Misunderstanding the harms of online misinformation.

The controversy over online misinformation and social media has opened a gap between public discourse and scientific research. Public intellectuals and journalists frequently make sweeping claims about the effects of exposure to false content online that are inconsistent with much of the current empirical evidence. Here we identify three common misperceptions: that average exposure to problematic content is high, that algorithms are largely responsible for this exposure and that social media is a primary cause of broader social problems such as polarization. In our review of behavioural science research on online misinformation, we document a pattern of low exposure to false and inflammatory content that is concentrated among a narrow fringe with strong motivations to seek out such information. In response, we recommend holding platforms accountable for facilitating exposure to false and extreme content in the tails of the distribution, where consumption is highest and the risk of real-world harm is greatest. We also call for increased platform transparency, including collaborations with outside researchers, to better evaluate the effects of online misinformation and the most effective responses to it. Taking these steps is especially important outside the USA and Western Europe, where research and data are scant and harms may be more severe.

Like-minded sources on Facebook are prevalent but not polarizing.

Many critics raise concerns about the prevalence of 'echo chambers' on social media and their potential role in increasing political polarization. However, the lack of available data and the challenges of conducting large-scale field experiments have made it difficult to assess the scope of the problem1,2. Here we present data from 2020 for the entire population of active adult Facebook users in the USA showing that content from 'like-minded' sources constitutes the majority of what people see on the platform, although political information and news represent only a small fraction of these exposures. To evaluate a potential response to concerns about the effects of echo chambers, we conducted a multi-wave field experiment on Facebook among 23,377 users for whom we reduced exposure to content from like-minded sources during the 2020 US presidential election by about one-third. We found that the intervention increased their exposure to content from cross-cutting sources and decreased exposure to uncivil language, but had no measurable effects on eight preregistered attitudinal measures such as affective polarization, ideological extremity, candidate evaluations and belief in false claims. These precisely estimated results suggest that although exposure to content from like-minded sources on social media is common, reducing its prevalence during the 2020 US presidential election did not correspondingly reduce polarization in beliefs or attitudes.

The effects of Facebook and Instagram on the 2020 election: A deactivation experiment.

We study the effect of Facebook and Instagram access on political beliefs, attitudes, and behavior by randomizing a subset of 19,857 Facebook users and 15,585 Instagram users to deactivate their accounts for 6 wk before the 2020 U.S. election. We report four key findings. First, both Facebook and Instagram deactivation reduced an index of political participation (driven mainly by reduced participation online). Second, Facebook deactivation had no significant effect on an index of knowledge, but secondary analyses suggest that it reduced knowledge of general news while possibly also decreasing belief in misinformation circulating online. Third, Facebook deactivation may have reduced self-reported net votes for Trump, though this effect does not meet our preregistered significance threshold. Finally, the effects of both Facebook and Instagram deactivation on affective and issue polarization, perceived legitimacy of the election, candidate favorability, and voter turnout were all precisely estimated and close to zero.

Time and skeptical opinion content erode the effects of science coverage on climate beliefs and attitudes.

Although experiments show that exposure to factual information can increase factual accuracy, the public remains stubbornly misinformed about many issues. Why do misperceptions persist even when factual interventions generally succeed at increasing the accuracy of people's beliefs? We seek to answer this question by testing the role of information exposure and decay effects in a four-wave panel experiment (n = 2,898 at wave 4) in which we randomize the media content that people in the United States see about climate change. Our results indicate that science coverage of climate change increases belief accuracy and support for government action immediately after exposure, including among Republicans and people who reject anthropogenic climate change. However, both effects decay over time and can be attenuated by exposure to skeptical opinion content (but not issue coverage featuring partisan conflict). These findings demonstrate that the increases in belief accuracy generated by science coverage are short lived and can be neutralized by skeptical opinion content.

Why the backfire effect does not explain the durability of political misperceptions.

Previous research indicated that corrective information can sometimes provoke a so-called "backfire effect" in which respondents more strongly endorsed a misperception about a controversial political or scientific issue when their beliefs or predispositions were challenged. I show how subsequent research and media coverage seized on this finding, distorting its generality and exaggerating its role relative to other factors in explaining the durability of political misperceptions. To the contrary, an emerging research consensus finds that corrective information is typically at least somewhat effective at increasing belief accuracy when received by respondents. However, the research that I review suggests that the accuracy-increasing effects of corrective information like fact checks often do not last or accumulate; instead, they frequently seem to decay or be overwhelmed by cues from elites and the media promoting more congenial but less accurate claims. As a result, misperceptions typically persist in public opinion for years after they have been debunked. Given these realities, the primary challenge for scientific communication is not to prevent backfire effects but instead, to understand how to target corrective information better and to make it more effective. Ultimately, however, the best approach is to disrupt the formation of linkages between group identities and false claims and to reduce the flow of cues reinforcing those claims from elites and the media. Doing so will require a shift from a strategy focused on providing information to the public to one that considers the roles of intermediaries in forming and maintaining belief systems.

