This essay appears in print in Economics After Neoliberalism.

The First Amendment has long been celebrated as the guardian of our democracy, a protector of the robust public discourse essential to self-determination. Today, however, the First Amendment is being shaped into something very different: a guardian of the interests of private companies that resist democratic regulation.

What has largely escaped public notice, is that the Supreme Court has begun reshaping the First Amendment into a tool to broadly undermine the regulatory state.

Many are familiar with Supreme Court cases such as Citizens United v. Federal Election Commission (2010) that use the First Amendment as a weapon against campaign finance restrictions. Others may also have heard of the 2018 decision Janus v. AFSCME, which struck a grave blow to public sector unions. In that case, a 5–4 majority flatly reversed more than 40 years of precedent, barring certain dues (so-called “fair-share fees”) on the grounds that they conflict with union members’ speech rights. What has largely escaped public notice, though, is that the Supreme Court has also begun reshaping the First Amendment into a tool to broadly undermine the regulatory state. Today, most Amercans are clamoring for more robust regulation of markets. But what companies cannot win through democratic politics, they are hoping to win from increasingly conservative courts, with First Amendment speech protections as an increasingly powerful weapon in their arsenal.

In a 2017 case, for example, five merchants challenged a New York law preventing businesses from adding a surcharge on credit card purchases. They did not argue that it was bad for business, or bad for consumers, or bad public policy, or a restriction on contractual liberty. Instead, they argued it violated their speech rights by regulating how they communicated with customers about prices. The Supreme Court agreed, sending it back to the lower court for review. Consider, as well, laws requiring companies to disclose whether they are trading in conflict diamonds, or mandating graphic warning labels on cigarettes. They, too, have been overturned for violating companies’ speech rights. As Justice Stephen Breyer wrote in a case last term, “Because much, perhaps most, human behavior takes place through speech,” the Supreme Court’s new approach to free speech law threatens a wave of new lawsuits “over the constitutional validity of much, perhaps most, government regulation.”

Particularly troubling are new First Amendment cases that chip away at the power of the U.S. Food and Drug Administration (FDA). Do drug companies have to provide scientifically reliable evidence for their claims to the agency before marketing a medicine? Do e-cigarette companies have to show that their products have health benefits before marketing them that way? For as long as we have had a regulatory state, these kinds of questions have been the domain of Congress and regulators. Today, courts are increasingly treating them as constitutional questions, answering them through a First Amendment doctrine that treats many forms of regulation as the illegitimate coercion of speech, rather than as the democratic prerogative of a public seeking to protect itself from the risks of deception and harm inherent to market society.

We have been here before, during the dark history of the Supreme Court’s Lochner Era (1890–1937). For more than four decades—extending well into the Depression—the Supreme Court effectively wrote laissez-faire economics into the Constitution by repeatedly striking down laws regulating working hours and conditions or seeking to stabilize the national economy. According to the Court, these laws and ordinances were illegitimate intrusions on the rights of contract.

Do drug companies have to provide scientific evidence for their claims? This used to be the domain of congress and regulators. Today, courts treat them as constitutional questions.

In 1937 the Court—wildly out of step with public sentiment and in the face of Franklin Roosevelt’s court-packing threat—retreated, and a new settlement was forged: courts would use their authority to reinforce democracy, focusing on voting rights, speech protections, discrimination, and other protections for vulnerable groups, leaving ordinary socioeconomic legislation to elected officials. The current Court, armed with a new interpretation of the First Amendment, is eroding that settlement. In the process, it is undermining our collective authority to solve critically important social problems and casting into doubt the fundamental structure of our regulatory state.

