There are many messages from science that bear repeating. One of them is that correlation does not mean causation. When headlines include a cause-and-effect conclusion based on correlative evidence, that is the false cause fallacy, and it can mislead the public with worrying consequences.
I have previously identified this misconception in unfair public judgments about minorities, helicopter parents, police officers, transgender athletes, people with fixed mindsets, and those who Zoom (Stalder, 2018, 2000, 2021), but the headlines keep coming. As the new year begins, I’m thinking about two high-profile victims of the past year’s false cause fallacy: immigration and social media.
Does immigration lead to an increase in crime?
Source: David Peinado/Pexels
The answer is no, despite many politicians vigorously claiming otherwise. Their “evidence” for the false conclusion seemed to be specific incidents in which an undocumented immigrant of color committed a terrible crime. But large-scale research overwhelmingly shows no connection and shows that immigrants are less likely to commit crimes (Hesson and Rosenberg, 2024; Seid et al., 2024).
No high-profile figure has correspondingly argued that being a white American citizen causes an increase in crime after such a person has committed a terrible crime. After a US-born citizen allegedly committed the recent truck attack in New Orleans, some politicians wrongly suggested that the perpetrator was actually an immigrant (Shabad, 2025).
The false alarm about immigrants appeared to be due to anti-immigrant bias and presidential politics. Another factor is that there are fewer immigrants in the US than non-immigrants. When two relatively less likely events occur at the same time (being a recently arrived immigrant and committing a terrible crime), our brains can hardly help but connect them – such a common misconception becomes a illusory correlation.
Is social media use causing an increase in mental health problems?
Compared to the obvious immigration problem, the social media question has become an empirical mess. Strong, complicated disagreements exist among experts.
Until about 2017, social media unfairly received a lot of hits while the underlying research was completely correlational (Stalder, 2018). For example, maybe teens with depression were more likely to use social media, and it wasn’t that social media was causing the depression.
Ultimately, some randomized controlled trials, or real experiments, showed a cause-and-effect link between social media use and poor mental health, but these early studies had too many limitations to warrant a summary conclusion. Moreover, there were also some real experiments that showed no harm and even some benefits.
Haidts The fearful generation
Source: ghcassel/Pixabay
Then in 2024, Jonathan Haidt’s best-selling book entitled The Anxious Generation: How the Great Rewiring of Childhood Is Causing an Epidemic of Mental Illness. Haidt gathered an impressive amount of potential evidence and wrote convincingly. To his credit, he directly addressed the issue of cause and effect. He referred to twenty real experiments, fourteen of which “found evidence of harm” and six that did not (although the six were only mentioned in a footnote, while the twenty were exaggerated to “dozens” that “confirmed Haidt’s view”). Even if every experiment had no limitations, is a ratio of 14 to 6 good enough, while most of the research was still correlational?
Candice Odgers, recipient of an American Psychological Association award for distinguished contributions “in the public interest,” wrote a scathing review in Nature. Odgers wrote that Haidt’s position is “not supported by science” and that Haidt “is a gifted storyteller, but his story is currently one in search of evidence” (Odgers, 2024).
In a later interview, Odgers expressed concern that Haidt was using a “fear story and scare tactic,” which the interviewer suggested deterred other “prominent psychologists” from criticizing Haidt “on the record.” Haidt admitted to the same interviewer that the book proposed social changes “before the scientific community has reached full agreement.” But Haidt thought it would be irresponsible to wait any longer and saw no harm in his proposals, which included a ban on cell phones at school. Odgers, on the other hand, saw potential harm, including shaming teens (Thorp, 2024a, 2024b).
A New York Times The review also stated that Haidt’s proposals could cause harm and that Haidt did a better job of acknowledging the scientific limitations “online” than in his book (Dennis-Tiwary, 2024). The conversation provided a timeline and some details of the recent back-and-forth among experts, calling it a “messy fight” (Meyerowitz-Katz and Jané, 2024).
My take on Haidt’s book
Ultimately, Haidt took a “what else could it be” approach, which is one of the pitfalls of the false cause fallacy. However, Haidt showed great diligence in investigating other possible causes, and Haidt directly addressed the issue of cause and effect, something left unaddressed by most critics. Haidt even has one ongoing blog where he welcomes input and new research. In recent months there have been a few new studies that support his position, and a few that contradict it. Two of the counterstudies do not yet appear on his blog (Jones et al., 2024; Taylor et al., 2024).
But in other parts of Haidt’s book, he routinely used terms like “damage” and “damage,” even when the specific research cited in that section was only correlative. Haidt also claimed that reduced in-person social interactions (caused by social media use) were a mediating source of harm, but that view opens yet another Pandora’s box.
A lot of social interaction is not necessarily a good thing. There is inevitable conformity with peers, which can contribute to teenage problems including drug use and crime (Warr, 2002) and more general group phenomena such as deindividuation, groupthink and even the bystander effect (Stalder, 2014). And psychology recognizes that children who interact minimally with peers can still do well, and sometimes even better, than extroverted children (Cain, 2013).
In sum
The immigration issue appears to have been scientifically resolved. The social media story is much more complicated. Haidt’s book had several strengths, but despite some attention to counterpoints, it still seemed to overstate the message. Failure to write more carefully or share more of the field-wide disagreements can overly contribute to anti-smartphone fervor and causation-equals-correlation thinking that is all too common among general readers.
