Background
Since 2012, rates of anxiety, depression, self-harm and sleeping disorders amongst young people in the UK have risen alarmingly, and this rise has been associated with the rapid adoption of social media in a growing body of research. High-profile cases of suicide, online grooming, cyber-bullying, and self-harm communities facilitated by online platforms have brought to public attention the ruinous consequences of social media overuse and addiction.
After Australia introduced a social media ban for those under sixteen years of age in 2025, many voters and politicians in Europe have been eager to implement a similar policy to curtail the woes experienced by children online. Social media bans are currently being prepared and passed in Denmark, France, Germany, Greece, Malaysia, Slovenia and Spain, and there is strong support for similar steps to be taken in the UK.
Seen as a collective action problem, the targeted ban for young people aims to forcefully dismantle the strong network effects that lure the most vulnerable demographic into a potentially malicious online environment. However, a ban is not the only solution. A blanket ‘Dopamine Tax’ aimed at distorting the incentives behind compulsive usage could be implemented in conjunction with a youth social media ban. A policy package involving both strategies could more effectively decrease the harmful effects of social media addiction which are evidenced across the entire user base.
A Pigouvian Proposition
One method of ameliorating the damages caused by the consumption of harmful goods is implementing a tax. By raising the cost of production or consumption, taxation reduces the quantity consumed and the negative effects of consumption. A tax which has the intention of benefiting society through the reduction of harmful consumption is called a Pigouvian tax when these damages come as externalities, or a sin tax when the process of consumption itself is seen as harmful. The Dopamine Tax, named after the neurochemical which is manipulated by the design of social media platforms to create compulsive tendencies in users, would be a mixture of a sin and a Pigouvian tax, for high rates of anxiety and depression are inherently undesirable for society and also carry negative externalities (e.g. decreased productivity, and higher costs of treating associated disorders). The aim of a Dopamine Tax would be to make the enterprise of compulsion and addiction to social media more costly for platforms or users, which would lead to a reduction in excess engagement, and a decrease in corresponding deleterious societal effects.
In practice, this tax would either be levied as a tax on sales or as a tax on usage, both of which have precedent within the UK tax system. A tax on usage would likely be more successful in this case as it targets the harmful aspect of addiction more precisely, and risks creating fewer perverse incentives. We will, however, firstly consider the potential structure of a tax on sales. The sale of advertising space by social media companies could be taxed to a greater degree, with an additional charge on top of the existing Digital Services Tax on gross revenues online. This increased cost could induce social media companies to reduce the supply of their services, but there are many issues with the sales tax approach in practice. Firms could just increase the price of advertisements, passing on costs to the advertisers and decreasing the demand for online advertising, not affecting usage or addiction. This distortion decreasing the demand for advertising space might even push social media companies to increase hours spent on their platforms to make up for the loss in profit, directly opposing the intended effect of the policy.
Taxing social media usage directly is a more favourable alternative because it creates more precise incentives to reduce the harmful effects of social media addiction. The government could request engagement metrics (e.g. hours spent per session) from social media companies, and charge levies based on the reported data. There is precedent to this approach in the UK tax system; Fuel Duties are charged per litre produced/imported for domestic use, and the Landfill Tax is set per tonne of waste disposed. Each of these taxes requires the collection and divulgence of data which is not contained in the market transaction. Similarly, the Dopamine Tax would require social media companies to track and report the engagement metrics of its users (which they already collect in detail internally), and like in the other cases this data would be subject to HMRC inspection.
A tax based on usage could then either be set on the firm or the user, and which approach is more favourable depends on personal views about the taxincidence theorem. In practice, it is likely that a tax imposed on the user would have a stronger effect on usage, since it is hard for a firm to reduce the supply of a good like social media. A tax on the social media consumer would be charged per session spent on an addictive platform, and would be proportional to the time spent in each session. This setup involving immediate, frequent payments would likely have the greatest impact on addictive usage, as it would both reveal the extent of one’s engagement and penalise excessive time spent. The present zero pecuniary cost to users consuming addictive online products distorts the equilibrium quantity above the social optimum as has been evidenced by the medical literature (see Ahmed et al, 2024, Journal of Affect Disorders) . By charging for the use of compulsive online services, the Dopamine Tax would internalise the externalities associated with exorbitant use which are not presently accounted for.
There are nevertheless a collection of issues with imposing a consumption tax on social media usage. Firstly, there is no precedent for collecting social media usage metrics for tax purposes in the UK, neither is there precedent for charging consumers based off of usage metrics (Fuel Duties and the Landfill Tax are charged to firms). Creating such administrative structures would undoubtedly be costly for HMRC. Each relevant social media account would also need an associated bank account from which a charge could be taken, a further administrative point of friction, though one which could be transferred from the government to the firms if payments are made through the social media companies. The tax would also be regressive, and likely unpopular amongst consumers who do not recognise the intended effects of the policy. While a bureaucratic burden, a Dopamine Tax would strongly disincentivise the unnatural indulgence of artificial compulsions manipulating and harming users of social media in the UK.
Taxation vs Prohibition
The ‘Social Media Ban’ is already a very popular policy proposal in the UK: the Conservative party has explicitly pledged to ban social media for under 16s if they are re-elected, the House of Lords backed an amendment to implement a similar block, over 60 Labour MPs have signed an open letter urging the prime minister to introduce a prohibition, and there are other petitions and parent-led campaigns calling for the restriction of access for young people. Supporters of the ban say that it would both make it harder for individual minors to access social media sites (the ban would likely involve some ‘age-assurance’ steps to be implemented on all relevant sites like in the Australian bill) and it would improve the leverage of authoritative agents like schools and parents over their children who can use the illegality as a reason to justify restricting access. There are of course issues with the social media ban, such as the tendency for young people to sidestep the ‘age-assurance’ barriers (ID upload, bank-verified credentials, or facial age estimation), bypassing the policy entirely. This issue of avoidance is shared with the Dopamine Tax if the tax is confined to the population under 16 years old. Other than the possible avoidance of the ban, critics say that the policy might also limit the freedom of expression for young people, and that social media might not be responsible for all the negative trends associated with it over the last ten years. These are issues which need further debate and more evidence to be resolved.
Importantly, however, there is significant evidence that social media overuse is associated with poor mental health outcomes in adults as well as children. Heavy consumption of social media has been associated with higher rates of depression, anxiety, stress, loneliness, and poor sleep in the adult population, and if this evidence is strong enough, then an optimal response would be a policy package involving a combined social media ban for young people and a blanket Dopamine Tax on the whole user base. Then the freedom of access to social media would be preserved, while the incentives to engage excessively are curtailed, even for young people who circumvent the ban. There is strong precedent for this tiered approach to regulate harmful industries without restricting freedom of access, for example, it is illegal to invite, cause, or permit under-18s in the UK to gamble in almost any form, but there are also specific betting and gaming duties on most forms of gambling for the rest of the population, and this tiered approach also applies to the consumption of alcohol and tobacco. By both banning access for under-16s and introducing a population-wide levy on the compulsive industry, the government could meaningfully reduce social media addiction in the UK and its associated harms, ushering in a new era of responsible use.