Children and social media: a new problem we’re tackling with old tools

Hearings continued before the US Senate sub-committee on the impact that new forms of entertainment have on young people. The most prominent guest was a psychiatrist of solid reputation, the author of studies on the mental health of Black children in segregated neighbourhoods – research which, in other forums, has earned him the esteem of judges and civil rights lawyers. “I won’t be talking to you about averages, sample surveys or correlation coefficients. I know these children: I see thousands of them every year.” The picture he painted was no different from that of recent days: children use the new medium every day, to an extent that no adult suspects; they do so in solitude, often in secret; if interrupted, they become irritable; their academic performance drops; they sleep less; they withdraw from their families and from their real-life friends. He added that he had come across cases where the content had acted as a catalyst for premature or abnormal sexual urges. When asked to quantify the damage, he replied that he was not there to compile statistics. Also yesterday, a representative from a national association of parents and teachers spoke, presenting letters received from families across the country: the recurring phrase is always the same, ‘I no longer recognise my child’.

When an industry executive took his place at the hearing, he was unable to hide his unease. He pointed out that the product’s main audience is not children; he implied that the most serious cases concern competitors’ products rather than his own; and he said he was ready to cooperate fully with the authorities. When questioned about a single piece of content – particularly graphic in nature – which a senator had shown in the chamber, he replied that, in his view, it did not exceed the bounds of good taste. Outside the chamber, a group of parents were handing out leaflets. In the preceding months, in various towns across the country, young people had been gathered in school and parish playgrounds to publicly burn the offending material.

On reviewing the sub-committee’s work, seven recurring themes can be identified.

Structural thesis: for the first time, an industry is speaking directly to minors, bypassing any adult mediation: the child accesses the content independently; the parent is unaware, the teacher is unaware, the priest is unaware.

The quantity argument: it is not just the quality of the content that is under discussion, but the sheer volume of consumption, which is unprecedented in the history of childhood.

Imitative approach: what you see is what you get.

The desensitisation theory: repeated exposure dulls the emotional reaction, and what was initially disturbing eventually ceases to be so.

The ‘subtraction’ thesis: those hours are not added to, but taken away from: reading, sport, sleep, and family conversation.

The theory behind the vulnerability: the boy’s mind, still developing, lacks the restraints of an adult’s.

The theory of intent: the most serious one: the industry knows everything. It is not a side effect, it is the product itself; those who sell it are aware of the harm and factor it in, because it is the harm that generates profit.

On this basis, it is likely that the sub-committee will rule in favour of certain proposals that have been circulating for months: setting an age limit for access, obliging the industry to include warnings on the product, and banning its distribution in schools. Which committee are we talking about? In reality, the contents of two hearings held some time apart have been combined here.

The first dates from April 1954: the Senate Sub-Committee on Juvenile Delinquency, chaired by Robert Hendrickson. The key witness was the psychiatrist Fredric Wertham. The medium deemed responsible for the collapse of American childhood was the comic strip. In October of that year, in an effort to avert further legislative action, publishers established the Comics Code Authority, which banned stories sympathetic to crime and imposed a largely black-and-white portrayal of crimes and criminals: the creativity of a generation of comic-book artists was stifled.

The second took place in December 1993: a joint hearing of two sub-committees – Joseph Lieberman’s Sub-committee on Governmental Affairs and Herb Kohl’s Sub-committee on Juvenile Justice – with video games in the dock, specifically ‘Mortal Kombat’ and ‘Night Trap’. Following a second session on 4 March 1994, the industry announced the Entertainment Software Rating Board (ESRB) in July, which has been classifying video games according to their content since the following September. Forty years on, the same alarmist arguments were levelled at comic books and video games – which have nothing in common. And today, decades after 1993, the same list still holds true: the same arguments recur in attacks on social media, and the evidence cited to support them is always similar. Clinical case studies are cited, without a control group. Correlations are used and interpreted as causal links. Reference is made to laboratory experiments conducted on proxy measures of aggression that no one has ever validated against real-world behaviour. And finally, there is the dose-response rhetoric, the most persuasive of all because it makes the medium resemble a toxic substance: those who use it for more than three hours a day show twice as many symptoms as those who use it for less than one.