Consider the case of Alfred Caronia, a sales representative for Jazz Pharmaceuticals, which manufactures the drug Xyrem. The main ingredient of Xyrem is a form of GHB, a central nervous system depressant notorious for its use as a date-rape drug. The drug has FDA approval only for the treatment of two kinds of narcolepsy in adults, and drugs with small patient populations— especially drugs as dangerous as Xyrem—generally do not sell well. So Jazz Pharmaceuticals did what many other companies in its shoes have done: it launched a marketing offensive. Alfred Caronia was one of the ground soldiers in this effort—and the one unlucky enough to be caught on tape in a federal sting. According to the trial transcript, he urged doctors to prescribe Xyrem for a long litany of ills: fibromyalgia, chronic pain, fatigue, restless leg syndrome. There is “no safer drug,” he declared, even recommending it for children as young as four. Caronia was convicted in 2008 under the Food and Drug Cosmetic Act (FDCA), New Deal legislation from 1938 which gave the FDA its powers. His efforts as a sales representative to market off-label uses of Xyrem were found to violate the FDCA’s prohibitions on “misbranding” drugs.

What Caronia did was illegal because, since the 1960s, the FDA has required that companies prove, before marketing a drug, that it is safe and effective for a specific use. (Doctors can prescribe drugs for off-label uses, but manufacturers were not supposed to market them for those uses.) This is because a drug’s side effects and efficacy can change in different patient populations—for example, in the young or the old—and for different kinds of uses. Addiction or overdose may be tolerable risks for a painkiller intended for terminal cancer patients, but not for those in mild pain.

Our modern drug regulation and innovation system relies on these marketing restrictions. Because they cannot promote medicines without showing that they work, companies conduct and hand over the clinical research the FDA and medical experts use to evaluate a drug’s effects. This powerful system of premarket review grew in response to a series of disasters, each revealing new risks in laissez-faire approaches to markets in food and medicine. By the early 1900s, markets in food and medicine had become increasingly anonymous and far-reaching, and profit-motivated sellers had found ways to hawk their wares that gravely endangered the public. While the patent medicine industry seized upon newspaper advertising, promising miracle cures but usually delivering nothing more than alcohol, opioids, or worse, food marketers discovered that suffusing food with borax and formaldehyde lengthened shelf life. Upton Sinclair’s The Jungle (1906) still provides the most indelible image of modern capitalism without a regulatory state: workmen slipping into a vat and emerging from the factory pressed into a packet of “Durham’s Pure Leaf Lard.”

What caused this shift in the Court’s opinion, and where are we headed next?

Recognizing that only a federal regulatory agency could assert control over the risks of these newly nationalizing markets, Congress created the FDA in 1906. Sellers would now be required to accurately label their wares and disclose ingredients such as opioids. After a patent medicine called Elixir Sulfanilamide killed dozens around the country, Congress in 1938 gave the agency authority beyond labeling and disclosure requirements, allowing it to bar dangerous products from the market. Then came Thalidomide. Prescribed widely in Europe for morning sickness in pregnancy, the drug turned out to cause horrible birth defects. LIFE published images of children with shortened or missing limbs, and the press celebrated the story of a regulator at the FDA who, through extraordinary persistence, prevented the drug’s approval in the United States. The outcry turned into a mandate for one of the most muscular regulatory agencies in the country and the world. By the 1960s, it was settled: pharmaceutical companies would have to provide robust, scientifically reliable evidence before marketing their wares.

Hence Caronia’s legal troubles: he was promoting Xyrem for uses for which it had never been approved and so violated the FDCA’s prohibitions on “misbranding” drugs. In the end, though, he was saved by the Supreme Court’s ruling in a new First Amendment case, Sorrell v. IMS Health (2011), to which I’ll return at length. After that decision, Caronia appealed his case, arguing that he was simply exercising his right to free speech when he marketed Xyrem for off-label uses. The U.S. Court of Appeals for the Second Circuit concluded that the Supreme Court’s new precedent was in conflict with six decades of FDA regulatory authority. Caronia won his appeal, and the First Amendment was at war with democratic regulation.

What caused this shift in the Court’s opinion, and where are we headed next?

