Subscribe now

Is that a sign of progress or decline? Our honest answer is “both,” which is not very satisfying. Below are 11 observations that explain why. Some surprised us.

1. The American economy today has a single driver of jobs: healthcare.

The rest of the jobs market has basically flatlined.

2. The growth is in headcount, not hours.

The growth has come almost exclusively on the back of new healthcare job creation, while hours per job fell.

3. The over-65 population has grown faster than healthcare jobs.

Healthcare jobs per older person have fallen over the past decade.

4. In California, New York, and Illinois, healthcare is all the job growth.

California, New York, and Illinois have had negative overall job growth since the pandemic when healthcare jobs are backed out. And none of the 15 largest states lost healthcare jobs.

5. Demographics explain hospital hiring, and little else.

One might expect this to be a clear story of an aging population driving up demand for healthcare workers. That is true to some extent, and explains most growth in hospital hiring. But based on population growth and aging alone, the actual increase in nursing home jobs is much lower than demographics would imply. Doctors’ offices, clinics, and home health went the other way: jobs more than doubled, against roughly 40 percent growth predicted by demographics.

6. Half the new jobs are home care aides.

The bigger surprise is who got hired. Half the jobs added since 2020 across major healthcare occupations went to home health and personal care aides, who earn about $36,000 a year.

7. Where Medicaid home-care spending grew, home-care jobs followed.

So what explains this? Medicaid. Compare the third of states where Medicaid home-care spending grew most with the third where it grew least. Home-care jobs grew about 15 percent faster in the first group, and the gap opened after federal pandemic money for home care arrived in 2021. Nursing home jobs moved the other way.

8. The work is done mostly by women and paid for largely by government.

The net result is fewer nursing home jobs than in 2019 and far more low-paid work in people’s homes. That work is done mostly by women, in home care disproportionately by immigrants, and paid for largely by government.

9. Jobs grew faster than pay.

As healthcare swells to nearly a fifth of the US economy, the extra money is going to headcount more than pay. In every fast-growing setting, jobs grew much faster than real pay. Since 2019, home care added 40 percent more jobs and raised real pay 16 percent. Outpatient centers added 22 percent more jobs and raised real pay not at all.

10. Output per healthcare job grew half as fast as the rest of the economy.

Zoom out, and one of the country’s slowest-improving industries is absorbing its workforce. Since 1998, healthcare jobs have grown 75 percent, four times the rate of the rest of the private economy. Output per healthcare job grew half as fast: 31 percent against 63.

11. Where productivity lags, prices rise and jobs pile up

This final chart shows the whole story: Baumol’s cost disease, the economic phenomenon where stagnant (often service) industries get steadily more expensive. Each bubble is an industry; the darker, the more its share of jobs grew.

In most service markets rising prices would push buyers away or they would substitute. You can always trade a $100 night out with friends for a night in with Netflix, or your fade-and-massage for fifteen minutes at Supercuts. You can’t really trade a colonoscopy or a trip to the ICU for WebMD. So when healthcare gets more expensive, we don’t buy less of it. We find a government form to fill out, and more people get hired to do the same thing the same old way.

Is America’s healthcare boom a sign of progress or decline? 🤔

Some of this is progress. Older Americans who once would have gone to nursing homes are getting care at home, which is what most of them say they want.

But the economics are grim. Every year, a larger share of American workers is employed where output per worker barely rises, much of it paid for by taxpayers, current and future. None of this is a knock on the people doing the work.

The picture that emerges is mixed. There’s progress for families getting care and a slower-growing economy for everyone paying for it.

Written by Anonymous Excel Guy

I am about to argue that I should be put out of a job.

My day-to-day job consists of collecting, validating, and ultimately reporting financial interactions between medical manufacturers and healthcare professionals to the US federal government for compliance with the Sunshine Act. In my position, most of my time is spent scrolling through thousands of rows of data in hopes of finding minor errors that, if not corrected, could lead to large financial penalties.

