Stock

Robert A. Levy

This blog post updates and substantially enlarges my previous post on December 20, 2024.

1. When are executive orders valid and when are they invalid?

The ground rules are: First, the president can issue executive orders pursuant to a grant of authority from Congress. Second, executive orders are legitimate if they relate to the president’s role as commander-in-chief, or to his other core powers—especially in the areas of foreign policy and national security. Third, executive orders are valid when they deal with the internal workings of executive agencies. Beyond those three categories, executive orders affecting the rights and obligations of private parties may well be unconstitutional. They are not intended as a way for the executive branch to bypass Congress. 

Justice Robert Jackson set out the constitutional framework in his concurring opinion in Youngstown Sheet & Tube v. Sawyer, the 1952 case that denied President Truman’s authority to seize the steel mills during the Korean War. Jackson offered this analysis: When the president acts pursuant to an express or implied authorization from Congress, “his authority is at its maximum.” When the president acts in the absence of either a congressional grant or denial of authority, “there is a zone of twilight in which he and Congress may have concurrent authority.” But when the president takes measures incompatible with the express or implied will of Congress, “his power is at its lowest.”

Ignoring that framework, Trump and his recent predecessors have unilaterally rewritten congressional enactments. Imagine if a future president were to annul environmental laws or financial regulations. Liberals would be up in arms. Or suppose a president alone implemented new regulations that Congress had considered and rejected. You’d expect conservatives to be on the warpath. Yet some observers believe that the Trump, Biden, and Obama presidencies overreached in both directions—ignoring existing laws and fashioning new ones.

2. What can Congress do to constrain presidential abuse of executive orders?

Here are two options if a president’s executive orders go too far: First, Congress could refuse to fund the activity. Second, Congress could pass legislation with guardrails limiting executive and administrative action. For example, the Congressional Review Act allows Congress to nullify selected agency acts within a specified time frame; and the REINS bill (Regulations from the Executive in Need of Scrutiny), if passed, would require Congress to approve all major administrative regulations. 

Let’s examine a few of President Trump’s more controversial augmentations of executive power.

3. Can the president impose tariffs without congressional approval?

Recent laws give the president substantial authority over tariffs that protect industries harmed by global trade. But blanket, across-the-board tariffs—or those unrelated to trade policy, such as assessments meant to compensate for the ravages of fentanyl—might not hold up. Moreover, our trade agreement with Mexico and Canada (negotiated by Trump in 2020) provides for duty-free transactions among the three countries. So, any new tariffs imposed on our immediate neighbors will be subject to challenge if they’re more than a bargaining ploy.

4. Can the president unilaterally abrogate a treaty? 

The Constitution tells us that presidents can “make Treaties, provided two thirds of the Senators present concur”; but it says nothing about termination or abrogation of treaties. Nor has the Supreme Court filled that gap. In 1979, President Carter withdrew from the Taiwan Mutual Defense Pact. The withdrawal was challenged in Goldwater v. Carter. But the court ruled that the dispute was a political question, not to be resolved by the judiciary. Perhaps if Congress had passed a resolution opposing termination of the treaty, the court would have rejected Carter’s unilateral action. In practice, however, presidents have repeatedly withdrawn from treaties without being successfully challenged.

Conceivably, a treaty could be implemented by statute rather than by Senate consent, in which case abrogation would require the repeal of the underlying law—unless, of course, the statute gave the president withdrawal discretion. At the other extreme, some treaty-like arrangements are structured as executive agreements, without Senate approval. Those agreements may be politically binding, but not legally binding. They could be canceled by the president alone.

5. Can the president bar transgenders from competing in female sports?

The Trump proposal to ban transgenders from female competition appears to enjoy substantial popular support. Department of Education regulations would apply to private schools, elementary through college, that receive federal funding. That raises multiple legal issues: (1) Is the federal government constitutionally authorized to regulate education? Remember, Trump proposed closing the DOE, in part because the 10th Amendment says that powers not enumerated in the Constitution are reserved to the states. (2) Can the president deny funding to private schools that don’t comply with his executive order? Probably not, if the denial is deemed to be coercive. (3) Does Title IX of the Civil Rights Act, which bars discrimination by sex—but not necessarily by sexual identification—allow differential treatment of transgender females? And finally (4) Does the president have the power by executive order, without congressional input, to direct DOE to implement the applicable regulations?

6. Can the president impound—i.e., refuse to spend—congressionally appropriated funds?

Trump has suggested that he would resurrect impoundment to control federal spending. But the Impoundment Control Act (1974) provides that rescission must be approved by both chambers within 45 days. Moreover, Congress is not even required to vote on rescission. In the past, legislators have ignored most presidential requests. Still, all recent presidents from Reagan forward have supported impoundment, as did many politicians across the ideological spectrum (e.g., John McCain, John Kerry, Al Gore, and Paul Ryan), as well as Elon Musk and Trump’s OMB pick, Russ Voight.

The Supreme Court hasn’t ruled on the constitutionality of the Impoundment Control Act, but in Train v. New York (1975), the court said that President Nixon could not refuse to fund an environmental project after Congress overrode his veto. And in 1998, the court held that the line-item veto was unconstitutional. It would have allowed the president to veto parts of a bill.

Most likely, the impoundment issue will be re-litigated. The court will have to consider, first, the Constitution’s Appropriations Clause, which sets a ceiling (but not a floor) on federal spending. It says, “No Money shall be drawn from the Treasury, but in Consequence of Appropriations made by Law.” That provision will be informed by separation of powers concerns and the president’s duty to faithfully execute the laws.

Arguably, the Court might view Trump’s proposed freeze or delay in spending differently than it would view impoundment, which is a refusal to spend. Unless a statute has a specific date by which expenditures must occur, the president might be justified in postponing the spending until he can certify that it complies with other laws and executive orders.

7. Is the Department of Government Efficiency (DOGE) legal? 

At the outset, there’s the Elon Musk question. His access to confidential information may violate various privacy and national security regulations. Further, he’s been designated as a “special government employee,” which limits him to 135 workdays in any year. He’s also subject to conflict-of-interest laws that bar his involvement with government matters that could affect his personal interests (e.g., procurement and research contracts). And if he’s deemed (as the head of DOGE) to be an “Officer of the United States,” he would have to be confirmed by the Senate. 

