Editor’s Pick

Gene Healy

Man, libertarians are a difficult bunch. Give them a government agency tasked with “delet[ing] entire agencies,” and they’ll start yowling about the agency-deleting agency.

The zanier version of John Galt has been set loose on the permanent bureaucracy, invoking Milton Friedman while he throws the US Agency for International Development (USAID) “into the woodchipper.” Yet rather than welcoming the new Department of Government Efficiency (DOGE), a lot of libertarians are filled with dread:

• Is DOGE “just the Deep State by a different name?”
• Or is it a modern-day CREEP, Nixon’s dirty-tricks-campaign operation, “on steroids”?
• Does Elon Musk think he’s “Dictator of America”? What he’s doing “bears all the hallmarks of a coup d’état”!

“You just can’t please these people,” I can hear our critics on the right saying, “It’s the Movement that Would Not Take Yes for an Answer.” But I like to think DOGE skepticism comes from a good place: Libertarians take “eternal vigilance” really seriously.

Even so, I am unconvinced—to put it mildly—that the agency-for-eliminating-unconstitutional-agencies is a serious threat to constitutional government. DOGE’s legal defects are a far bigger threat to the agency’s bureaucracy-slashing mission than they are to our liberties. And since a lot more than USAID deserves to go into the woodchipper, libertarians should want those defects cured because we want DOGE’s mission to succeed.

The Stuff of Libertarian Nightmares?

Be careful what you wish for, Veronique de Rugy cautions in Reason. When one president “circumvents legal constraints to impose libertarian-leaning policies,” the next can use “the same unchecked powers to expand government.” In the worst case, DOGE could leave us with a “presidency on steroids,” “opening the door to the same abuse when the left is in power.”

That’s sound advice in general, but in this particular case, I’m struggling to come up with true nightmare scenarios.

There are three main areas where DOGE might further weaken the remaining constraints on executive power:

• Impoundment: The administration insists the “power of the pen” trumps Congress’s power of the purse, and the president can simply nullify appropriations he doesn’t like.
• Bypassing Senate confirmation: They’ve also played fast and loose with Musk’s legal status, arguably flouting the Constitution’s Appointments Clause.
• Expanded removal powers: And they’ve claimed sweeping authority to pink-slip federal employees en masse, civil-service protections notwithstanding.

Which of these moves, if successful, would spell doomsday for libertarians when “the left is in power”?

Is the big threat here impoundment? Beware, lest come 2029, President Gavin Newsom starts zeroing out all the agencies libertarians like? (NASA geeks and “state capacity libertarians” hardest hit).

Is it bypassing the Senate? Will Newsom govern through unconfirmed apparatchiks? What if he enthrones a “pay czar” to determine CEO salaries or a “car czar” who calls the shots for US auto manufacturers?

Or is it the threat to federal civil-service laws that’s supposed to keep libertarians up at night? If Trump’s “Schedule F” gambit succeeds, will President Newsom pack the commanding heights of the administrative state with Democratic loyalists?

He could, I guess. I doubt we’d notice the difference. In the 2024 cycle, 84 percent of federal employees’ campaign contributions went to Harris-Walz—99 percent of EPA donations, 100 percent at the Department of Education. Americans are already governed by a bureaucratic class well to the left of the median voter.

Are there risks to Musk’s berserker approach? Sure, but given where we are now, catastrophizing about DOGE is a little like telling Javier Milei to slow down because his successor might be a Peronist.

To be fair, I think impoundment is a power grab worth worrying about—but not because I fear a future president may cut too much (don’t threaten me with an imaginary good time). Instead, the risk is that presidents will use budget-slashing authority to punish legislators who don’t fall into line—further accelerating our slide toward one-man rule.

The other two issues—Musk’s slippery constitutional status and the push for greater authority to fire bureaucrats—cause me far less agita, for reasons I’ll explain below.

What Is Elon Musk?

Trump’s legal team has proved comically incapable of a straight answer to that question. “I don’t have any information beyond he’s a close adviser to the president,” an administration lawyer burbled in a DC district court hearing last month. But he’s definitely not the DOGE administrator—that’s somebody named Amy Gleason. 