The first amendment declares that Congress “shall make no law. . . abridging the freedom of speech, or of the press.” Taken literally, these words would render government as we know it impossible. But courts have never taken them literally. People are penalized every day—they are put in jail, they lose their livelihoods—because of the words they utter. Doctors are liable for what they say to their patients, and if you lie to investigators you can be charged with perjury. The government bars witnesses in lawsuits from giving expert testimony about phrenology and astrology. Companies are required to disclose certain risks to investors and are subject to the law of fraud, a crime often consummated entirely through speech. To bar a sale of something is to restrict speech, because offers for sale are speech. And what does the law of contracts regulate if not words?

Buried here is a basic point about legal interpretation. Does a ban on “vehicles in the park” prevent a parks commission from permitting baby strollers or a decommissioned tank? You cannot decide unless you have a sense of the purpose of the ban. This general point applies to the First Amendment, too. Courts need a sense of the purpose of the constitutional free speech guarantee in order to determine how far they should reach into the power to legislate. Courts only began to assert free speech protections in a serious way in the 1920s, and since then have often emphasized the importance of speech protections for democratic governance.

We protect speech in order to facilitate “public discourse.” First Amendment law is centrally about protecting the formation of public opinion because, as Robert Post has described, only if we can freely speak in public can we understand and render our government legitimately our own. As Justice Louis Brandeis once put it, “The right of a citizen of the United States to take part, for his own or the country’s benefit, in the making of federal laws and in the conduct of the government, necessarily includes the right to speak or write about them; to endeavor to make his own opinion concerning laws existing or contemplated prevail; and, to this end, to teach the truth as he sees it.”

One consequence of the Court’s democratic focus was that, for decades, it simply did not apply the First Amendment to commercial speech—just as today it still does not apply it to the rules of evidence or to the law of contracts. As Justice Hugo Black, one of the Court’s early champions of the modern conception of free speech, succinctly put it, the First Amendment had nothing to do with a “merchant who goes from door to door selling pots.”

This all changed in the 1970s, when the Supreme Court decided that commercial speech in fact did warrant constitutional protection. In striking down a state law barring pharmacists from advertising the prices of prescription drugs, the Court announced that “the free flow of commercial information is indispensable . . . to the proper allocation of resources in a free enterprise system” and “also indispensable to the formation of intelligent opinions as to how that system ought to be regulated or altered.” The First Amendment protects commercial speech, the Court declared, not out of a concern for the speakers, but because listeners need that information for public decision-making. A new spirit was on the move through our law.

In 2011, the legal argument shifted from its earlier focus on citizens’ need for information and toward a newfound solicitude for the rights of corporate speakers.

This listener-based argument felt a little strained, and no one was more unsparing about its shortcomings than Justice William Rehnquist, a Richard Nixon appointee and leading conservative: Americans may “regard the choice of shampoo as just as important as who may be elected to local, state, or national political office,” he scoffed, “but that does not automatically bring information about competing shampoos within the protection of the First Amendment.” What kind of world was the Court ushering in, he asked? One where a pharmacist might run an ad in the local paper saying, as he put it, “Don’t spend another sleepless night. Ask your doctor to prescribe Seconal without delay”?

You likely know how this story ends. That decision, Virginia State Pharmacy Board v. Virginia Citizens Consumer Council (1976), did indeed help to unleash direct-to-consumer advertising of drugs, by ensuring that something that is illegal in most of the world could not be barred here. It also gave ammunition to industry-funded conservative legal groups such as the Washington Legal Foundation. Though the cases received little attention at the time, the group used the new commercial speech doctrine to slowly deregulate drug and dietary supplement advertising.

It was not until the 1990s, however, that conservative justices realized the true power of the First Amendment to undermine the regulatory state. By 1995 Rehnquist, who had been so appalled at the notion of crass drug advertising in the 1970s, joined the Court in invalidating restrictions on putting alcohol content on beer labels. In a wave of subsequent decisions, a new polarity emerged, with the Court’s conservative wing deploying the First Amendment in more and more cases to strike down ordinary economic regulation.