Then there’s DOGE itself—supposedly an advisory entity and therefore subject to the transparency rules of the Federal Advisory Committee Act. That Act also limits the delegation of authority to private parties. Additionally, DOGE may not be authorized to spend (much less impound) money without the approval of Congress.

8. Can the president close an agency?

President Trump plans to close USAID (the Agency for International Development) as well as the Department of Education. Can he do so? Generally, the president can close an agency under these conditions: (1) the agency was authorized by executive order, not by statute; or (2) the statute establishing the agency gives the president termination discretion. But if Congress has already appropriated funds to run the agency, then the president’s flexibility may be limited—unless Congress has also given the president the option not to spend the funds, or the president can legally impound the funds (a subject discussed above).

9. Can the president remove agency heads before their terms expire?

The Constitution covers the appointment of agency heads, but not removal. Until 1935, the president could fire subordinates freely. But then, in Humphreys Executor v. United States, the Supreme Court concluded that the Federal Trade Commission did not exercise executive power. Accordingly, if the president, as head of the executive branch, wanted to remove an FTC commissioner, he could do so only “for cause.” Again in 1988, the court upheld a congressionally-imposed, for-cause requirement. That case, Morrison v. Olson, was about terminating an independent counsel. The court found that limiting the president’s removal authority would not impede the operations of the executive branch. The president would still be able to faithfully execute the laws.

However, in a 2020 case, Seila Law v. Consumer Financial Protection Bureau, Chief Justice John Roberts held that a for-cause constraint on the president would violate the separation-of-powers doctrine because the director of the CFPB exercised substantial executive power. The ground rules, established in Seila, preclude a for-cause requirement except when (1) the president is firing an inferior officer, such as an independent counsel, who has limited duties and no policymaking or administrative authority, or (2) the agency is headed by a commission balanced along partisan lines, appointed to staggered terms, and performing mainly non-executive functions. (That guidepost would not, however, prevent the president from removing the chair of an agency from his leadership role, without removing him from the agency itself.)

10. Can the president exert control over so-called independent agencies?

A Trump executive order, published on February 18, imposes executive branch supervision over independent agencies—such as the SEC, FCC, FTC, NLRB, and CFPB—that are not technically characterized as “executive” even though they have law-enforcement authority. Trump asserts that he would not be able to “take Care that the Laws be faithfully executed,” as mandated by Article II of the Constitution, without supervising those agencies. 

Under Trump’s order, independent agencies would have to submit their rulemaking proposals to OIRA (the Office of Information and Regulatory Affairs), which is part of the Office of Management and Budget. The rules would then be reviewed within the White House for conformity to the president’s agenda and cost-benefit constraints.

One of Trump’s stated goals is to increase accountability by substituting an elected official (the president) for unelected agency heads. That would represent a major change in governance—effectively reinstating the Framers’ original design that established three branches, which did not include independent agencies. The effect, if approved by the Supreme Court, would overrule the 1935 Humphrey’s Executor decision. And because many of the independent agencies also exercise legislative power, the longer-term impact might be to incentivize Congress to resume the lawmaking role that it has abdicated in favor of the administrative state.

At the root of Trump’s claim for expanded authority is the so-called unitary executive theory. It says the president wields exclusive control over the executive branch. A stronger version of the theory states that laws conferring independence on executive-type agencies or officials are unconstitutional. A moderated view, advocated by constitutional scholar Yuval Levin, is that “the president does command the executive branch, but the executive branch does not command our government.” Levin has nevertheless embraced the restoration of presidential authority over so-called independent agencies, which seems consistent with our tripartite constitutional plan.

11. Can the president fire government employees and bypass Civil Service requirements?

In January, Trump fired lower-level employees—FBI agents, Justice Department attorneys, and inspectors general—who are part of the executive branch, but may be protected by contract and Civil Service regulations. He also offered to buy out some of their contracts. Presumably, the buyouts would be funded by eliminating compensation that would otherwise be paid to the resigning employees. If, however, incremental funding were needed, that would require a congressional appropriation. Additionally, federal law calls for written reasons and 30-days notice to fire inspectors general.

On a related front, labor unions are challenging Trump’s Schedule F reform, which would remove Civil Service protection for selected high-ranking employees. But the president’s legal argument seems well-grounded: The Civil Service Reform Act of 1978 exempts positions “determined to be of a confidential, policy-determining, policy-making or policy-advocating character.”

12. Can the president make recess appointments without Senate approval?

Normally, high-level presidential appointments require Senate approval. But the Constitution allows short-term appointments for vacancies that arise when the Senate isn’t in session. Those appointments must, however, be confirmed by the Senate by the end of the succeeding session of Congress —i.e., within a maximum of two years—or else the appointments expire.

The Constitution also requires both Houses to approve any recess longer than three days. But if the two Houses disagree as to the “time of adjournment,” the Constitution allows the president to “adjourn [Congress] to such time as he shall think proper.” That’s what Trump said he would do if the Senate prolonged his confirmations. But, to block the president from making recess appointments, the Senate has occasionally held pro forma sessions, which are called to order, but no business is conducted. In NLRB v. Canning (2014), a unanimous Supreme Court held that those sessions are not recesses, because Congress retains the capacity to conduct business. Instead, a ten-day minimum break is necessary to declare a recess. In short, said the court, Congress itself determines when Congress is in session. 

Of course, it’s perfectly normal for presidents to begin their terms on January 20 with most cabinet nominees not yet confirmed. The proper fix is for the president to make temporary appointments, not try to bypass the confirmation process by declaring a Senate recess. That’s why Congress enacted the Federal Vacancies Reform Act in 1998. It provides for interim appointees, who can serve up to 300 days. Persons serving as interim officials cannot be the ultimate nominee, but they can be another already-confirmed person, or a senior employee meeting certain criteria.

13. Can President Trump bar federal grantees and contractors from promoting DEI?

Trump has ordered that federal grants and contracts be withheld or canceled for entities that have implemented diversity, equity, and inclusion programs, on the ground that they violate antidiscrimination laws. Opponents of DEI contend that “diversity” is a code word for a representative mix of various identity groups—usually based on characteristics such as race, gender, or sexual orientation; “equity” seeks to impose equal outcomes (not just opportunities) for group members and non-members, independent of merit; and “inclusion” urges acceptance of group members, without regard to character, effort, or competence. 