Subscribe now

Here’s what this looks like in practice. When healthcare professionals attend an industry-sponsored event, nearly everyone has a flight, hotel room, meals, and ground transportation. A meeting planner might inadvertently allocate a hotel charge to the wrong attendee or divide room expenses incorrectly. That’s where I come in. I analyze the data, make corrections for reporting, and communicate back to the meeting planners, who surely have much more important things on their mind than remembering the Sunshine Act. It’s tedious but necessary. An allocation error can create compliance exposure, so we spend considerable time making sure every dollar ends up in the right spreadsheet cell.

This is the kind of work that keeps people like me employed. And I don’t think it’s making healthcare any better.

Did you know there’s a website where you can look up your family’s doctors and see what kinds of financial relationships they have with the medical industry? Every consulting payment, every royalty payout, and every reportable meal paid for by a pharmaceutical company is meticulously published on an annual basis. If you didn’t know, you’re not alone. In my estimation, not only do most Americans not know this program exists, most doctors don’t remember it does either. It’s called the Open Payments program, and it is run by the Centers for Medicare & Medicaid Services (CMS).

The program was established by the Sunshine Act, passed in 2010 as part of the Affordable Care Act. Championed by Republican Senator Chuck Grassley, it launched in 2013. Since then, pharmaceutical and medical device companies have had to devote staff and resources to collecting and reporting financial transactions, lest they face potentially significant civil penalties for violating federal law. The intent of the Sunshine Act is noble. Particularly in the wake of the opioid epidemic, it’s reasonable to demand transparency in this space. Relationships between industry and doctors can influence medical decisions. If a doctor is getting paid tens of thousands of dollars by a company that manufactures a drug, how do we know that their prescription for that drug is based on what the patient actually needs?

But the Sunshine Act has three problems. First, it is far too expansive in scope, with mission creep already evident in the expansion of the program in 2021. Second, it imposes substantial burdens on industry without a clear corresponding benefit for patients. Third, while it requires extensive disclosure, it does little by itself to prevent misconduct.

The goal of transparency is good. The problem is a system that demands paperwork for a $25 lunch as well as a $25,000 consulting contract.

It’s Too Expansive

Out of more than 17 million transactions reported for calendar year 2025, almost 15 million were for food and beverage payments alone. Nearly $450 million in food and beverages was reported by 1,435 manufacturers. All the big players are here: Pfizer, Stryker, Eli Lilly. What is this telling us? That physicians across the country are in the pocket of Big Pharma? That the next time you see your primary care physician, you should ask about the $20 meal purchased on their behalf by AstraZeneca?

Anyone familiar with this element of the medical industry understands why it’s a questionable reporting metric. Meals are low in dollar value and high in transaction count. Not only are expensive steak dinners reported, but so are relatively small transactions that cross the applicable reporting thresholds. That’s every qualifying box of donuts brought in by a rep to a surgery center, every Starbucks coffee, every snack at a conference paid for by industry.

Business is conducted over meals, not only in the medical industry but in every industry. If you want to convince medical professionals that your new product is superior to the one they’re currently using, you might do that over food. Training is frequently conducted over meals as well. If your company has just signed a contract with a health system to purchase one of your products, you’ll need seminars to educate its staff. You can bet your employees are bringing in catering. Multiply that annoyance across every hospital and clinic a company serves, then across every manufacturer in the country.

These meal transactions tell us very little. They’re a metric of how much business the medical industry is conducting, not necessarily how much money nefarious otoscope manufacturers are putting into the pockets of doctors they want to undermine. Real compensation — consulting fees, honoraria, royalties — is significantly more meaningful. Those payments show a direct financial relationship between a company and a doctor rather than a minor transfer of value on the doctor’s behalf.

There is evidence that even small industry-sponsored meals can influence prescribing behavior, so the concern shouldn’t be dismissed outright. But that doesn’t establish that the current reporting requirements are proportionate to their benefits. A database containing millions of minor transactions doesn’t necessarily help patients distinguish meaningful conflicts of interest from ordinary business interactions. As of 2026, if a single covered recipient crosses the total $138.13 annual threshold, every single transaction for them is reportable, even if it’s only $1.