If a federal officer exercises “significant authority,” he’s supposed to go through the constitutional appointments process; for “principal officers,” that means getting confirmed by the Senate. Musk hasn’t been, to the consternation of multiple federal judges.

The constitutional formalities matter; still, in this case you might have a hard time explaining to a normal person how skirting them has made Musk uniquely dangerous: “What do you mean he’s not under democratic control? He works for the president. Trump can overrule anything he does and fire him at any time.”

And is Musk exercising “significant authority” in the constitutional sense: making final decisions that bind the government? The legal scholar Michael McConnell argues that he’s basically another executive-branch “czar” with “no line authority but a great deal of influence.” DOGE’s power “is simply to make recommendations to the agency”—a point that seems to have been impressed on Musk at a contentious Cabinet meeting on March 6.

Either way, it seems to me that the best reason to follow the formalities is that it would put DOGE on sounder legal footing and help ensure that the cuts and the firings stick.

“You’re Fired”

Then there’s the administration’s bid for expanded removal authority: Trump’s “Schedule F” order authorizes stripping civil-service protections from thousands of senior executive-branch officials and shifting them to employment-at-will.

Proponents of Unitary Executive Theory insist that’s perfectly constitutional: Federal civil service laws can’t “limit the president’s power to remove policymaking officials” without encroaching on core Article II authority. And while Unitarians have made all manner of extravagant claims about the scope of “the executive power,” their narrowest and most defensible position is that Article II makes the president the boss of the executive branch, with the power to fire anyone using their borrowed authority in ways he disapproves.

Still, you don’t even have to reach the constitutional question if, as my colleague Bob Levy suggests, Trump already has statutory authority to make the change. “The president’s legal argument seems well-grounded,” Bob writes: “The Civil Service Reform Act of 1978 exempts positions ‘determined to be of a confidential, policy-determining, policy-making or policy-advocating character.’” Likewise, Trump’s bid for “the restoration of presidential authority over so-called independent agencies … seems consistent with our tripartite constitutional plan.”

That these moves are legally defensible doesn’t mean they’re risk-free. Even the narrower versions of Unitary Executive Theory offer presidents broad latitude for mischief.

But the biggest risks of “a presidency on steroids” aren’t coming from DOGE. Trump’s initial flurry of executive orders includes several that assert a George W. Bush–style royal prerogative to ignore or rewrite the law. Another order dusts off the 1798 Alien Enemies Act as authority for summary deportations and, apparently, warrantless searches of homes. Then there’s Trump’s attempt to convert the International Emergency Economic Powers Act of 1977 (IEEPA) into a trade-war weapon. No prior president has ever used IEEPA for across-the-board tariffs on allies and major trading partners. But that’s exactly how Trump is using it, in service of “the dumbest trade war in history.” If, amid the excesses of George W. Bush’s “Terror Presidency,” libertarians spent their time fretting about whether Dick Cheney’s National Energy Task Force was FACA-compliant, you’d worry they’d lost the plot.

It’s Illegality All the Way Down

To put things in perspective, look at the agencies DOGE has gone after hardest: USAID, the National Endowment for Democracy (NED), the Consumer Financial Protection Bureau (CFPB), Federal Emergency Management Agency (FEMA), and the Department of Education.

They’re all unconstitutional. Some, like CFBP, wield lawmaking powers Congress has no business delegating; others, like the Department of Education, commandeer responsibilities the Constitution leaves to the states and the people. Not one has an enumerated power that can plausibly support it.

What’s more, they’re all agencies Cato scholars have called for shuttering, in most cases for at least three decades. FEMA has been a perennial target of Chris Edwards’ Downsizing Government project; CFPB the target of multiple Cato amicus briefs from its inception. USAID, NED, and the Department of Education were on the Institute’s hit list in the first Cato Handbook for Congress (1995), the last time Washington was forced to have a serious conversation about serious cuts.

The agencies in DOGE’s crosshairs shouldn’t exist. That doesn’t justify a by-any-means-necessary approach to eliminating them. To the contrary: Since the federal leviathan is a creature of statutory law and congressional appropriations, there’s no dismantling it without buy-in from the branch that got us into this mess.