Overconfidence in news judgments is associated with false news susceptibility.

We examine the role of overconfidence in news judgment using two large nationally representative survey samples. First, we show that three in four Americans overestimate their relative ability to distinguish between legitimate and false news headlines; respondents place themselves 22 percentiles higher than warranted on average. This overconfidence is, in turn, correlated with consequential differences in real-world beliefs and behavior. We show that overconfident individuals are more likely to visit untrustworthy websites in behavioral data; to fail to successfully distinguish between true and false claims about current events in survey questions; and to report greater willingness to like or share false content on social media, especially when it is politically congenial. In all, these results paint a worrying picture: The individuals who are least equipped to identify false news content are also the least aware of their own limitations and, therefore, more susceptible to believing it and spreading it further.

Elite rhetoric can undermine democratic norms.

Democratic stability depends on citizens on the losing side accepting election outcomes. Can rhetoric by political leaders undermine this norm? Using a panel survey experiment, we evaluate the effects of exposure to multiple statements from former president Donald Trump attacking the legitimacy of the 2020 US presidential election. Although exposure to these statements does not measurably affect general support for political violence or belief in democracy, it erodes trust and confidence in elections and increases belief that the election is rigged among people who approve of Trump's job performance. These results suggest that rhetoric from political elites can undermine respect for critical democratic norms among their supporters.

A digital media literacy intervention increases discernment between mainstream and false news in the United States and India.

Widespread belief in misinformation circulating online is a critical challenge for modern societies. While research to date has focused on psychological and political antecedents to this phenomenon, few studies have explored the role of digital media literacy shortfalls. Using data from preregistered survey experiments conducted around recent elections in the United States and India, we assess the effectiveness of an intervention modeled closely on the world's largest media literacy campaign, which provided "tips" on how to spot false news to people in 14 countries. Our results indicate that exposure to this intervention reduced the perceived accuracy of both mainstream and false news headlines, but effects on the latter were significantly larger. As a result, the intervention improved discernment between mainstream and false news headlines among both a nationally representative sample in the United States (by 26.5%) and a highly educated online sample in India (by 17.5%). This increase in discernment remained measurable several weeks later in the United States (but not in India). However, we find no effects among a representative sample of respondents in a largely rural area of northern India, where rates of social media use are far lower.

Information From Same-Race/Ethnicity Experts Online Does Not Increase Vaccine Interest or Intention to Vaccinate.

Policy Points Mass vaccination is essential for bringing the COVID-19 pandemic to a close, yet substantial disparities remain between whites and racial and ethnic minorities within the United States. Online messaging campaigns featuring expert endorsements are a low-cost way to increase vaccine awareness among minoritized populations, yet the efficacy of same-race/ethnicity expert messaging in increasing uptake remains unknown. Our preregistered analysis of an online vaccine endorsement campaign, which randomly varied the racial/ethnic identity of the expert, revealed no evidence that information from same race/ethnicity experts affected vaccine interest or the intention to vaccinate. Our results do not rule out the possibility that other low-cost endorsement campaigns may be more effective in increasing vaccine uptake, but do suggest that public health campaigns might profitably focus on issues of access and convenience when targeting minoritized populations in the United States. CONTEXT: The COVID-19 pandemic in the United States has been unequally experienced across racial and ethnic groups. Mass vaccination is the most effective way to bring the pandemic to an end and to manage its public health consequences. But the racialization of public health delivery in the United States has produced a sizable racial/ethnic gap in vaccination rates. Closing this gap in vaccine uptake is therefore essential to ending the pandemic. METHODS: We conducted a preregistered, well-powered (N = 2,117) between-subjects survey experiment, fielded March 24 to April 5, 2021, in which participants from YouGov's online panel-including oversamples of Black (n = 471), Hispanic/Latino/a (n = 430), and Asian American (n = 319) participants-were randomly assigned to see COVID-19 vaccine information endorsed by same- or different-race/ethnicity experts or to a control condition. We then measured respondents' vaccination intentions, intention to encourage others to get vaccinated, and interest in learning more information and sharing information with others. FINDINGS: Same-race/ethnicity expert endorsements had no measurable effect on nonwhite or white respondents' willingness to get the COVID-19 vaccine, to encourage others to get the vaccine, or to learn more or share information with others. CONCLUSIONS: Our study provides empirical evidence suggesting online endorsements from same-race/ethnicity experts do not increase vaccine interest, advocacy, or uptake, though same-race/ethnicity endorsements may be effective in other venues or mediums.

