In recent years there has been growing awareness of the potential harms caused by smartphones and social media, from Jonathan Haidt’s The Anxious Generation to the hit drama Adolescence.1
While not an absolute slam-dunk, we are seeing increasing evidence showing causation as well as correlation,2 not least because there are obvious plausible mechanisms that almost all of us can see in terms of how to much time on our phones social media can affect ourselves.3
But beneath the headlines, there are two distinct hypotheses4 as to how that harm arises - which I will call here The Content Hypothesis and The Medium Hypothesis.
These are not mutually exclusive - and one could believe in both simultaneously. But in practice, most people seem to lean towards one or the other, sometimes heavily - and they have very different implications in what they imply for policy solutions at a governmental, school and family level.
The Content Hypothesis
“The righteous need not cower before the drumbeat of human progress. Though the song of yesterday fades into the challenge of tomorrow, God still watches and judges us. Evil lurks in the datalinks as it lurked in the streets of yesteryear. But it was never the streets that were evil.”
– Sister Miriam Godwinson, “A Blessed Struggle”
Under the content hypothesis, the principal harm caused by smartphones is the access they give people - particularly, but not exclusively, children - to harmful content.
Depending on your particualar concerns, this could includesome or all of pornography, graphic violence, pro-anorexia sites, vaccine denialism or radicalising material, whether that is Andrew Tate-style misogyny, Islamism, far-right content or extreme environmentalist beliefs. It could include ‘misinformation’, or conspiracy theories, which appear to have become both significantly more prevalent and more outrageous over the last decade.
For those who favour the content hypothesis, the principal concern is removing harmful content - or at least preventing children from getting their hands on it. You may support bans on, or enforced age requirements, for things such as pornography, and wish social media companies to take stricter action to remove ‘misinformation’ or harmful content. You may support laws such as the UK’s Online Safety Act5 (or think it should go further) want the government or regulators to take stronger action to enforce this, or to fine companies that do not comply. You may want ISPs to make certain sites unavailable, or to ban or restrict the use of VPNs, to prevent these restrictions from being bypassed.
Favouring the content hypothesis doesn’t necessarily mean you have no concerns about free speech, or don’t recognise that governments or big corporates could wrongly classify views they don’t like as ‘misinformation’. But it likely does mean that you will be willing to accept trade-offs here, to keep harmful material off the internet and away from children.
Conversely, if you favour the Content hypothesis, you may be happy to give a smartphone to your children at a younger age, provided they have good parental controls on.6 You may be happy for smartphones and the internet to be used extensively in schools, provided the children’s use is monitored and/or bad sites are blocked. The biggest factor on which social media sites you are comfortable to use is likely to be the extent to which they adequately police content, rather than the form of content or the algorithms used.
The Medium Hypothesis
“The Medium is the Message.”
– Marshall McLuhan
In the Medium hypothesis, the principal harm caused by smartphones and social media results from the nature of the medium itself and its impact upon our brains.
This could include concerns about the impact on mental health, including anxiety, depression and self-esteem; the addictive nature of phones that displaces other, healthier activities (such as outdoor play, or reading); the contribution of phone availability to loneliness; the impact of short-form video on attention spans; the way algorithmic feeds can create filter bubbles or reinforce extreme or conspiracy-based ideas; or the way in which features such as endless scroll or feedback from ‘likes’ and other community mechanisms can reinforce unhealthy behaviours or beliefs.
Favouring the Medium hypothesis doesn’t mean having no concerns about content.7 But it does mean believing that, even if all the content was benign, that these devices and apps would still be harmful - and that addressing the harm being done by the medium is a higher priority than trying to police content.
Those who favour the Medium hypothesis may be more likely to favour tougher age controls or outright bans on the use of smartphones and/or social media by children. They are more likely to support phone-free schools and to avoid giving their own children a smartphone. Where regulation or corporate activity is concerned, they are more likely to see actions to address the nature of the Medium itself as a priority - such as mandating a chronological feed, algorithm transparency, or banning endless scroll. They may be more likely to support features that help adults control their use, such as the ability for devices or apps to set timelimits on activity, and be more cautious about their own use.
Conversely, if you favour the Medium hypothesis, you may be more concerned about censorship and free speech. You may be cautious of action by governments or social media companies in deciding what is or is not ‘harmful’ or ‘misinformation’. You may believe that while age restrictions on content for children are acceptable, these should not be applied to adults. When it comes to which social media sites to use, they may be more concerned about the algorithms and type of content being shown - for example, by avoiding short-form video sites due to considering them innately harmful.
So, which is right - and why does it matter?
To emphasise again, these hypotheses are not mutually exclusive. You can believe that both the content and the medium are harmful.8
But we all live in the world in which smartphones and social media exist, so unless we swear off them entirely (difficult!), we have to make choices and trade-offs - both in our own use and in that of our children. Governments have a limited appetite to regulate, particularly against powerful lobbies, and campaigners must also choose where to spend their efforts: so, at this level, too one must decide which interventions to prioritise.