In 2011 the commercial speech train jumped the tracks. The legal argument shifted decisively from its earlier focus on citizens’ need for information and toward a newfound solicitude for the rights of corporate speakers. In Sorrell v. IMS Health, the Supreme Court sided with pharmaceutical companies against a Vermont law passed to protect doctors from intrusive pharmaceutical marketers. The Vermont law required doctors to affirmatively consent before the details of their prescribing practices could be sold to data miners and then used by the Alfred Caronias of the world to badger them into prescribing more and more. The law had some sensible exceptions though—academic researchers could access the data more easily, for example.

The exceptions killed the rule. Fixating on the fact that some parties could more easily access and use data about doctors’ prescribing practices than others, the Supreme Court found that the Vermont law was not a sensible protection of doctors and patients but was unconstitutionally discriminating against companies. The law not only overtly disfavored marketing as a kind of speech and so regulated by reference to “content”—long a warning flag where political speech regulations were concerned—but also “disfavors specific speakers, namely pharmaceutical manufacturers.”

What could this possibly have to do with upholding democracy? This was one of the critical questions posed by Sorrell v. IMS Health, and the majority opinion, written by Justice Anthony Kennedy, simply sailed past it. The majority also routed around the broad framework of analysis that the Court had used for decades to distinguish restrictions on commercial speech from restrictions on political speech. Under that framework, lawmakers can, for example, bar false and misleading commercial speech outright. In contrast, lies are protected in politics, and Congress will rarely be able to impose restrictions turning on a political viewpoint: for example, applying restrictions to pro-life arguments that are not applied to pro-choice ones. Even true and non-misleading commercial speech can be regulated, if the government had a substantial interest and the law directly advanced that interest.

In Sorrell, the Court cast doubt on this historical approach, introducing into the mix a new set of questions about content neutrality and viewpoint discrimination. In dissent, Justice Stephen Breyer sounded the alarm: the Court was opening an avenue to meddle in an extraordinary range of legislative and regulatory decisions. Regulators often train their attention on some subjects and not others. Cosmetic companies, for example, might be required to substantiate the claim that a “product contains ‘cleansing grains that scrub away dirt and excess oil’” while “opponents of cosmetics use need not substantiate their claims.” Appliance companies might be required to publicize ways to save energy, though their industrial counterparts are not. Or, the FDA might forbid drug companies from promoting drugs for unapproved uses, though academic researchers may recommend for or against the same uses.

The First Amendment has become a new kind of ‘guardian of our democracy’—a guardian of we the pharmaceutical companies, we the gun salesmen, we the e-cigarette makers.

Untethered from its historical moorings, Breyer argued, this new First Amendment threatened to undermine the historical deference given to legislatures to govern “ordinary commercial or regulatory legislation.” He was referring to the great Lochner settlement, the one that brought the Court back from the brink of disaster in the 1930s. From the ashes, constitutional lawyers had to construct a new sense of what courts were for and a new argument for why courts should be allowed to overturn the judgments of elected bodies. They landed on the idea that an essential part of their purpose was to protect democratic self-government, by protecting speech, ensuring voting rights, and making sure that all voices—including minorities who might be excluded—were heard in the process. The way they distinguished the new order from the constitutional cliff that Lochner had driven the Court toward was in what they wouldn’t do: the Court no longer had a role in second-guessing the judgment of Congress when it passes ordinary socioeconomic laws. That was outside of the competency of courts and a domain left to the people.

The new First Amendment is undoing this settlement. And implicit in the language of discrimination that has consolidated among the conservative justices lies an account as to why: the problem is that the people are disrupting efficient markets and discriminating against corporations. To protect against this, the First Amendment today has become a new kind of “guardian of our democracy”—a guardian of we the pharmaceutical companies, we the gun salesmen, we the e-cigarette makers.

The cases now being litigated by conservative legal groups and industry paint a grim picture of where we might be headed. For years, conservative legal groups have railed against occupational licensing. These are the rules that require training and exams for teachers, lawyers, and doctors and that, in extreme cases, permit professionals to be sanctioned, including by losing their ability to practice. All of these restrict efficiency from the perspective of a neoclassical economics that assumes markets are replete with good information and that disavows any role for groups in the cultivation of something like virtue or professional expertise.