DOGE Needs Congress

DOGE’s legal defects are fixable, but they need to be fixed by Congress. I offered one method in a recent piece for Reason: renewed reorganization authority of the sort Congress delegated to multiple presidents from 1932 to 1984. The Reorganizing Government Act of 2025, introduced last month by Sen. Mike Lee (R–UT) and Rep. James Comer (R–KY), would give the administration sweeping power to restructure and downsize the bureaucracy, subject to a fast-tracked approval vote in Congress. It would also insulate those cuts from legal challenges.

Other ideas come from Sen. Rand Paul (R‑KY), who’s said DOGE’s cuts should be “bundled into a rescission package and sent back to Congress”; and Yale Law’s Jed Rubenfeld, who argues that “Congress Can Easily Green-Light DOGE” via a single sentence approving cuts in a reconciliation bill. 

Is any of this about to happen? Don’t hold your breath. Two weeks ago, when Senator Paul forced a vote on codifying DOGE’s USAID cuts, nearly half of the GOP caucus balked. Congress shows every inclination to keep kicking the can down the road, using media hyperventilation about DOGE as cover for doing nothing.

We should be clear-eyed about this: The smart money says DOGE will probably fail, leaving the permanent bureaucracy little worse for the wear. But realism shouldn’t mean fatalism. In this town, opportunities for radical cuts come maybe once a generation. Libertarians should be looking for ways to even up the odds.

Adam N. Michel and Joshua Loucks

When Congress passed the Inflation Reduction Act (IRA), it was told the new energy tax credits would cost about $270 billion over a decade. Revised official estimates put the cost at multiple times that amount, as much as $786 billion. But congressional scorekeepers may still be getting the long-term cost of the IRA energy subsidies wrong, and fixing this mistake is key to unlocking permanent pro-growth tax cuts.

A new Cato report by Travis Fisher and Joshua Loucks (one of the authors here) argues that two of the most expensive IRA tax credits could be functionally unlimited subsidies, costing as much as $180 billion annually by 2050. With no expiration in sight, the total cost of the IRA green energy subsidies could accumulate to as much as $4.7 trillion by the middle of the century.

These infinite tax credits can offset permanent extensions to the most pro-growth features of the Tax Cuts and Jobs Act of 2017, which largely expires at the end of the year. Without bending long-standing budget rules, Congress must ensure that any permanent tax cuts outside the 10-year budget window are offset with spending cuts or higher revenue. This rule is why tax cuts are often temporary.

Repealing two of the IRA’s open-ended tax credits could more than offset the revenue loss from cutting the corporate tax rate to 15 percent, restoring R&D expensing, fixing the interest deduction limit, and enacting full expensing.

Budget Scorekeepers Blew It; Fixing Their Mistake Is Critical

The Joint Committee on Taxation (JCT) badly underestimated the cost of the IRA energy tax subsidies. It has since revised many of the original estimates to more accurately reflect the consensus estimates that put the 10-year cost of the IRA at roughly $1 trillion. Government scorekeepers are very likely still not fully incorporating the long-run, uncapped costs of the production tax credit (PTC) and the investment tax credit (ITC). 

As Fisher has repeatedly pointed out, the ITC and PTC are functionally uncapped because they are not time-limited, like most other temporary tax credits. The ITC and PTC only phase down when the level of greenhouse gas (GHG) emissions from the electricity sector is reduced by 75 percent of the 2022 baseline. 

Figure 1 is adapted from Fisher and Loucks’ recent report. It shows GHG projections from the Energy Information Administration, which show that electricity-sector emissions will remain far above the IRA’s target of a 75 percent reduction in the 2022 level through 2050. The GHG phasedown trigger is not met even in the scenario with a high uptake of IRA subsidies.

The Treasury Department’s most recent tax expenditure estimates explicitly state they assume the subsidies will begin to phase out as early as 2034. The JCT could have made a similar mistake in its original score of the IRA. Had the cost estimates appropriately accounted for both the magnitude and the open-ended credits, the bill could not have passed using the budget reconciliation process, which requires that legislation not increase the deficit beyond a 10-year window.

A Golden Opportunity for Pro-Growth Reform

As Republicans work to repeal or revise the IRA credits in reconciliation, it is critical that the JCT properly accounts for the full cost of the credits outside the budget window. By 2034, the ITC and PTC alone will cost about $130 billion per year—enough to permanently finance four of the most pro-growth tax cuts under consideration. 