DEI programs, according to Trump, run afoul of the 14th Amendment’s Equal Protection Clause and Title VI of the Civil Rights Act, which applies to recipients of federal funds. Ammunition for that argument comes from the recent Supreme Court affirmative action cases involving Harvard University and the University of North Carolina. The court held that race could not be used as an admissions criterion to achieve a diverse student body.

The counterargument, which was endorsed on February 20 by a federal judge in Maryland, is that Trump’s executive order “is textbook viewpoint-based discrimination”—i.e., suppression of speech based on the position taken by the speaker, in violation of the First Amendment. The judge temporarily halted enforcement of Trump’s executive order. He concluded not only that it threatened the expression of views supportive of DEI, but also that its lack of clarity violated Fifth Amendment rights to due process.

14. What are the limitations, if any, on the president’s pardon power?

Here are the guidelines: (1) The president can pardon federal criminal acts, not state crimes or civil liability. (2) A pardon cannot protect against future crimes—only past actions. (3) The pardon power extends to all “Offenses against the United States,” which doesn’t distinguish between those crimes already prosecuted and those not yet prosecuted. There are at least two open questions: First, can the president pardon himself? Probably not, but it’s unsettled. Second, with respect to preemptive pardons—for crimes that may have been committed but have not been charged—must the pardon clearly specify the conduct and time frame? Hunter Biden’s pardon, for example, specified the time period, but contained only a vague description of the crimes. Ditto for Nixon’s pardon, which applied to “all offenses against the United States which he … may have committed or taken part in” during a specified period. A Supreme Court pronouncement on the requisite degree of specificity would certainly be helpful.

15. What impact can President Trump have on the TikTok controversy?

In TikTok v. Garland (2025), the question for the Supreme Court was whether the Protecting Americans from Foreign Adversary Controlled Applications Act violated the First Amendment. The Act required TikTok’s Chinese parent, Byte Dance, to sell TikTok by January 19. Otherwise, TikTok’s U.S. operations would be banned. Two days before the deadline, the court upheld the law, 9–0, ruling that restrictions on TikTok were based on the ownership and privacy of the information, not its content. Accordingly, the justices concluded, the First Amendment issues were less compelling, and the law advanced an important government interest without excessively burdening speech—notwithstanding the right of 170 million US customers to post and have access to the information.

Prior to his inauguration, Trump had asked the Supremes to postpone the January 19 sale-or-ban deadline. He filed an amicus brief suggesting that he would be able to negotiate a settlement. The court declined Trump’s request; but then, post-inauguration, he ordered the US ban delayed for 75 days. That did not comply with the PAFACA statute, which allowed a delay only if sale negotiations were well underway. They were not. And the statute also required that any delay had to be ordered before the January 19 deadline. Plainly, Trump’s executive order flouted the law. 

Yet, it appears as though the delay will not be challenged. Plaintiffs generally would not have legal standing to insist that the Justice Department enforce the TikTok ban. Discretionary enforcement choices are seldom subject to judicial review unless the statute provides a clear right of action for private parties. Nothing in the PAFACA Act creates such a right.

16. Can President Trump eliminate birthright citizenship for the children of illegal aliens?

The 14th Amendment states that “All persons born or naturalized in the United States and subject to the jurisdiction thereof, are citizens.” The goal was to overturn the 1857 Dred Scott decision, which denied citizenship to slaves and their children. The Framers added the phrase “subject to the jurisdiction” to withhold citizenship from children of foreign diplomats and children of enemy forces occupying the United States. Neither group was deemed to be subject to our jurisdiction. But Congress surely intended that children of permanent legal residents would be citizens. Indeed, the Supreme Court said so in United States v. Wong Kim Ark (1898). That decision did not, however, resolve the status of temporary resident aliens, e.g., students and tourists, or illegal aliens.

Those persons are subject to our laws—at least while they’re here—and they can be convicted of violations. So, a natural reading of the phrase “subject to the jurisdiction” suggests that their children are citizens. But in 1868, when the 14th Amendment was ratified, there were no laws restricting immigration. So, birthright citizenship wasn’t about illegal aliens. More likely, “subject to the jurisdiction” meant primary allegiance to the United States as sovereign and principal lawgiver. On the other hand, some victims of the slave trade were brought here illegally. Their allegiance was surely questionable; yet their children were among the designated beneficiaries of birthright citizenship. Moreover, in Plyler v. Doe (1982)—involving illegals’ access to education—the Supreme Court said there was “no plausible distinction with respect to 14th Amendment ‘jurisdiction’ … between resident aliens whose entry into the United States was lawful, and resident aliens whose entry was unlawful.”

My sense of the constitutional question goes like this: If Congress were to deny citizenship to children of illegal and non-permanent aliens, a conservative Supreme Court might (but probably wouldn’t) uphold that law. It’s even less likely that the court would uphold a unilateral declaration by the president without parallel legislation.

17. Finally, what is the recourse if the president ignores a court order?

Although executive branch subordinates can be held in contempt, fined, or even disbarred, Justice Department policy bars prosecution of a sitting president. In the past, courts appointed prosecutors who were independent of the executive branch, e.g., Ken Starr, who investigated President Clinton’s involvement in the Whitewater case. That process was affirmed by the Supreme Court in Morrison v. Olson, but the underlying statute has since expired; it would have to be re-enacted. Other than impeachment, the remedy when a president ignores a court order is likely to be political rather than prosecutorial.

One concluding comment: The imperial presidency has indeed become a problem. But the problem pre-dates recent power grabs by Donald Trump. Nor does culpability rest exclusively with the Republican Party. After all, it was Joe Biden who attempted to bypass Congress and cancel roughly $400 billion in student loans. And it was Joe Biden who then tried to circumvent a Supreme Court decision by conjuring alternative means to the same end. And it was Joe Biden’s Department of Labor that unsuccessfully claimed unilateral authority to mandate vaccinations by all large private sector companies. In each case, Biden had no legislative assent. Nonetheless, he garnered enthusiastic backing from congressional Democrats—willingly condoning executive overreach when their man was in the White House. Of course, Republicans behaved in like manner. 