Critical dynamics in population vaccinating behavior.

Vaccine refusal can lead to renewed outbreaks of previously eliminated diseases and even delay global eradication. Vaccinating decisions exemplify a complex, coupled system where vaccinating behavior and disease dynamics influence one another. Such systems often exhibit critical phenomena-special dynamics close to a tipping point leading to a new dynamical regime. For instance, critical slowing down (declining rate of recovery from small perturbations) may emerge as a tipping point is approached. Here, we collected and geocoded tweets about measles-mumps-rubella vaccine and classified their sentiment using machine-learning algorithms. We also extracted data on measles-related Google searches. We find critical slowing down in the data at the level of California and the United States in the years before and after the 2014-2015 Disneyland, California measles outbreak. Critical slowing down starts growing appreciably several years before the Disneyland outbreak as vaccine uptake declines and the population approaches the tipping point. However, due to the adaptive nature of coupled behavior-disease systems, the population responds to the outbreak by moving away from the tipping point, causing "critical speeding up" whereby resilience to perturbations increases. A mathematical model of measles transmission and vaccine sentiment predicts the same qualitative patterns in the neighborhood of a tipping point to greatly reduced vaccine uptake and large epidemics. These results support the hypothesis that population vaccinating behavior near the disease elimination threshold is a critical phenomenon. Developing new analytical tools to detect these patterns in digital social data might help us identify populations at heightened risk of widespread vaccine refusal.

Treatment versus Punishment: Understanding Racial Inequalities in Drug Policy.

CONTEXT: Many observers believe that the policy response to the opioid crisis is less punitive than the crack scare and that the reason is that victims are (stereotypically) white. METHODS: To assess this conjecture, we compile new longitudinal data on district-level drug-related deaths and (co)sponsorship of legislation on drug abuse in the House of Representatives over the past four decades. Using legislator fixed effects models, we then test how changes in drug-related death rates in legislators' districts predict changes in (co)sponsorship of treatment-oriented or punitive legislation in the subsequent year and assess whether these relationships vary by race of victim or drug type. FINDINGS: Policy makers were more likely to introduce punitive drug-related bills during the crack scare and are more likely to introduce treatment-oriented bills during the current opioid crisis. The relationship between district-level drug deaths and subsequent sponsorship of treatment-oriented legislation is greater for opioid deaths than for cocaine-related deaths and for white victims than for black victims. By contrast, district-level drug deaths are not significantly related to sponsorship of punishment-oriented bills. CONCLUSIONS: These results suggest that the racial inequalities and double standards of drug policy still persist but in different forms.

Interventions to counter misinformation: Lessons from the Global North and applications to the Global South.

We synthesize evidence from 176 experimental estimates of 11 interventions intended to combat misinformation in the Global North and Global South, which we classify as informational, educational, sociopsychological, or institutional. Among these, we find the most consistent positive evidence for two informational interventions in both Global North and Global South contexts: inoculation/prebunking and debunking. In a complementary survey of 138 misinformation scholars and practitioners, we find that experts tend to be most optimistic about interventions that have been least widely studied or that have been shown to be mostly ineffective. We provide a searchable database of misinformation randomized controlled trials and suggest avenues for future research to close the gap between expert opinion and academic research.

Thinking about default enrollment lowers vaccination intentions and public support in G7 countries.

Policymakers often face a conundrum between being transparent about policies and ensuring that those policies are effective. This challenge is particularly relevant for behavioral nudges, which are not usually disclosed. Rather than avoiding transparency, we suggest that policymakers encourage citizens to reflect on nudges to help them understand their own views and align those views with their behaviors. Using data from an online survey experiment with 24,303 respondents in G7 countries, we examine the impact of reflection on a hypothetical default nudge policy for COVID-19 booster appointments. Contrary to expectations, participants say they would be less likely to get the booster when automatically enrolled compared with a control condition. Similarly, encouraging citizens to think about the status quo (baseline) policy also reduces intentions for boosters. These interventions have no effect on approval of the policy. Further, encouraging people to think about automatic enrollment decreases approval of the policy and further decreases their intentions to get vaccinated. These findings suggest that reflection on a nudge can increase backlash from a nudge and also elicit policy disapproval, thereby aligning policy support with behavioral intentions.

Public support for more stringent vaccine policies increases with vaccine effectiveness.