Figure 2 compares Cato’s $130 billion cost estimate for the uncapped energy credits with the Tax Foundation’s static estimates of four pro-growth tax cuts that reduce revenue by approximately $113 billion in 2034. The four changes are (1) cutting the corporate tax rate from 21 percent to 15 percent, (2) renewing full investment expensing, (3) returning the interest limitation to its less restrictive definition, and (4) allowing full research spending deductions. 

The PTC and ITC are projected to continue to expand past 2034. In 2050, the two credits will cost as much as $180 billion annually.

Other pro-growth reforms, such as neutral cost recovery, have increasing static fiscal costs outside the budget window. Although neutral cost recovery is not included in our estimates, the escalating open-ended cost of the PTC, ITC, and other IRA provisions, such as the advanced manufacturing credit sub-provisions for critical minerals, would more than pay for neutral cost recovery even in the years beyond the budget window.

The IRA’s unlimited tax credits are a fiscal time bomb, costing between $2.04 trillion and $4.67 trillion by 2050. They represent an unchecked expansion of government spending with no clear end date. Policymakers have a rare opportunity to repeal these costly provisions and use the savings to fund permanent, pro-growth tax reform that benefits the entire economy—not just politically favored industries. Getting the scoring right is critical. 

Chris Edwards

Those acquainted with English history know that in 1215 AD, land-owning barons pressed King John to approve the Magna Carta, which limited his power and helped secure individual liberties. John “was a very poor king” who “levied frequent taxes,” according one history on my bookshelf.

However, an article in BBC History magazine informs me that “Magna Carta” historically referred to a revised 1225 version of the charter. It is this later version that influenced government and law in England for centuries. So, happy 800th birthday, Magna Carta.

The battles to rein in kings and limit government power went on for centuries. The Magna Carta formed the basis of subsequent statutes and played a key role in educating each generation about their liberties. But even before the early 1200s, the Magna Carta had precursors, including a “Charter of Liberties” issued by King Henry I in 1100 AD.

America’s limited-government Founding could not have happened without the centuries of struggle and progress made by the English. Our National Archives holds a 1297 version of the Magna Carta, courtesy of David Rubenstein. The Archives Foundation notes, “Magna Carta also guaranteed due process of law, freedom from arbitrary imprisonment, trial by a jury of peers, and other fundamental rights that inspired and informed the Founding Fathers of our nation when they wrote the Declaration of Independence, United States Constitution, and Bill of Rights.”

Some people worry that our current president is acting like an unhinged king, such as the chaotic way he is unilaterally imposing hefty trade barriers. Interestingly, the Magna Carta tried to guard against this precise abuse: “All merchants are to be safe and secure in departing from and coming to England, and in their residing and movements in England, by both land and water, for buying and selling, without any evil exactions but only paying the ancient and rightful customs.” (Clause 41 of the 1215 charter).

Here are excerpts from the BBC History article by David Carpenter, a professor of medieval history at King’s College, London:

That [1215] document, sealed under duress from rebel barons and famously subjecting the king to the law and promising justice to all, is widely lauded today as a milestone in the establishment of English law. Yet the fact is that, in the centuries after 1215, the term Magna Carta was hardly ever used for John’s document. Rather, that was the name given to the charter issued nearly a decade later by King John’s son, Henry III, at Westminster on 11 February 1225.

Although Henry’s charter was based on his father’s, the earlier iteration was usually called simply the “Charter of Runnymede.” So when, in 1297 and 1300, Henry’s son Edward I confirmed Magna Carta, he meant the charter of 1225—and the same was true of all later confirmations.

King John quashed the 1215 charter, and the barons deposed him. In 1217, John’s minor son, King Henry III, and his court made another deal with the barons to issue a new version of the charter. This charter was

accompanied by an entirely new charter regulating the running of the royal forest. It was at this point that the term Magna Carta (“Great Charter”) was first introduced, referring not to its grandeur by to its size, to distinguish that larger charter from the smaller Charter of the Forest.

In 1225, a grown Henry III wanted to fight a war in France to be funded by higher taxes, and he needed buy-in from the barons. He agreed to a new version of the Magna Carta. The 1215 and 2025 versions were cowritten by Stephen Langton, the Archbishop of Canterbury.

