How We Write A Law
Reading the evidence on social media and the mental health of minors.
I came into this piece wanting to write about social media and mental health, platform by platform. I pulled the studies, started reading, and realized I was writing the wrong article.
The more useful piece was about how to read this kind of evidence in the first place. Science literacy improves our decision making in every facet of life. It matters when we vote, when we parent, when we sit on a school board, when we read a headline that tells us coffee causes cancer one week and prevents it the next. For readers who already know the difference between a cross-sectional correlation and a randomized experiment, this is a refresher built around a question that matters. For everyone else, it is an introduction to a skill worth keeping.
By the end you will have a tool for reading science, a clear picture of what we actually know about social media and kids, and a framework for state legislation along with how you can personally advocate for it.
Three studies are particularly illustrative, and each shows a different kind of strong evidence.
Luca Braghieri, Roee Levy, and Alexey Makarin had an unusual opportunity. Facebook rolled out across US college campuses one school at a time between 2004 and 2006, which meant some campuses had Facebook and others did not, with the timing varying for reasons that had nothing to do with student mental health. The researchers could compare students before and after Facebook arrived at each school, and compare schools that already had Facebook to schools that did not yet. The arrival produced measurable increases in depression and anxiety. Students predicted to be most susceptible saw the largest effects, and more of those students used campus mental health services after Facebook arrived. The mechanism evidence pointed to unfavorable social comparison, which tells legislators not just that the harm exists but where to aim a bill that addresses it. Their work was peer reviewed, meaning other researchers in the field read the paper looking for problems, checked whether the methods supported the conclusions, and required changes before it could be published.
The second piece of evidence is what scientists call a meta-analysis, which is essentially a study of studies, pooling results from multiple experiments to get a larger combined sample and a more reliable estimate of the true effect than any single study can give you. In 2025, researchers published a meta-analysis of randomized controlled trials on social media reduction, pooling results from multiple experiments where participants were randomly assigned to either reduce their social media use or continue as usual. The combined effect on depression translates into something concrete. If you pick one person who cut back on social media and one person who did not, the one who cut back will have lower depression about 57 percent of the time. Without the intervention, it would be a coin flip at 50 percent. That seven-point shift sounds modest applied to any single pair of people, but across hundreds of millions of users it is the kind of number that justifies legislative attention. The finding remained statistically significant after the researchers adjusted for publication bias, the tendency of journals to prefer positive findings over null ones, which can inflate apparent effects if it is not corrected for.
A third body of evidence shows what happens when scientists ask the same question with different methods and the answers agree. School phone bans have been tested by a randomized controlled trial of 17,000 college students, a quasi-experimental study of more than 130,000 students in Florida, and frontline observation from teachers in districts that have adopted bell-to-bell policies. Academic gains are concentrated among lower-performing students, which matters because educational interventions that help the kids furthest behind are among the most valuable a state can fund. The mental health evidence is thinner, partly because kids who lose phone access during the school day tend to make up much of the lost screen time after school, which dilutes the overall reduction. When studies using different designs all point to the same conclusion, we have a finding worth taking seriously even if any individual study has limitations.
Science is much better at telling us what is happening than why it is happening. We can measure that Facebook’s arrival on campuses increased depression. We can measure that cutting back on social media reduces it. We can measure that school phone bans help lower-performing students. What we cannot see clearly is what is happening inside the platforms themselves. Which algorithmic choices drive harm? Which design features push compulsive use? Which content types do the most damage to which kids? The platforms refuse to share the internal data that would let researchers answer those questions with precision. Every study above had to work around that refusal.
Public health policy has never required perfect understanding of mechanisms before acting. We knew tobacco caused lung cancer for decades before we understood the cellular pathway. The what is sufficient for making informed public policy that benefits us broadly.
Other countries are not waiting for perfect understanding. Australia banned under-16 accounts on Facebook, Instagram, TikTok, Snapchat, YouTube, Reddit, X, Threads, Twitch, and Kick. The law took effect December 10, 2025, with penalties on platforms up to 49.5 million Australian dollars for non-compliance. Meta removed roughly 4.7 million minor accounts before the deadline. France passed its own under-15 ban on January 26, 2026, by a vote of 130 to 21, set to take effect that September. Denmark followed in November 2025 with an agreement to ban under-15 access. Portugal’s parliament approved an under-16 bill in February 2026. Germany commissioned a study to be delivered in fall 2026 and is debating an under-14 ban. The EU is considering bloc-wide minimums.
The European approach carries a serious cost that the headlines do not always cover. To enforce an age-based ban, the platform has to verify everyone’s age, not just the kids. Every adult user has to hand over a government ID, a face scan, or a credit card to a company that has spent twenty years building a business on harvesting and monetizing exactly that kind of information. Each verification mandate creates a new database of identity documents that will be breached eventually. The Tea app leak in 2025 exposed face scans and government IDs of users who thought they were just verifying for a dating service. UNICEF Australia and major privacy organizations have argued that the better fix is regulating what platforms can do to the kids who are on them, rather than building a national identity verification system that affects every adult.
