Europe and America: A Comparative Study.” Political Studies 70(2): 483-501.
Wood, Michael J., and Karen M. Douglas. 2013. “‘What About Building 7?’ A Social Psychological Study of Online Discussion of 9/11 Conspiracy Theories.” Frontiers in Psychology 4: 409.
Zaller, John. 1992. The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press.
Zwierlein, Cornel. 2020. “Conspiracy Theories in the Middle Ages and the Early Modern Period.” In Routledge Handbook of Conspiracy Theories, eds. Michael Butter and Peter Knight. New York: Routledge. 542-54.
We also note that those engaged in illegal or undesirable behaviors exhibit other, nonconspiratorial beliefs and traits that have escaped scrutiny. For example, many of the rioters at the January 6 attack were evangelical Christians, and Christian iconography was prominent at the riot (Armaly, Buckley, and Enders 2022). While our focus is on conspiratorial speech, we believe it is worth noting that many of the focal events that have prompted concerns about conspiracy theories and related lies may very well be the causal product of other traits and beliefs (e.g., a general support for political violence, anti-social personalities, and anti-establishment proclivities).
That said, some scholars imply or explicitly claim that exposure to or belief in conspiracy theories directly cause societal harm (Dow et al. 2021; Pummerer 2022; Jolley, Mari, and Douglas 2020). Some limited experimental work has also demonstrated a link between exposure to conspiracy theories and behavioral intentions (Jolley and Douglas 2014; van der Linden 2015). However, most of the identified associations between conspiracy theories and consequences are correlational in nature, and, as such, cannot determine causality. For example, it could be that conspiracy theories are adopted as post hoc explanations for behaviors people would have engaged in anyway; the linkage could also be spurious, meaning that conspiracy beliefs and dangerous behaviors are both caused by the same underlying dispositions, such as deep-seated distrust, strong partisan or other social identities, or prejudice, for example.
We also assume some nontrivial, albeit idiosyncratic, impact of political or social circumstances. For example, Donald Trump inviting supporters to Washington, D.C., for a rally on January 6, 2021, and subsequently encouraging them to march to the Capitol surely increased the probability of violence.
It is very likely the case that social media algorithms promote conspiracy beliefs among those already seeking out conspiratorial content. This is congruent with our argument that motivations matter—the more conspiratorial content one is motivated to seek out, the more will come their way by algorithmic suggestion. That said, there is no evidence that conspiracy beliefs have increased over time (Uscinski et al. 2022a), even during the pandemic when more people were online (Romer and Jamieson 2020).
Each data point corresponds to a national U.S. survey of between 1,040 and 2,021 adults that is designed to be representative of the population based on age, gender, educational attainment, and race.
Adam M. Enders is an associate professor of political science at the University of Louisville.
Joseph Uscinski is a professor of political science at the University of Miami.