For Langton, one of the great biblical scholars of the age, [the charter] responded to deeply held ideas about the need for rulers to be limited by the law.

… Langton, responding to ideas about how just rule should benefit the whole community, also did something to make Magna Carta more inclusive. The concessions in the charters of 1216 and 1217, like those in 1215, had been made only to people who were free—thus excluding the unfree peasantry who made up the largest part of the population. In 1225, this changed. The statement that the concessions had been made to the free remained, but it was qualified—indeed, contradicted by the new preamble in which these concessions were granted to “everyone in the kingdom.”

Carpenter notes that the Magna Carta was not without flaws, but

The texts of the charters also became widely known and, thus, sank deep roots into English society. Copies were made by religious houses, magnates, ministers, knights, freemen and lawyers. Of the various versions, that of 1225 was easily the most copied.

… Those examining the texts—and marginal annotations bear witness to intensive study—could see that Magna Carta was far more than just a baronial document. Chapters such as those regulating the levying of fines and the running of local government benefited wide sections of society.

Scott Lincicome

Romina Boccia and Dominik Lett

The Department of Government Efficiency (DOGE) has encountered repeated legal setbacks as a result of its efforts to shrink the federal workforce, cut costs, and reshape the executive branch. Part of the challenge that the administration faces is a constitutional one—Congress, not the executive or judicial branches, is vested with the power of the purse. This means that congressional action is unavoidable if the administration wants to deliver substantial, lasting savings. 

To add legitimacy to DOGE’s efforts to cut spending and encourage Congress to take responsibility, the president should submit a rescissions package (or multiple) under the Impoundment Control Act that officially requests Congress to cancel certain spending.

Once a regular feature of the budget process, presidents across both parties used to routinely submit rescission packages to Congress, which recommended the cancellation of already appropriated spending. Reviving these rescission packages would be a return to responsible fiscal governance, realizing DOGE cost savings and ensuring courts do not roll back the administration’s actions.

Below, we detail the Impoundment Control Act (ICA) rescissions process, its historical track record, and the path forward.

Rescissions 101

Under the ICA, Congress established a special, expedited process to accelerate the cancellation of previously approved spending. This budget process, known as rescission, allows the president to propose cancelling specific spending by submitting a detailed request to Congress. This starts a 45-day shot clock, during which the president can legally pause congressionally approved spending, and the legislature considers the rescission request. Notably, this process only requires a simple majority in both chambers, bypassing the Senate filibuster. If approved and signed by the president, this spending is legally cancelled.

In prior decades, presidents and Congress used this presidentially initiated rescissions process, in combination with congressionally initiated rescissions, to achieve modest savings.

The Track Record on Presidential Rescissions

According to a Congressional Research Service (CRS) report, presidentially proposed rescissions were once a more common feature of the budget process but have since become less prevalent. That’s unfortunate given that lawmakers used to save taxpayers billions of dollars using the process. As the graph below shows, Congress and the president jointly cancelled roughly $152 billion in budget authority over three decades (1974—2003). These modest savings deserve recognition, in part because they demonstrate that Congress, together with the president, can actually cut spending if lawmakers put their minds to it.

While Presidents Ford and Carter both proposed a series of small rescissions, Reagan demonstrated that presidentially initiated rescissions were a viable mechanism to achieve major spending cuts. In 1981, Reagan requested $15 billion be rescinded—roughly a 4 percent cut to discretionary budget authority at the time. While Congress did not approve all his cuts, they did eventually enact nearly $15 billion in rescissions that year and approved another $5 billion in cuts in 1982.

In subsequent years, most rescissions were initiated by Congress rather than presidents. Granted, both Presidents George H.W. Bush and Clinton successfully leveraged presidentially proposed rescissions, and each oversaw the enactment of significant rescission packages during their administrations. Clinton, the last president to have a budget surplus in his administration, was also the last president to have a presidentially initiated rescission enacted. In total, Congress approved roughly $25 billion in presidential rescission requests across five administrations but enacted a much larger $152 billion in total rescissions across the same period.