In the United States, more than a dozen states have passed age-verification or parental-consent laws since 2023. Most have been blocked or partially blocked by federal courts on First Amendment grounds. NetChoice, the industry group whose members include Meta, TikTok, and X, has built a litigation machine that challenges every state law in this space. The pattern in the rulings is clear: courts have struck down laws that try to keep kids off the platforms, and have let stand laws that tell platforms what they cannot do to the kids who are already on them. The first kind treats minors as the problem and gets blocked, because minors have free speech rights too. The second kind treats the platforms as the problem and survives, because regulating a company’s business practices is something government has always been allowed to do. California’s SB 976 and New York’s SAFE for Kids Act take the second approach, regulating algorithmic addictive feeds rather than account access. Those are the bills still on the books.
While Congress has not acted, individual states have. Every meaningful regulatory experiment in the United States on this question for the last five years has come from a state legislature or a state attorney general.
I wanted this piece to end with something readers could actually do. So I sat down and tried to design a piece of legislation that does three things at once. It has to protect kids in ways the science supports. It has to survive the First Amendment challenge NetChoice will bring, because they bring one against every state law in this space. And it has to generate enough revenue that a state legislator sees a reason to champion it beyond the moral argument.
The science constraint shapes what the bill targets. Braghieri Levy and Makarin pointed to social comparison as the mechanism behind Facebook’s effect on student mental health, which means algorithmic amplification of appearance-based content to minors deserves specific attention. The randomized trials show that reducing compulsive use produces measurable improvements in depression and anxiety, which means the design features that drive that compulsion deserve specific attention too. The science supports protections aimed at the specific mechanisms producing harm.
The First Amendment constraint shapes how the bill is written. Laws that block minors from accessing platforms keep losing in court because minors have free speech rights too. Laws that regulate how a commercial company designs its product survive because that is commercial regulation, which government has always been allowed to do. The bill therefore targets specific platform conduct toward accounts the platform has reason to know belong to minors. No algorithmic amplification of content the platform knows is harmful to a minor, such as eating disorder content, self-harm content, or content related to suicide methods. No variable reward features on minor accounts, such as autoplay, infinite scroll, visible like counts, and streaks. No push notifications to minor accounts during defined sleeping hours. No targeted advertising to minor accounts based on inferred psychological state. The bill does not force every adult to verify their identity to use the internet. It tells platforms what they cannot do to minors using their products.
The third piece is how the bill makes money. When we draft model legislation at The Existentialist Republic, we look for funding mechanisms because we need both the moral and practical elements to build robust legislation. Funding lets the state enforce the law, pay for the services that address the harms the law is responding to, and in some cases tax the harmful behavior itself as a disincentive, the way cigarette taxes both fund public health programs and make smoking more expensive. The mechanism here is a private right of action for parents, with statutory damages set per violation. This is modeled on the Telephone Consumer Protection Act, the consumer protection statute that has generated billions in plaintiff recoveries since 1991 because it created an incentive structure that lets lawyers take cases without charging the client. The lawyer covers the costs up front and only gets paid if the case wins, with the fee coming out of the recovery. Fee-shifting is what makes this work. Without it, only wealthy parents can sue. With it, every parent can. Two additional layers stack on top: civil penalties paid to the state for each violation, modeled on Florida’s penalty structure but attached to design violations rather than the account-access provisions that got Florida’s law blocked, and the option for the state attorney general to sue on behalf of harmed minors as a class, modeled on the 1998 tobacco master settlement and Texas’s 1.4 billion dollar settlement from Meta in 2024.
The bill regulates platform conduct rather than user access. It puts enforcement in the hands of parents and their lawyers rather than depending on a single elected official to act. It creates the financial incentive structure that has forced corporate behavioral change in this country since the tobacco settlement. And it is small enough to pass quietly in a single state, after which the math changes for every other state that sees the recovery.
If you want to do something as an activist today, you can email this article to your state representative and ask whether they would propose a bill like the one described above in the next legislative session, or at minimum publicly endorse one. You can find your state representative and their contact information at openstates.org. If they say yes, reach out to The Existentialist Republic at AC@theexistentialistrepublic.com and we will supply a full bill drafted and tailored to your state’s statutes.
But just so you know, we need 10 people to become subscribers every article to keep this publication running.
We have millions of readers and tens of thousands of activists participating in the daily calls to action we share.
Don’t let this be the reason you skip a meal or miss rent.
But if you can manage it, will you become one of the 10 today and keep this all alive for your fellow makers of change?
Conservatism: America’s Personality Disorder — physical copy / free download
Intro to Soft Secession — physical copy / free download
Oppositional Federalism and You — physical copy / free download
Toppling Tyrants: A Field Guide to Dismantling American Fascism — physical copy / free download
Grab Them By The E.A.R.R.: How to Get Politicians to Do What You Want — physical copy / free download
More Free downloads:
Soft Secession: 100 Policies That Pass
Being Dangerous: Go From Activist to Operative
All Four Completed Model Legislation Bills
The Opposition Guide to Tax Warfare
Six-Panel Soft Secession Brochure
Prosecution Memo: Jonathan Ross
The Opposition Guide to Tax Warfare
Six-Panel Soft Secession Brochure
As a Water/Plumbing Subject Matter Expert, I have been involved to the way Building Codes and Water policies are established. Engineers and science are used, not FaceBook MetaData. There is not 1st, Amendment for plumbing, so my world is a bit more straightfroward. But, manufactures do get in the way so to speak by having clout and a hand in the lobbiests.
It is too bad that we as a nation have been told that the Government is the problem, when the special interests are what sours the end results of government policy.
GenX reflex and now ear worm:
Schoolhouse Rocky
A chip off the block
Of your favorite schoolhouse
Schoolhouse Rock!