So, the threshold problem is neither Biden nor Trump. It’s congressional abdication—the solution for which is Congress’s re-assertion of its authority as a co-equal branch, and reinvigoration of its legislative and oversight roles, as envisioned by the Framers.

Stephen Slivinski

Ten years ago, on February 25, the Supreme Court of the United States reminded us that maybe the Sherman Antitrust Act, the federal statute passed in 1890 intended to enforce and increase market competition, should be aimed at the only lasting anti-competitive force in the modern market economy: the government itself. 

It all started nearly 20 years ago when the North Carolina Board of Dental Examiners declared war on shopping mall teeth whiteners, none of whom were dentists. The board sent more than forty-five cease-and-desist letters to entrepreneurs offering conventional teeth-whitening services—not in spacious and well-appointed offices but more like modest spaces for people to apply the products themselves with some help and maybe an infrared lamp provided by the staff. The dental board also convinced the state’s Board of Cosmetic Arts Examiners to issue similar letters. In 2007, they also began sending letters to shopping mall managers recommending eviction of the offending businesses.

This all happened because the board members wanted to bully their competitors out of business because the dental board consisted mainly of licensed dentists—six out of the eight board members—who saw these upstarts as a threat to their existing businesses. So, using the power the state government delegated to them to enforce occupational licensing laws, the board defined the practice of dentistry to also include teeth whitening. This meant that the shopping mall teeth whiteners, to continue to operate their business, needed to spend thousands of dollars and years of classwork to get an accredited dental degree and then jump through the additional hoops to get state dental licensing. 

It was obviously meant to be a death sentence for these competitors, meted out by the judge, jury, and executioner that was the North Carolina Board of Dental Examiners.

Then, on February 25, 2015, the Supreme Court of the United States reprimanded the North Carolina board for this shakedown. The SCOTUS majority effectively stated that state licensing boards made up of a majority of “active market participants” and not subject to active legislative supervision were effectively acting anti-competitively—like a private cartel—by using a regulatory power that the government handed to them to protect their incumbent market interests. 

You’d think that over the past ten years most state legislatures and boards would have wised up and at least tried to look less suspicious even as they continued to box out their economic rivals. But you’d be wrong.

A recent report, co-authored by me and two attorneys from the Pacific Legal Foundation, took a look at how many “active market participants” sit on dental boards today versus ten years ago. Most of them look the same as they did then, and a few even look worse. 

Since the NC Dental case was decided, no state has changed its laws to decrease the industry control of dental boards. Three states—Georgia, Michigan, and North Dakota—have actually increased the number of board seats held by people with clear conflicts of interest. And eight states allow lobbyist control of dental licensing board nominations by requiring that the governor choose board members only from a list provided by a state or national dental association.

This isn’t only a problem on dental licensing boards. Of the more than 1,700 occupational licensing boards across the country, 85 percent are required by statute to award a majority (and sometimes a supermajority) of voting seats on the board to those already licensed and currently active in the profession.

What should this mean to liberty-minded legislators and litigators? You can tune in to this February 27 Cato webinar to hear our thoughts. And you can read the report here.

Jeffrey A. Singer

When Mark Ibsen, MD, spoke as a panelist in the Cato online event “Pain Refugees: Collateral Damage in the War on Drugs” last December, he had no idea that representatives of the Montana Board of Medical Examiners (BOME) were watching. Three days later, he was shocked to receive a letter from BOME demanding he justify mentioning deceased patients by name (possibly violating the Health Insurance Portability and Accountability Act, or HIPAA)—a charge that would later be dropped but sent a clear message: speaking out has consequences. The letter stated:

While investigating the original Complaint, the Department became aware that you identified six former patients by name in the following online forums, which were available to the public:

1) December 22, 2019, written post responding to a blog article entitled “No Help, no medication. I want out. I’m not strong enough” on the Pharmacist Steve website.

2) October 6, 2020, video post on your Facebook page.

3) December 2, 2024, video post on your Facebook page.

4) December 3, 2024, panel discussion hosted on the Cato Institute’s website and video post of the panel discussion on that website.

… Please provide a written response to these additional allegations by the end of the day on December 16, 2024. Please include in your response whether you had written permission from your former patients or other legally authorized individual to disclose the patient’s protected health information in a public setting. If so, please provide copies with your response. (emphasis added)

In 2015, agents from the Drug Enforcement Administration (DEA) raided the office of Christopher Christensen, a Montana physician who was then 67 years old, and arrested him for inappropriately prescribing opioids to his patients. The DEA charged Dr. Christensen with 400 felony drug offenses and two counts of negligent homicide following the overdoses of two patients who had taken his prescribed medication. A jury convicted Christensen in 2017, and the court sentenced him to 20 years in prison, 10 of which were suspended. The Montana Supreme Court overturned the negligent homicide convictions but still required him to serve time in prison for nine counts of criminal endangerment and 11 counts of criminal distribution of dangerous drugs.

Dr. Christensen’s arrest by cops practicing medicine frightened Dr. Ibsen, a Montana primary care and urgent care physician. Dr. Ibsen wrote me: “Before ceasing writing prescriptions for opiates, reported DEA incursions had me thinking that every opiate prescription I wrote could be my last. I worried if I could trust patients to honor their promise that they would not kill themselves with my prescription, as that would take me down as well.” He ceased prescribing opioids for pain in 2017. “I thought I was going to be given an award for weaning 80% of my patients off opioids.”

He tapered his patients’ doses, gave them notice that he would no longer treat them, referred them to other doctors, and prescribed them a 30-day supply of pain medications to tide them over until they saw the other doctors. Unfortunately, other doctors were oftentimes afraid of taking on new pain patients. Some of Ibsen’s former patients grew desperate, and like many across the country in similar straits, some turned to suicide.

At Cato’s pain refugee event, Dr. Ibsen, holding back tears, stated:

Six of my patients died from 2016 to 2018 after I stopped prescribing opiates when Dr. Chris Christensen was convicted. Three of my patients died by gunshot wounds … two by alcohol complications … and one by a non-opioid overdose: Chris Storseht — Jennifer Adams – Robert Mason – Jennifer Beausoleil – John Burke- Lynette Chadwick.

At the BOME Screening Panel hearing this month, the BOME attorney never mentioned HIPAA violations when summarizing the Board’s concerns. The physician chairing the panel didn’t mention HIPAA violations either.