It’s worth noting an important limitation built into the presidentially proposed rescissions process: the president can only request rescissions to discretionary spending. Because discretionary spending accounts for just over a quarter of all federal spending, it limits potential savings from pursuing rescissions. Still, with discretionary budget authority at $1.6 trillion per year and counting, neither the president nor Congress has any excuse to sidestep much-needed spending restraint, and ICA rescissions remain an important tool to bring rampant, wasteful spending under control.

Why Hasn’t Trump Proposed a Rescission?

Given the administration’s stated cost-cutting enthusiasm with DOGE and the OMB, it is surprising that the administration has not proposed an ICA rescission package. After all, it was the Trump administration in 2018 that proposed the first rescission package to Congress since Clinton, requesting $15 billion in cuts (just 1 percent of discretionary budget authority). While the House narrowly approved this proposal, it ultimately failed in the Senate, with two Republicans defecting. This failure might deter the administration today. If Congress didn’t have the appetite for cuts then, why try now? Additionally, there is reportedly some internal debate over the constitutionality of the ICA itself, likely reducing interest in the rescissions process.

Arguably, the current political landscape is more favorable for a presidential rescissions package than it was in 2018. The budget is bloated from a pandemic spending spree. Meanwhile, high bond yields, inflation, and DOGE have helped generate a national conversation about the need for spending cuts. While it is true that Republicans have a slimmer House majority, the most recent budget framework is evidence that the House is actively identifying spending cuts. And compared to 2018, Republicans have an extra two senators that could make a rescission package politically feasible.

There is no time like the present to cut spending, and it helps to have an administration that throws its weight behind a spending cuts package. As Rollcall’s David Lerman has reported, legislators have expressed frustration about “being kept in the dark as departments and agencies fire workers.” A presidentially proposed rescission package could build political support for DOGE’s spending cuts and workforce reductions by bringing transparency to the current process, involving legislators directly, and pursuing cuts via a legal and lasting mechanism.

Optimistically, Elon Musk recently expressed openness to the idea of presidentially proposed rescissions during a lunch with Republican senators (it doesn’t appear Musk was aware of the ICA process prior to the lunch). As Sen. Rand Paul (R‑KY) rightly argued, the administration should consider clawing back hundreds of billions in already appropriated funds through rescissions, something that could “make those of us who are skeptical about reconciliation, adding to the debt, […] be a little bit more open to it if we’re actually going to … make the spending cuts real and permanent.”

The Path Forward

Moving forward, DOGE and the administration should actively collaborate with Congress and regularly submit rescission packages to cancel wasteful and inappropriate spending. The administration could break these rescission requests into multiple smaller packages, following the agency-by-agency approach of DOGE, or they could submit a single, comprehensive rescissions package covering a wide array of policy areas. Either approach could add pressure for Congress to rise to the occasion and govern responsibly, bringing transparency, legitimacy, and permanence to an executive cost-cutting process that has otherwise been characterized by its secrecy, chaos, and policy whiplash.

As Elon Musk has noted, the federal government is a “target-rich environment” for cuts. We agree. It’s now time to put pen to paper and codify those cuts into law. Rescission packages are just the tool to do so.

Jeffrey Miron

On February 1, 2025, the White House announced a series of aggressive tariffs against Canada, Mexico, and China (since adjusted in numerous ways). One ostensible justification for the tariffs is to combat fentanyl smuggling. According to the statement, Customs and Border Protection seized more than 21,000 pounds of fentanyl at American borders last year, a drug that contributes to over 75,000 overdose deaths annually.

This approach will fail; fentanyl smuggling occurs because fentanyl is illegal, and higher tariffs do nothing to reduce the incentives prohibition creates for illicit trade.

Prohibition also spreads violence and corruption, both domestically and in source countries. Underground markets also impede quality control because consumers must guess the potency or composition of the drugs they consume, quadrupling drug-related poisonings and increasing emergency room visits by 60 percent in destination countries. A legal market would reduce these risks dramatically.

In short, punitive tariffs are not a solution to the problem of fentanyl smuggling. Instead, the U.S. should legalize not just fentanyl but all drugs. This would eliminate the black market and its attendant adverse consequences.

This article appeared on Substack on March 26, 2025. Jonah Karafiol, a student at Harvard College, co-wrote this post.