The expansion of Open Payments in 2021 also extended the definition of “covered recipient” beyond physicians to include advanced practice registered nurses, physician assistants, and anesthesiologist assistants. Again, the intent is noble. Nurse practitioners can prescribe drugs and often function as primary care providers. But the expansion also covers certified registered nurse anesthetists and anesthesiologist assistants, whose roles differ substantially from those of prescribing physicians. The more categories of professionals and transactions we add, the greater the compliance headache.

It’s Unnecessarily Burdensome

A small medical device company is unlikely to have a dedicated employee for complying with this regulation. Its transaction volume is too low. More likely, it’ll pull someone in accounting who has access to NetSuite and devote a few hours each month to the task. For a large company, entire teams may be dedicated to collecting transactions, cleaning up reportable data, and managing submissions. The penalties for noncompliance can be substantial, and the reputational consequences of an enforcement action can be even more damaging.

What we end up with is a laborious process of diving into one system for airfare payments, another for grants, and a third for consulting contracts. Much of the effort isn’t directed at finding misconduct but at assembling information for reporting.

Of course, every regulation imposes some burden. The question is whether the benefits justify the costs. For much of the Sunshine Act, I believe it does not.

Startups in healthcare are trying to create novel solutions to existing problems. It’s one of the best parts of the American healthcare system, regardless of all its faults, that we have a churn of baby companies trying something new. Threatening them with substantial penalties over administrative minutiae adds another burden to an already complicated regulatory environment.

For large, established companies, this is yet another department in a Byzantine administrative system. They can certainly afford it, but is it worth it? I’m not so concerned about the bottom lines of billion-dollar companies (Lord knows there’s enough bloat in there with or without the Sunshine Act), but I question whether a company manufacturing pacemakers for Americans’ hearts should devote so many resources to tracking relatively minor transactions that could otherwise be used elsewhere.

The Sunshine Act Doesn’t Actually Prevent Misconduct

Open Payments is a transparency program. It doesn’t directly prevent or prohibit financial relationships. Anti-kickback laws are what make it illegal for pharmaceutical companies to enter into certain backroom deals with doctors to prescribe unnecessary drugs in return for hidden payments. Those are the good laws — laws that make it illegal to harm patients who trust their doctors to uphold the Hippocratic oath — and they existed long before the Sunshine Act.

The Sunshine Act requires companies to report financial interactions to the government and publishes them on a website relatively few patients appear to know about. It’s true that investigators, journalists, and researchers can use those disclosures to uncover suspicious relationships. That’s a legitimate benefit, but reporting itself doesn’t prohibit misconduct. Kickbacks were illegal before the Sunshine Act and remain illegal today.

Imagine a sales rep who’s college friends with a surgeon. They conspire to make more money together: The surgeon agrees to overuse the company’s spinal fusion implants, the rep receives a massive sales bonus, and the surgeon secretly gets a cut. The only parties hurt are the patient and Medicare, which pays for an unnecessary procedure. That’s already illegal. Where exactly does Open Payments come into play? Do we really think the surgeon and rep are going to put their arrangement on record?

Sure, disclosure requirements can help investigators uncover suspicious patterns, and the possibility of public scrutiny may deter some questionable relationships. But those benefits have to be weighed against the costs of collecting and publishing millions of transactions that have little connection to misconduct.

If you want fewer inappropriate opioid prescriptions, regulate prescribing directly. If you want harsher penalties for kickbacks, make that change. Instead, we have a system that requires a great deal of reporting without a clear measure of what it accomplishes.

In my experience, many covered recipients don’t even check the transactions reported on their behalf, and many physicians barely remember the Sunshine Act exists. As for patients, how many know they can search for their doctor’s name? I’d be interested to know how many searches come from patients rather than manufacturers checking their own data.

If we’re going to require companies to spend countless hours collecting information, shouldn’t we have a better understanding of who uses that information and what value it provides?

Solutions

We should reform the Sunshine Act. Its expansive nature may be difficult to rein in by excluding particular categories of healthcare professionals, but we can remove mandatory reporting requirements for meals, travel, and lodging. These transactions are often secondary to the underlying financial relationship. If a physician is a consultant under contract, the company may pay for airfare and hotels. Is it the travel and lodging that’s important to report, or the consulting payments themselves?