One of Dr. Ibsen’s chronic pain patients from 2013 provided an affidavit stating:

Information, pictures, and names of chronic pain patients who committed suicide, have been public knowledge for a number of years. … There was a conference put on by the Helena Department of Public Health & Human Services in 2018. A few of us chronic pain patients that belong to an organization called “Don’t Punish Pain” were invited to be on a panel. We talked about, (six) chronic pain patients just in Montana! We held up their pictures as we said each of their names.

But the affidavit wasn’t needed. The panel voted to terminate the investigation.

This might end the Board’s persecution of Dr. Ibsen, but he will always be haunted by the memories of pain patients who suffered so much that they took their own lives.

Ibsen’s story is emblematic of a broader pattern: regulators and law enforcement drive physicians out of pain management, leaving patients nowhere to turn. His decision to stop prescribing opioids after Dr. Christensen’s conviction—out of fear that he could be next—shows the chilling effect of aggressive prosecution. And the tragic outcome for his former patients mirrors what has happened nationwide as pain management has been criminalized.

Check out my new book, Your Body, Your Health Care, scheduled for release on April 8.

Clark Neily

Jennifer Huddleston

On February 20, the Federal Trade Commission (FTC) released a request for information (RFI) on “Technology Platform Censorship.” The last section of the RFI asks, “Were platforms’ adverse actions made possible by a lack of competition? Did the practices and policies affect competition?” Such questions indicate that, once again, the connection between antitrust and concerns about content moderation may be under consideration. Yet this misunderstands the appropriate use of antitrust enforcement, the likely outcomes for content moderation under such enforcement, and the ways the current liability protection of Section 230 actually encourages competition in content moderation.

Antitrust Enforcement is Not an Appropriate Tool for Concerns About Content Moderation

Antitrust enforcement and competition policy are powerful tools that can result in significant government intrusion into a market. The current US standards of enforcement support such intervention only in specific per se cases or when there is sufficient evidence of harm to consumer welfare. Neither of these scenarios is true when it comes to content moderation policy, nor would they support intervention into an otherwise competitive market for concerns about content moderation.

Elon Musk’s X has a different content moderation strategy than Facebook or Reddit. Despite being general-purpose social media platforms, they have found distinct roles in the markets for both consumers and advertisers. This is not to mention the plethora of other user-generated, content-based sites that might compete for more specific audiences, whether it be BlueSky, President-Trump-owned Truth Social, or more targeted sites like Goodreads or Tripadvisor. Users may choose a social media site in part for its content moderation policies, as seen when changes lead users to seek substitutes that are readily available. If anything, the ongoing evolution of content moderation policies by platforms illustrates the continuing competition in this regard.

Using antitrust policy beyond its intended and objective purposes risks harming consumers rather than helping them. In the last administration, Chair Lina Khan’s FTC sought to deviate from these objective standards in favor of policy goals or their belief in what the market should look like. This resulted in losses on cases brought by the FTC that relied on creative claims or unrealistic market definitions, such as in future markets in virtual reality fitness. A theory that points to disliked content moderation decisions as a violation of antitrust laws would likely similarly fail in the courts.

Such deviations from traditional economic-based analysis also risk allowing the use of antitrust in the future for policy objectives it is ill-designed to achieve. That could allow significant government intervention into competitive markets. The result is a shift away from a focus on consumers to a focus that emphasizes competitors or the government’s preferred policies at the expense of the benefits they would have otherwise had from the market.

Antitrust Remedies Would Not Resolve Content Moderation Concerns

Say a content moderation-based antitrust case was successful in court. Would the remedies actually improve content moderation the way those concerned about censorship believe? In short, it is unlikely, and it could even make social media experiences worse.

As Adam Thierer and I first wrote about in 2019, there is no reason to believe that structural remedies would resolve other tech policy concerns like “censorship.”

First, there is no reason to believe that these two separate platforms still trying to retain their audience would change their content moderation policies just by their separation. Just separating a company would not change what a platform feels best serves either its audience or advertisers in the type of content it seeks to curate.

But things could actually get worse rather than better. These now-separate platforms might have fewer resources to respond to more obvious problematic content like animal cruelty or spam, thus diminishing the consumer experience. They would also be less able to invest in or deploy state-of-the-art tools to respond to novel problems that might arise, like the viral “Tide Pods” challenge or AI-generated sexual content. 

Additionally, rather than resolving content moderation concerns, a forced separation might result in more intense content moderation based not on what audiences want but on what advertisers want be able to afford to keep the platform afloat.

Section 230 and Current Content Moderation Policy Encourages Competition

In conversation around social media “censorship,” Section 230 is often the elephant in the room. This 1996 law has two parts that create liability protection for platforms hosting user-generated content from litigation related to their users’ content and their content moderation decisions in most cases. Critics of the law will often say it is a handout to big tech, but Section 230’s liability protection actually encourages competition in the market of social media and other user-generated content.

While Section 230 is important to large platforms engaging in content moderation at scale, it also enables small platforms to host user-generated content without the fear that a bad actor or a wrong decision could result in business-ending litigation costs. Even if proven justified in court, the cost of defending a case can easily reach more than six figures and derail a potential new entrant into the market. Such protection is particularly critical for those whose platforms might allow the discussion of controversial ideas or serve marginalized communities.

While content moderation should not be thought of as an antitrust or competition issue, it is important to recognize how the legal frameworks that support private actors’ content moderation decisions can support competition. Changes that risk the possibility of litigation around content moderation decisions risk locking in the decisions of the largest players who can afford to engage in litigation and could result in many platforms further limiting discussion or voices for fear of potential lawsuits. Once again, the consumers would be the ones ultimately harmed by such changes.

Conclusion

There is much further discussion to be had on claims of social media “censorship”; however, it is important to separate any such claims from conversations around competition policy or antitrust in the technology sector. The use of antitrust enforcement for such a policy approach risks removing the important, consumer-focused objective standard in ways that could allow further intervention into a wide array of markets and violate the First Amendment but is also unlikely to provide a remedy for the alleged concerns.

Despite positioning itself in the request as seeking to protect consumers against “censorship,” the reality is that government action that dictates the decisions private actors can make regarding the content they host or amplify on their platforms raises far more significant concerns about government censorship and the First Amendment than the actions of the platforms themselves. The current policy approach to content moderation under Section 230 has allowed competition for platforms hosting user-generated content to flourish. Changes could limit rather than encourage that competition.