Just because the experts were wrong about comics and video games doesn’t necessarily mean they’re wrong again about social media: the latter is certainly more pervasive. And there is one piece of data that cannot be dismissed with a joke: around 2012, in several English-speaking and Nordic countries simultaneously, indicators of mental distress among adolescents – particularly girls – began to trend upwards. However one explains it, that bend in the curve is there. The problem, then, is this: why, when faced with a new problem, does a community of experts find it so easy to trace it back to an old case? To understand what is happening, it is worth borrowing a tool from the philosophy of science. Imre Lakatos observed that theories are almost never abandoned one by one, as Popper claimed, when they are ‘falsified’ (that is, contradicted even by a single piece of data): scientists work within what he called research programmes. Every programme has two parts.

There is a core: the fundamental assumption that none of the adherents is prepared to question, because to abandon it would mean leaving the programme. And there is a protective belt of auxiliary hypotheses: the more specific claims, those that are truly open to empirical testing and which, when an experiment goes wrong, are sacrificed in place of the core. Lakatos, however, added a criterion for distinguishing living programmes from those in decline. A programme is progressive when adjustments to the belt lead to predictions of new facts, which are then actually found: the hypothesis of the invisible planet is progressive because it leads to Neptune. It is degenerative when the adjustments serve merely to accommodate falsifications – the theory is corrected after the event, an explanation is given for why the experiment went wrong, and no new prediction is derived that was not already known. A degenerative programme is not false: it is sterile, and the difference only becomes apparent in the medium term.

Let us therefore examine how this research programme operates. Over the past seventy years, its core premise has not shifted by a single millimetre: the new medium damages young people’s minds. The framework has been revised seven times. It began with delinquency; when, in the 1990s, youth crime plummeted whilst media consumption rose, the focus shifted to aggression; when meta-analyses reduced the effects on aggression to a statistical speck, the focus shifted to desensitisation; then to addiction; then to attention deficits; then to loneliness; and today to anxiety. Seven formulations, and not a single one that predicted in advance a new finding that was subsequently confirmed. Each revision came after the previous one had been disproved, never before, and each time left the core exactly where it was. Clinical psychology and psychiatry are therefore routinely enlisted to support the same argument, even when it concerns entirely different forms of entertainment. A condition defined by thresholds of behaviour and subjective distress can be broadened to cover whatever an era deems excessive.

This is no mere suspicion: gaming disorder was included in the ICD-11 in June 2018, based on research that has been widely contested by researchers in the field themselves, and its adoption has legitimised night-time curfew policies for minors in East Asia. This did not prevent Seoul from repealing its own policy in August 2021, after a decade in which the most visible outcome had been to prompt minors to obtain adults’ identification codes. A diagnosis transforms a social preference into a medical fact, and a medical fact into a pretext for public intervention. This is the debate that needs to be had, rather than a ‘social media yes or no’ debate based on a selective use of data that ministers and MPs are simply falling in line with.

There are numerous cases in which doctors have invented imaginary mental illnesses, stemming from social paranoia, which then disappeared, taking the illness with them, and the list is instructive. In the late 18th century, there was talk of ‘Lesesucht’, the compulsion to read novels, complete with a treatment regimen: cold baths, exercise and diet. There was the ‘Werther panic’ of 1774, the first wave of suicides attributed to a medium – except that documentary evidence of that wave of suicides following the reading of Goethe’s novel is virtually non-existent: it nevertheless survives in literature as the ‘Werther effect’, an expression coined by David Phillips in 1974, exactly two centuries after the novel’s publication. There had been Johannes Hofer’s ‘nostalgia’ (1688), a condition with a specific course and mortality rate, now reduced to a mere noun. There was Cartwright’s drapetomania (1851), the condition affecting black slaves attempting to escape; masturbatory insanity, which, unlike the others, resulted in actual surgical mutilations on real patients; and ‘railway spine’, which affected those travelling by train in the early decades of the railway. In cases where new developments are associated with new habits and alter the cognitive landscape, creating new social complexities, the outcome is always the same: to ban access to a means of transport because it causes a mental illness. Which would require that such an illness actually existed.

Lakatos would undoubtedly have recognised this as a degenerative research programme. Anyone who believes that the ‘core’ should be applied to social media – and not merely yet another ‘belt’ – has a simple way of proving it: state now what new fact their hypothesis predicts, make a prediction about the future effects of social media, and spell out the conditions under which the hypothesis can be falsified. Wait and see. In 2012, Carol Tilley, whilst examining Wertham’s papers – which had been made available for consultation at the Library of Congress in 2010 – documented that the clinical cases reported in "The Seduction of the Innocent" – the great "j’accuse" against hard-boiled comics – had been manipulated: ages altered, statements pieced together, contexts omitted, and conclusions not supported by the evidence. The question to ask is not whether those who are now accusing social media are right this time, but why the answer has been ready for seventy years, merely awaiting a new medium to which it can be applied.