In making those choices, it helps to be clear about one’s theory of harm.
In high status circles - i.e. the ‘centre’9 and progressive centre - there is much more willingness to believe and to talk about the Content hypothesis. Just look at the way Adolescence went viral, compared to the foot-dragging on phone-free schools. Quite why the left has chosen this moment to indulge in scepticism that massive corporations might push harmful products on consumers for profit, ignoring any potential harms, I am unsure, but there we are.10
My personal view - though this is a belief I hold with only moderate confidence - is that the Medium hypothesis is more compelling.11 I’m not saying there’s no bad content out there - and I’m supportive, for example, of the recent age-limits on porn sites - but overall I think the biggest harm is being caused by the innate nature of these sites.
There are a few reasons for this:
The mental health issues - anxiety, depression, etc - that Haidt, Twenge and others report on track strongly with volume of social media use. They are a mass phenomenon, and seem to match much more with large numbers of people using an innately harmful product, than a much smaller number coming across harmful material. To the extent it is real and not overdiagnosis, the mental health epidemic is having a massive impact on everything, from schools to welfare to birthrates to just simple human happiness.
The loneliness epidemic again, appears to result simply from mass use of phones and social media - and is also one of the ways in which life simply seems to be getting worse for many people, leaving aside its societal impact.
I worry about the stupidogenic qualities of short-form video and the dawn of the post-literate society written about by
and others. A collapse in attention spans seems pretty bad for society as a whole.Polarisation is heavily driven by the algorithms - and it doesn’t need to be false, graphic or filled with racial slurs to cause problems. A feed which shows you endless pictures of crimes committed by immigrants, or climate change-related forecasts and weather events, can take you to a pretty dark place, even if every individual item was factually true. Even if we were able to perfectly and objectively identify ‘hate’ and ‘misinformation’ in a way we all agreed with,12 there’d still be plenty of problems.
I can feel the way smartphones and social media can affect me myself - and can here how friends talk about how it affects them. Even if it’s just simple FOMO, increasing anger or jealousy, or spending too long on a mind-numbing game, it’s abundantly clear in our own lives that these devices and apps are not entirely harmless.
Ultimately though, I don’t mind if you disagree with me on this - there’s certainly a lot of bad content out there, and that’s not a good thing either. Regardless of which hypothesis you think is dominant, there are harms on both sides.
But what does matter - whether you’re a parent, individual, campaigner, social media executive or government - is to think carefully about your theory of harm and take proportionate actions accordingly.
What, in other words, are you most worried about - and will what you’re planning to do address it?
Much of it collated on Haidt’s substack.
As a parent, one often has to make decisions without perfect evidence - I set out here why I think the evidence is sufficiently persuasive that caution around phones and social media is warranted.
Or, to be more precises, two thesis clusters.
Unless you happen to be the Secretary of State who was instrumental in passing the Act but then find yourself defecting to Reform, upon which you are required to have a Damascene moment and realise the real harms were the Acts we passed along the way.
And providing that you trust the parental controls work.
After all, it would be hard for algorithms to reinforce conspiracy theories if there were no conspiracy theories.
Or that neither are.
Civil service, academia, high status media such as the BBC, polite company more broadly.
This is slightly unfair to the left, as there are good people raising concerns about phones across the political spectrum, and this is - fortunately - not a completely polarised issue. But it is a bit peculiar that, to the extent there is a political slant, it is on the right that Haidt’s views have gained more traction, when it is normally the left who is concerned about Big Oil, Big Pharma, Big Tobacco, etc.
Is that highly convenient for me, as someone who cares a lot about free speech and is worried about governments and companies censoring views they don’t like as ‘misinformation’ or ‘hate’, and thinks this has happened to legitimate views in the past, for example in the pandemic or in the debates around trans? Why yes, it is highly convenient, and you can discount my views accordingly for how much you think my views as to which hypothesis I’d like to be true is influencing my beliefs, when you assess my arguments.
Ha, ha, ha.
What I find most interesting in all this discourse is that it is framed as "smartphones" when the actual discussion is usually then about social media, porn, or something else that is just as easily accessed on a tablet, laptop or PC. In fact in Adolescence the internet influence wasn't portrayed as mainly phone based, it was the computer in a bedroom. Nonetheless this 'medium' aspect is relevant because a phone is portable, ever present, and hard to see what someone else is accessing. But it isn't the whole story, and discussions that centre around smartphones feel like they are probably not clear enough on what they are proposing - the framing is off.
(That awareness of individual isolation being a concern is also the key to what we've attempted to do as parents - encompass all of the concerns with a simple principle of access in public. No electronics upstairs, so whatever they're doing, they're doing it where everyone else can at least vaguely see what sort of thing is happening. We have also had other rules of engagement which fall on the stricter side of things - but it's the staying involved with what they're up to that I think is the most powerful principle.)
For me, I worry about smartphone bans etc because for the weird kids it's a lifeline to be able to contact their people. Not to mention all the assistive apps that make life with vision and hearing defects, or existing in a second language, easier.