Clark Neily

In his new book examining the “explosion” of freedom-restricting laws over the past few decades, Over Ruled: The Human Toll of Too Much Law, Justice Neil Gorsuch advances a thesis that will resonate with many: America is suffering from an excess of legal rules and regulations that increasingly constrain individual liberty while defying common sense. 

Through vivid examples and careful analysis, Gorsuch documents how the sheer volume, complexity, and severity of modern American law have created a system where, paradoxically, there may be too much law for anyone to follow but too little justice for those caught in its web. 

Consider the federal government’s bizarre regulatory crusade against the Ernest Hemingway Home and Museum in Key West, Florida. The Department of Agriculture deemed the museum’s resident cats—descendants of Hemingway’s beloved felines—to be subject to federal animal exhibition standards, leading to prolonged litigation over proper cat-care protocols that are not obviously within the purview of a federal government whose powers were meant to be “few and defined.” 

Or take the 1997 prosecution of racing legend Bobby Unser, who faced federal charges for accidentally snowmobiling onto protected land while desperately seeking shelter from a life-threatening blizzard. That Unser was ultimately convicted and fined (albeit only $75) after a trial without a jury speaks volumes about how technical violations can trump common sense even in extreme circumstances. 

Adding insult to constitutional injury, Unser was denied even the basic right to a jury trial due to the judicially invented “petty offense doctrine,” which provides that crimes punishable by less than six months’ imprisonment do not trigger the jury trial right—this despite the explicit command in Article III of the Constitution that “the trial of all crimes shall be by jury” and the Sixth Amendment’s guarantee of the right to a jury trial “in all criminal prosecutions.”

Notably, any suggestion that federal prosecutors have grown more judicious in their case-selection practices is belied by the ongoing prosecution of charter-boat operators John Moore and Tanner Mansell for interfering with a federally authorized, long-line shark-fishing operation near Jupiter, Florida, which they plausibly—though mistakenly—believed to be the work of poachers. Moore’s and Mansell’s felony theft convictions were reluctantly affirmed by the Eleventh Circuit, and they will be seeking Supreme Court review with amicus support from Cato and others. 

These and numerous other examples illustrate Gorsuch’s broader argument about the corrosive effects of statutory and regulatory proliferation. When experts estimate that at least 70 percent of American adults have unknowingly committed imprisonable offenses, we confront the troubling reality described by law professor William Stuntz who, as quoted by Gorsuch, observed that “too much law amounts to no law at all,” because “when the law makes everyone an offender, the relevant offenses have no meaning independent of law enforcers’ will. The formal rule of law yields to the functional rule of official discretion.” A more succinct—and ominous—summary of the current regulatory landscape is hard to imagine.

Gorsuch’s diagnosis of these symptoms is compelling, and his concern for individual liberty—as expressed in his book and his judicial rulings—is inspiring. But the analysis presented in Over Ruled contains a rather glaring omission that will have many libertarian constitutional litigators (and perhaps others as well) scratching their heads and/​or grinding their teeth, and that is the role of the federal judiciary in facilitating the very proliferation of law Gorsuch decries.

The Constitution’s framers anticipated that members of the legislative and executive branches would seek to expand their powers beyond constitutional limits. Accordingly, they created an independent judiciary to check this tendency and protect individual rights. Yet modern courts have largely abdicated that responsibility through a series of doctrinal developments that systematically favor government power over individual liberty and tilt the litigation playing field sharply in favor of power-wielding government actors. These include:

Judicially invented immunity doctrines and excessive use of “avoidance doctrines” such as standing, ripeness, mootness, and abstention to prevent otherwise viable challenges to government power from being heard on the merits;
Application of rubber-stamp “rational basis review” as the default standard in constitutional cases;
Embrace of functionally limitless theories of enumerated federal power, such as Wickard v. Filburn’s “aggregation doctrine”; and
Acquiescence to the practical elimination of criminal jury trials through mass plea bargaining, which vastly reduces the cost of pursuing even marginal charges and enables prosecutors to obtain convictions for pennies on the constitutional dollar compared to the more onerous procedure described and prescribed by the Bill of Rights, which devotes more than half its words to criminal procedure and jury trials.

Besides allowing the government to exercise forbidden powers and trample individual rights in particular cases, this retreat from meaningful judicial review creates a regulatory moral hazard. When lobbyists, legislators, and administrators know their actions are unlikely to face serious constitutional scrutiny, they have strong incentives to pursue regulations (often at the behest of powerful interest groups seeking to sandbag potential competitors) that they might otherwise have eschewed based on rational cost-benefit analysis regarding the prospects of having their regulatory handiwork undone by a subsequent reviewing court. The judiciary’s excessive deference thus actively encourages the proliferation of laws and regulations that might not survive more rigorous review.

Another point Over Ruled helpfully has addressed in grappling with the root causes of excessive regulation is the possibility that excessive judicial deference to dubious laws may flow in part from the federal judiciary’s institutional composition. A 2018 study by my Cato colleagues and me documents that approximately 44 percent of federal judges previously served as government advocates in court, while only about 6 percent spent significant time opposing government power as public defenders, public interest lawyers, or similar advocates. And while individual judges can certainly remain neutral notwithstanding their own prior professional experiences, this dramatic institutional skew toward those who went from advancing often quite aggressive arguments for government power to adjudicating those same kinds of arguments almost certainly has a marked effect on the development of legal doctrine.

None of this negates the value of Gorsuch’s core observation about the freedom-stifling proliferation of law or his compelling examples of regulatory and prosecutorial absurdity. But addressing “too much law” requires examining not just the producers of law but also the institutional constraints—or lack thereof—that shape their incentives. A judiciary that consistently enforces constitutional limits on government power will discourage excessive regulation by raising its costs. Conversely, a judiciary composed disproportionately of former prosecutors and other courtroom advocates for government that defers too much to the other branches will encourage precisely the kind of regulatory excess that Gorsuch decries.