If you want to wage a successful litigation campaign, the rule is to start small, with sympathetic plaintiffs. In the licensing contest, the camel’s nose is made of tour guides and tire engineers. To impose licensing requirements on these professionals, conservative advocacy groups argue, is to prohibit speech about monuments and tire treads (content regulation). And it is discriminatory, because tour guides and tire engineers are forbidden to say things that other people are allowed to discuss freely. Federal appeals courts in the last few years have come to agree.

Last term, Justice Clarence Thomas put the pedal to the floor with one sentence in a case about crisis pregnancy centers. Speaking for a majority of five, he declared, “This Court’s precedents do not recognize such a tradition for a category called ‘professional speech’.” In a pen stroke, he wiped out decisions in lower courts that had for years treated professional speech regulations as exempt from First Amendment review.

Next stop: if the state cannot prevent tour guides from speaking without a license, why can it stop teachers and veterinarians? And then, why can states prevent someone from giving medical or legal advice without a license? These are precisely the questions coming to the fore, through cleverly designed cases, each intended to take us another step down this road, eventually leading to the explicit deregulation of the professions. Even the Lochner court did not tread this far. Over and over, at a time when laissez-faire reigned, it yielded to states’ authority to regulate professions such as medicine. “There is no right to practice medicine,” the Court declared, “which is not subordinate to the police power of the States.”

Corporate disclosures are a target, too. Can we compel companies to inform us, for example, what is in their products and the conditions under which they are made? Such reporting was long considered unproblematic, because protections of commercial speech were supposed to be about protecting listeners, not speakers. You and I cannot be forced, in contrast, to salute the flag or to avow even true statements about our political beliefs. That, the Court has long concluded, would implicate our autonomy as citizens. But even this distinction is eroding as courts are coming to treat corporations as speakers whose integrity can be undermined if they are forced to speak in a voice that is not theirs. “Uncontroversial” factual disclosures are allowed, but nothing that smacks of “opinions.”

Every labeling and advertising law for every commodity is now in the crosshairs. Calorie disclosures: do they undermine a company’s right to convey a happier image about its products?

You don’t need a high-priced corporate lawyer to drive a wide-bodied truck through this opening. A few years back, the D.C. Circuit Court of Appeals struck down a requirement that companies report whether their gems were obtained from conflict zones in the Democratic Republic of Congo (DRC). And tobacco companies won a major victory striking down the FDA’s graphic warning labels on cigarettes. Both measures, courts concluded, illegitimately compelled commercial speech. If democratically elected officials cannot require a Jehovah’s Witness to salute the flag, how can they require a large mining company to speak to something as “controversial” as whether its operations might be supporting violence in the DRC? That is the new First Amendment at work.

In fact, every labeling and advertising law for every commodity is now in the crosshairs. Calorie disclosures: do they overemphasize calories, undermining a company’s right to convey a happier image about its products? Does a required warning about cell phone radiation disrupt corporate messaging and so run afoul of the Constitution? What about sugar warnings? Companies have pushed all of these arguments in the courts—so far with limited success. But attuned to the power of the small businessman in the American imagination and looking ahead to the growing wave of Federalist Society judges appointed by Trump, right-wing litigation shops are building an army of more sympathetic complainants. They chose, for example, the South Mountain Creamery for their assault on the FDA’s authority over product labels. The farmers asked a seemingly innocuous question: How can the FDA mandate that the business call its skim milk “imitation,” when it has, in fact, skimmed the fat off the top of whole milk? The FDA has long required that food products be labeled “imitation” when essential nutrients are removed in processes such as skimming but are not restored. What is masquerading as a narrow challenge to labeling requirements for the benefit of small dairy farmers is in reality a wholesale offensive against the FDA’s ability to protect shoppers from foods misleadingly lacking the nutritional value people have come to associate with them. (The case is still pending.)

