US Surgeon General Vivek Murthy proposed last week that Congress mandate warning labels for social media platforms, comparable to those that have been required on cigarettes since 1965 and on containers of wine, beer, and other alcoholic beverages since 1989.

“The mental health crisis among young people is an emergency — and social media has emerged as an important contributor,” Murthy wrote in a New York Times op-ed. “Adolescents who spend more than three hours a day on social media face double the risk of anxiety and depression symptoms.”

This isn’t the first time the surgeon general has raised an alarm about social media’s harmful effects. In a 25-page advisory last year, he reviewed the scientific literature, providing citations to scores of studies. He concluded that while social media use might have potential benefits — at its best, it strengthens a sense of community and connection with others, provides access to important information, and enables self-expression — it is also associated with serious negative effects. The most alarming of these are a worsening of mental health, the spread of dangerous self-harm “challenges,” exposure to hate-based content, and a decrease in both the quantity and quality of sleep.

“Nearly every teenager in America uses social media,” Murthy wrote at the time, “and yet we do not have enough evidence to conclude that it is sufficiently safe for them.” He suggested numerous steps that policymakers, technology companies, parents, and researchers could take to address and better understand the risks connected to social media use. Those recommendations, he wrote last week, “remain the priority.” Now he also is urging Congress to require a warning label — not because it would make social media safer but because it would “regularly remind parents and adolescents that social media has not been proved safe.”

Is this a good idea?

As a philosophical matter, I’m always wary of government encroachments on free speech. That includes government policies that compel private individuals and organizations to express a particular point of view. I’m also skeptical of letting government officials decide what we need to be warned against — both because such warnings can be patronizing and because they can be egregiously wrong.

There is a well-established principle that the First Amendment gives the government wider leeway to regulate commercial speech than personal or political speech. As the decades-long requirement of warning labels on tobacco and alcohol suggests, the Constitution does not bar the government from requiring vendors to disclose pertinent and accurate information. Thus there is no legal bar to requiring every wine or beer label to include a warning that “consumption of alcoholic beverages impairs your ability to drive a car or operate machinery” or to mandating that packs of cigarettes remind buyers that “smoking by pregnant women may result in fetal injury, premature birth, and low birth weight.”

But the rules are different when speech itself is being regulated.

Legal scholar Eugene Volokh, a professor at UCLA Law School who specializes in the First Amendment, emphasized in an email that the Supreme Court frowns on mandatory disclosures or warnings that may impede fully protected speech. He pointed to a 1988 case in which the justices struck down a requirement that charitable fundraisers disclose to potential donors “the percentage of charitable contributions collected during the previous 12 months that were actually turned over to charity.” Donors might well benefit from knowing that information. But since fundraising appeals are incontestably free speech, the government cannot compel charities to announce that information before every appeal. Otherwise, it would be interfering with the charity’s right to freely express itself in its own way — something the First Amendment does not allow.

In Volokh’s view, social media platforms are similarly protected. The very purpose of those applications is to facilitate speech and expression. He predicts that a law forcing them to warn users about the potential dangers of such speech — perhaps by requiring a pop-up that materializes every time an app is opened, or a permanent banner at the top of a web page — would be struck down as unconstitutional.

The dangers of social media use are still being debated. Unlike the connection between smoking and cancer or between drinking and impaired driving, there are plenty of unanswered questions about the harms associated with social media. It isn’t even obvious what would constitute “social media” for purposes of a warning label. The study cited by the surgeon general — the one indicating a doubling of anxiety and depression among adolescents who spend three hours daily on social media — cast a very wide net. It included teens who said they visit “social networks like Facebook, Google Plus, YouTube, MySpace, Linkedin, Twitter, Tumblr, Instagram, Pinterest, or Snapchat.” Those sites are far from interchangeable! Does someone who spends three hours a day watching documentaries on YouTube or perusing home décor ideas on Pinterest suffer the same potentially negative effects as an Instagram or Snapchat addict?

I am by no means complacent about social media, not least because of the alarms sounded by some of the people intimately involved in developing these networks. As far back as 2017 — to mention just one example — former Facebook executive Chamath Palihapitiya said that he feels “tremendous guilt” for his role in creating “tools that are ripping apart the social fabric.” He had come to realize, he said, that “the short-term, dopamine-driven feedback loops that we have created are destroying how society works. No civil discourse; no cooperation; misinformation; mistruth.”

Such concerns by now are widespread and bipartisan. At a January congressional hearing about online safety and abuse, Republicans and Democrats alike excoriated Big Tech CEOs for the harmfulness of their products. In a column last month, I expressed the hope that, just as young people in recent generations gradually turned away from smoking cigarettes, the craving of today’s adolescents to be online will likewise fade over time.

Still, we should be cautious about mandatory warning labels. They are no panacea, as Murthy himself acknowledges. The law cannot require vendors to slap warnings on every potentially dangerous product and service. If that is true of consumer goods, how much more so is it true when the product or service involves words and ideas?

The surgeon general’s proposal is a fine conversation starter. His idea is well worth discussing. But Congress ought to be in no hurry to enact it into law.