Justice Gorsuch performs a valuable service by vividly cataloging the shocking extent of overregulation in America today. But neither he nor we should be surprised when legislators and administrators do precisely what the Founders expected them to do—namely, constantly test the outer limits of their own power and relentlessly push for more. Thus, real reform will require confronting uncomfortable questions about the judiciary’s role in enabling regulatory excess and considering how judicial doctrine and institutional culture may systematically favor extra-constitutional government power over individual rights. 

Only by restoring meaningful judicial review, aka “judicial engagement,” can we begin to restore the full measure of constitutionally prescribed limits on government power.

David Inserra

Matt Mittelsteadt

At the recent AI Action Summit, Vice President JD Vance delivered a barnburner debut on the international policy stage. In his fifteen-minute address, the vice president distilled Trump-era AI policy: colored by optimism for AI potential, a dismissal of AI safety regulation, a muscular resistance to European rules, pro-worker policy, and the explicit sense that AI will be a potent tool to deter American adversaries.

The stand against European regulators and AI safety regulations made the speech an instant hit in free-market circles. Vice President Vance is right to call out such overregulation and its problematic consequences. However, the speech’s overall nationalist tone and countless overlooked market intervention dog whistles should raise concerns. For example, in the speech, Vance explicitly promises to center labor unions in AI policy decisions—inviting the possibility of automation halting red tape—and claims “the Trump administration will ensure that the most powerful AI systems are built in the US with American-designed and manufactured chips”—a gesture at large-scale tech tariffs and industrial policy. A pure hands-off approach, this is not. 

Vance’s most concerning assertion was that:

“[T]he Trump administration will ensure that AI systems developed in America are free from ideological bias and never restrict our citizens’ right to free speech.”

This statement should set off alarm bells. In recent years, there have been countless Republican-led legislative attempts to steer digital content in the name of combating “conservative censorship.” While the Supreme Court struck these bills down and affirmed the right to editorial discretion for digital platforms, these impulses live on and are now being applied to artificial intelligence.

Despite Vance’s free speech overtures, “to ensure” AI models are free from ideology indicates intervention. If Vance wants to stamp out what he views as bias, he would need either an explicit AI content regulatory regime or a significant pressure campaign to compel Silicon Valley to follow his party line.

While some might claim such market intervention clashes with the primary thesis of the speech, worry over the administration’s interventionist impulses is indeed well founded. On February 18, President Trump announced his administration would proceed with semiconductor tariffs—a truly earthshaking, explicit tech market intervention. This move on tariffs demonstrates that while AI safety intervention specifically might be off the table, intervention in general is still ago.

The Consequences of AI Ideological Purity Tests

Even if one has sympathy for Vance’s desire to stamp out AI bias, any direct rules or efforts to influence Silicon Valley design would carry powerful consequences.

1. Freedom

The most immediate consequence would be the impact on freedom. By restricting what AI models can generate, the administration would limit individuals’ ability to build what they want and interact with AI on their own terms.

Across ideological divides, the AI marketplace is awash with intentionally biased systems. In Lucerne, Switzerland, an AI-powered Jesus, naturally biased towards the Christian faith, was spun up to chat with curious parishioners. Meanwhile, on CharacterAI, one can find a Karl Marx bot who assuredly has a lot to say in support of communism. One can disagree with either of these bots, but people have a right to make them.

Such intentional biases also exist in leading models. For instance, xAI’s Grok3 expresses deliberate bias against legacy media while pointing users to x.com for news. OpenAI’s ChatGPT is also biased, albeit more subtly, with ideological choices outlined in their public “Model Spec.” Despite statements that OpenAI’s technology will “assume an objective point of view,” the avoidance of racial stereotypes and the confident condemnation of genocide as an evil are widely accepted, yet still ideological choices.

It is a fact that all AI models will be biased in some way and that true “objectivity” cannot exist. The best way to prevent an AI information monoculture and ensure that AI systems trend towards good information is not through strict government standards but through a strong diversity of market choices. Because organizations like OpenAI and xAI are free to compete on ideological lines, consumers now enjoy the freedom to choose the model that fits their views and even compare model responses to uncover biases.

2. Chilling Innovation

Demanding ideological orthodoxy would also chill innovation. Today, creative risks and fast model releases color the AI market and have enabled rapid discovery, learning, and course correction. Under any “objectivity” mandate, releases will slow, as any model not vetted for political correctness would be a liability.

This would sacrifice the open market’s powerful learning opportunities. In 2016, Microsoft released Tay, an AI chatbot that quickly descended into generating offensive content due to Microsoft’s decision to enable dynamic learning through interactions with Twitter users. Tay was a PR debacle, but it was also a learning experience. Within 48 hours, Tay was pulled offline, and engineers quickly internalized the stark (and now obvious) lesson: don’t let Twitter users dictate your AI’s design. This was both innovation and the market at work. Since then, safeguards and best practices have developed to avoid a Tay part 2.

While hopefully mistakes as glaring as Tay are behind us, future missteps will happen. Only by enabling continuous iteration can we ensure such mistakes are increasingly rare and increasingly low stakes. Under government pressure, however, such improvement will take a back seat to concern that a release might step on someone’s toes.

3. Market Access and Global Competitiveness

Perhaps the most serious long-term risk of government-imposed AI ideology is its impact on international perceptions. If overt political influence poisons American AI models, they risk being seen not as tools but as instruments of US government propaganda.

We already see this phenomenon with China’s DeepSeek R1, an AI model whose cloud release is rife with predictable government restrictions on topics like Tiananmen Square. While DeepSeek’s release could have been a pure story of technical achievement, Xi Jinping’s political taint saddled the company with immediate global skepticism and even country-wide bans.

Deepseek should serve as a stark warning: liberalism has market value. If American technology reflects the political preferences of its ruling party, global trust in US AI innovation will erode and markets will be lost.

We are already seeing early signs of this concern. Recently in the Financial Times, a European contributor wrote that entanglements between US tech firms and the Trump administration pose “a direct threat to European sovereignty and value.” While I’d like to handwave this as par for the course with longstanding EU-US tech policy disagreements, I worry this reflects a growing global distrust of politicized American technology, a distrust that could easily spread beyond Europe to markets with longstanding anti-imperialist traditions.

For continued American innovation and success, foreign markets are essential. If our AI models become a tool of propaganda, no one will want to use them, and America will fall behind.