Ordinary economic transactions also involve speech, so they too are at risk. This seems fantastic but consider the case of Nordyke v. Santa Clara County (1996). In it, a federal district court ruled that Santa Clara’s attempt to forbid gun sales at its fairgrounds was an infringement on speech. As the district court said, “a gun may not be sold in silence, without any exchange of verbal communication whatsoever.” Once this was defined as an issue of speech, the court was empowered to ask about its justification: Why ban sales at the fair and not at the gun shop down the street? Finding the lawmaker’s proof lacking, the court threw the law out. When appealed, the case was upheld on a similar logic: the county was regulating speech and had not provided sufficient evidence for its choice.

Some of the laws being targeted may be unnecessary, even idiotic. But whether credit card surcharges or skim milk labeling rules are stupid has never in the modern era been a constitutional question. Rather, the question was, who decides? And the answer was, emphatically, the people through the officials they elect. Recall that constitutional obligations cannot be revised except by constitutional amendment—or by a reshaped Supreme Court.

The stakes for our democracy are deep. But so too are the stakes for our safety. Among the most pernicious cases are those targeting the FDA’s powers over the tobacco and pharmaceutical industry, threatening return to a world in which we know little about the drugs we put in our bodies and snake-oil salesmen reign supreme. In its ruling in favor of Caronia, the Second Circuit made much of the fact that doctors are allowed to prescribe off label, though companies may not promote off label—here again was the dreaded “discrimination.” In context, though, the differentiation makes sense. The regulatory scheme is about the production of evidence, and it is drug companies, and not individual doctors, that are in a position to fund and conduct the studies needed to guide practice. Our pharmaceutical industry was built—thrived, in fact—under a strong regulatory thumb: companies had to provide high-quality evidence or stay out of the market. But if it costs less to run catchy ad campaigns than to run studies, drug companies will choose that path every time.

The Caronia case still permits the FDA to bar false and misleading marketing. But it ensures the people who judge that fact will have neither the evidence nor the expertise to make good on that task. In a follow-up case to Caronia, a federal district court judge overruled the FDA, permitting a company to market a fish oil pill to a broad, low-risk population as a heart disease remedy. The FDA argued that the pill did nothing for that group. But without the benefit of fact-finding or expert witnesses, the judge disagreed. This was the same judge who, during the hearing, said to the lawyers: “You’re talking to somebody who has difficulty using a toaster. I’m the last person who should opine on this.”

Lining up after the pharmaceutical industry are the e-cigarette makers. They are targeting a federal law that requires them to provide evidence before they claim their products are lower risk than conventional cigarettes. The law was passed after Congress found that tobacco companies had misled the public for years, marketing “low tar” and “light” cigarettes that offer no health benefits. But a wave of current lawsuits argues that the Constitution forbids Congress and the FDA from learning from this dark history and from demanding evidence before a new generation is addicted.

The stakes for our democracy are deep. But so too are the stakes for our safety.

Consider, finally, the implications of the new First Amendment for a world where so much is digital, mediated by software and measured in terabytes. What is a file that instructs a 3D printer how to make a gun? Speech, of course. That is the argument made by Defense Distributed, a nonprofit that makes executable CAD files for plastic pistols and AR-15 components that can be printed from the comfort of your home. The State Department, citing a law designed to prevent the export of certain weapons, asked the company to remove the files from the Internet. The company sued, pointing to its First Amendment rights to speech. A lower court deferred the question, and shortly thereafter, the Trump administration settled the suit, voluntarily agreeing to let the files be disseminated around the world.

Are we a people with the authority to hold market actors accountable? Or are we in a new era, our sovereignty transferred to their sovereignty? Will “Congress shall make no law” come to mean no effective law against commercial fraud and no effective regulation against the addictive and toxic products that companies slide into the stream of commerce? The minorities that courts once understood as underrepresented persons are now marginalized corporations, and the power that they are being protected from is our own. As Justice Elena Kagan put it in a searing dissent in 2018:

Speech is everywhere—a part of every human activity (employment, health care, securities trading, you name it). For that reason, almost all economic and regulatory policy affects or touches speech. So, the majority’s road runs long. And at every stop are black-robed rulers overriding citizens’ choices. The First Amendment was meant for better things.

And so were we.