Under what conditions do citizens support coercive public policies? Although recent research suggests that people prefer policies that preserve freedom of choice, such as behavioural nudges, many citizens accepted stringent policy interventions like fines and mandates to promote vaccination during the COVID-19 pandemic-a pattern that may be linked to the unusually high effectiveness of COVID-19 vaccines. We conducted a large online survey experiment (N = 42,417) in the Group of Seven (G-7) countries investigating the relationship between a policy's effectiveness and public support for stringent policies. Our results indicate that public support for stringent vaccination policies increases as vaccine effectiveness increases, but at a modest scale. This relationship flattens at higher levels of vaccine effectiveness. These results suggest that intervention effectiveness can be a significant predictor of support for coercive policies but only up to some threshold of effectiveness.

The hazards of correcting myths about health care reform.

CONTEXT: Misperceptions are a major problem in debates about health care reform and other controversial health issues. METHODS: We conducted an experiment to determine if more aggressive media fact-checking could correct the false belief that the Affordable Care Act would create "death panels." Participants from an opt-in Internet panel were randomly assigned to either a control group in which they read an article on Sarah Palin's claims about "death panels" or an intervention group in which the article also contained corrective information refuting Palin. FINDINGS: The correction reduced belief in death panels and strong opposition to the reform bill among those who view Palin unfavorably and those who view her favorably but have low political knowledge. However, it backfired among politically knowledgeable Palin supporters, who were more likely to believe in death panels and to strongly oppose reform if they received the correction. CONCLUSIONS: These results underscore the difficulty of reducing misperceptions about health care reform among individuals with the motivation and sophistication to reject corrective information.

Subscriptions and external links help drive resentful users to alternative and extremist YouTube channels.

Do online platforms facilitate the consumption of potentially harmful content? Using paired behavioral and survey data provided by participants recruited from a representative sample in 2020 (n = 1181), we show that exposure to alternative and extremist channel videos on YouTube is heavily concentrated among a small group of people with high prior levels of gender and racial resentment. These viewers often subscribe to these channels (prompting recommendations to their videos) and follow external links to them. In contrast, nonsubscribers rarely see or follow recommendations to videos from these channels. Our findings suggest that YouTube's algorithms were not sending people down "rabbit holes" during our observation window in 2020, possibly due to changes that the company made to its recommender system in 2019. However, the platform continues to play a key role in facilitating exposure to content from alternative and extremist channels among dedicated audiences.

Asymmetric ideological segregation in exposure to political news on Facebook.

Does Facebook enable ideological segregation in political news consumption? We analyzed exposure to news during the US 2020 election using aggregated data for 208 million US Facebook users. We compared the inventory of all political news that users could have seen in their feeds with the information that they saw (after algorithmic curation) and the information with which they engaged. We show that (i) ideological segregation is high and increases as we shift from potential exposure to actual exposure to engagement; (ii) there is an asymmetry between conservative and liberal audiences, with a substantial corner of the news ecosystem consumed exclusively by conservatives; and (iii) most misinformation, as identified by Meta's Third-Party Fact-Checking Program, exists within this homogeneously conservative corner, which has no equivalent on the liberal side. Sources favored by conservative audiences were more prevalent on Facebook's news ecosystem than those favored by liberals.

How do social media feed algorithms affect attitudes and behavior in an election campaign?

We investigated the effects of Facebook's and Instagram's feed algorithms during the 2020 US election. We assigned a sample of consenting users to reverse-chronologically-ordered feeds instead of the default algorithms. Moving users out of algorithmic feeds substantially decreased the time they spent on the platforms and their activity. The chronological feed also affected exposure to content: The amount of political and untrustworthy content they saw increased on both platforms, the amount of content classified as uncivil or containing slur words they saw decreased on Facebook, and the amount of content from moderate friends and sources with ideologically mixed audiences they saw increased on Facebook. Despite these substantial changes in users' on-platform experience, the chronological feed did not significantly alter levels of issue polarization, affective polarization, political knowledge, or other key attitudes during the 3-month study period.

Reshares on social media amplify political news but do not detectably affect beliefs or opinions.

We studied the effects of exposure to reshared content on Facebook during the 2020 US election by assigning a random set of consenting, US-based users to feeds that did not contain any reshares over a 3-month period. We find that removing reshared content substantially decreases the amount of political news, including content from untrustworthy sources, to which users are exposed; decreases overall clicks and reactions; and reduces partisan news clicks. Further, we observe that removing reshared content produces clear decreases in news knowledge within the sample, although there is some uncertainty about how this would generalize to all users. Contrary to expectations, the treatment does not significantly affect political polarization or any measure of individual-level political attitudes.