The Continued Need for Non-Intervention

A heavily regulatory approach to AI policy under Trump is not inevitable, yet it is concerningly possible given the anti-tech and pro-industrial policy pushed.

Just because the administration criticized European AI regulations and rescinded President Biden’s AI executive order does not mean his administration’s approach won’t consider its own problematic regulations of these important technologies. Four years is a long time, and AI policy is still in its formative stages, and regulatory intervention could have consequences that change the trajectory or eliminate beneficial uses along with harms.

For those who value freedom, innovation, and global competitiveness, the message is clear: stay vigilant. The regulatory trajectory of AI in the US is far from settled and the consequences could be profound.

Walter Olson

Colin Grabow

Citing alleged unfairness in the current trading system, President Donald Trump appears to be laying the groundwork for the adoption of a tariff policy based on reciprocity. Whatever tariffs other countries levy on US imports will be the tariff charged by the United States on their same products. It’s not a new position for Trump, who also enthused about reciprocal tariffs during his first term.

The approach no doubt has superficial appeal. Why not give countries a taste of their own medicine? What’s good for the goose is good for the gander, and all that.

Closer scrutiny, however, reveals numerous flaws in such thinking. 

Tariff reciprocity means higher US tariffs. Although fairly average compared to other wealthy, industrialized countries, US tariff rates are at the lower end of the global scale. This necessarily means that a reciprocal tariff policy will lead to higher tariffs. Consumers will face higher prices for the goods they purchase, and businesses will pay for more critical inputs, reducing their competitiveness.

This is antithetical to US prosperity. It cannot be overemphasized that a policy of relatively low tariffs and trade barriers is not a favor the United States grants others but a key ingredient in its own economic success. To abandon this would be folly.

Tariff reciprocity surrenders US decision-making over trade policy. US tariff policy should be dictated by self-interest. A policy based on reciprocity, however, essentially hands policymaking to US trade partners. If these countries opt to pursue a policy of high tariffs, that’s regrettable. But it is not a decision the United States should be compelled to mimic. If others decide to jump off bridges (in a manner of speaking), it’s no reason for the United States to follow suit.

Tariff reciprocity means greater complexity. Under the current US tariff schedule, most items face one of two duty rates. The general one, applied to imports from most countries, serves as a default, while a lower, preferential rate applies to goods from countries the United States has free trade agreements with or that fall under special exceptions such as the Generalized System of Preferences. A tariff schedule based on reciprocity would prove enormously more complicated.

As economist Douglas Irwin recently pointed out, the US tariff schedule consists of approximately 13,000 line items detailing the various duty rates on imported items. Given that the United States trades with approximately 200 countries, this means about 2.6 million individual tariff rates. Spare a thought for both the headaches and hassle this would cause for a company trying to build an efficient supply chain, as well as the resources devoted to compliance.

Administering the tariffs, meanwhile, would be no easy feat either and is sharply at odds with efforts to improve government efficiency.

Not all foreign tariff reductions will prove beneficial. A lowering of tariffs by US trade partners in response to reciprocity would offset some of the pain of this new approach. But one must also consider that many (most?) countries will not do so. After all, tariff rates are not set at random but often reflect the pressures of domestic constituencies that may not dissipate in the face of higher US duties.

Furthermore (and perhaps more importantly), even a response that produces lower tariff rates could have little benefit. Consider the example of coffee. Brazil is the top supplier of coffee beans to the United States, to which a tariff rate of zero percent is applied. In contrast, Brazil applies a tariff rate of nine percent to US coffee exports.

Grossly unfair, no? Perhaps.

But here’s the rub: if Brazil responded to reciprocal US tariffs by slashing its own tariff rate to zero, the economic payoff to the United States would be almost nil. A mere 11.5 million pounds is produced in the United States (all of it in Hawaii), which is just seven-tenths of one percent of US coffee consumption. Only a subset of that production would be for export, and only a subset of that would conceivably go to Brazil.

The payoff of a reciprocal Brazilian tariff reduction would be tiny, while should Brazil fail to respond with a tariff lowering, American consumers (and businesses such as coffee shops and retailers) would be stuck with costlier cups of joe.

Even tariff workarounds bring added costs. President Trump has highlighted a seemingly simple means of avoiding US tariffs: transferring production to the United States. While true (to an extent, as imported inputs used in production could still face tariffs), it’s not apparent that this is actually desirable. If products are currently manufactured abroad, it’s a good indication that it’s more efficient to do so. Setting up shop in the United States may escape the burden of tariffs but will bring with it higher costs than had production been maintained overseas. 

Those higher costs will mean reduced consumption or less money for Americans to spend and invest elsewhere in the economy. American prosperity is promoted by making products more affordable, not the reverse. The promise of new tariff-induced jobs and investments may seem attractive, but they can only come at the expense of other jobs and investments frittered away due to inevitably higher costs.

The United States should no more aspire to produce all of the goods it consumes than an individual should try to produce the food they eat or clothes they wear. Rather, the US (which is to say, its people and businesses) should leverage the rewards of comparative advantage. To do otherwise is a formula for self-impoverishment.

There’s a better path to reciprocity. If President Trump desires reciprocal tariffs, free trade agreements (FTAs) are a better option. Under such deals, the United States and partner countries agree to remove or reduce trade barriers, including tariffs.

The Trans-Pacific Partnership trade deal signed by the United States and a number of its top trading partners in 2016, for example, would have eliminated 18,000 tariffs. (Unfortunately, President Trump withdrew from that agreement as one of his first acts after taking office in 2017.) A more recent agreement reached by the European Union and the Mercosur trading bloc eliminates tariffs covering 90 percent of bilateral trade. 

FTAs offer a straightforward approach for those interested in low, reciprocal tariffs. Sadly, however, the United States has not successfully negotiated and signed into law a new free trade agreement since its conclusion of bilateral deals with Colombia, Panama, and South Korea that took effect in 2012. 

Summing up

US policy should be determined not by what is perceived as fair but by what is in the country’s economic self-interest. Reciprocal tariffs are very much at odds with this. Such a policy would surrender US policymaking to foreign governments, increase prices for American consumers, and create new burdens for the country’s businesses.

To achieve a more favorable trading environment, President Trump should occupy the US chair at the negotiating table and negotiate new free trade agreements in which all parties reduce their tariffs to zero. It’s a seat that has been cold for too long.