My first proposal is to make food and beverage, travel, and lodging reporting voluntary. This would retain the infrastructure built by CMS and industry while removing the obligation to track thousands of relatively minor transactions under threat of penalties. Companies that want to stay squeaky clean in the eyes of the public could continue publishing everything. Others could devote their resources to more substantial payments that are less frequent and more revealing of potential conflicts of interest.

A narrower alternative would be to raise reporting thresholds substantially and focus mandatory disclosures on payments large enough to suggest a meaningful financial relationship. Either approach would reduce administrative burdens while preserving transparency where it matters most.

Industry codes of ethics can also complement government oversight. Organizations like AdvaMed and MedTech Europe maintain standards governing relationships between manufacturers and healthcare professionals. These aren't substitutes for enforcing anti-kickback laws, but they demonstrate that transparency and professional accountability need not depend exclusively on exhaustive federal reporting.

My second proposal is to significantly narrow how the Sunshine Act applies to medical device manufacturers. A company that makes insulin pumps for diabetics is simply not the same as a pharmaceutical company. What exactly is the concern when examining the financial relationship between an endocrinologist and companies like Tandem or MiniMed? Are they not supposed to have any? Is a diabetic patient at risk of receiving an unnecessary insulin pump simply because the physician has a relationship with its manufacturer?

Medical device companies can certainly create conflicts of interest, particularly involving implants and surgical equipment. But they also have legitimate reasons to work closely with physicians on training, product development, and technical support.

Rather than treating every interaction as equally suspect, we should consider reporting requirements that distinguish routine training and technical support from substantial financial arrangements that might influence clinical decisions. A surgeon receiving significant consulting payments from an implant manufacturer presents a different potential conflict than a physician attending a training session on a new insulin pump. Our disclosure requirements should reflect that distinction.

Most importantly, I’m not proposing that we get rid of Open Payments entirely. I understand its intent, and I believe the program has some value. The problem isn’t that it exists, but how it exists. Its scope is massive, its burden on companies is substantial, and its benefits are often unclear. If we think it’s inappropriate for industry and physicians to have any financial relationships, we risk strangling medical innovation. If we want to prevent kickbacks and corruption, separate laws already prohibit them. So if the goal is to shine some sunlight on these relationships, what exactly is it we want to see?

I spend my days searching for minor discrepancies in transactions that almost nobody will ever look at. I know firsthand how much work this system requires. What I don’t know is whether it’s making patients any safer. As someone who makes a living helping companies comply with these requirements, I'd welcome reform, even if it means less work for people like me.

Anonymous Excel Guy is a healthcare compliance worker who works for a large American medical device manufacturer writing under a pseudonym.

When Peter and Will joined, we envisioned it as a short-term experiment. It went so well that we extended it for a year, with a plan to evaluate at the end. That’s where we are now.

Both have made meaningful contributions. Peter helped sharpen our strategy, wrote our most popular essay ever, and pioneered our essay contests. Will improved operations, led research that grounded each sprint in facts, and made incredible charts that I honestly think are Pulitzer-worthy.

But the results haven’t justified our cost structure. We knew from the outset that our bar for success was high, and Peter and I agree this is the right moment for a change. We’re working together to make the transition as smooth as possible.

For now, I’ll lead our editorial work as we finish the healthcare sprint. To keep our focus during the transition, we’re canceling the essay contest, and we’ve already contacted those who submitted. Instead, over the next several weeks, we’ll publish a series of provocative contributed essays focused on improving American healthcare. I’m excited about them.

The plan is to spend October completing the transition and lining up essays, November closing out the healthcare sprint and reflecting on Boyd’s direction, and December setting priorities for 2027.

One advantage of being a small, independent institution is that we can experiment, learn, and adapt. I expect Boyd to look different next year, likely leaner and more decentralized.

The mission hasn’t changed. We’re still here to generate fresh ideas about America’s future. I believe that shifting our political culture toward solving real problems matters more than ever, and I’m looking forward to figuring out the best, most asymmetric way to do it.

If you care about this work and want to help shape what Boyd becomes, whether as a thought partner or donor, I’d love to hear from you. Feel free to email me directly at jeff@boydinstitute.org.

I’ll share more as our plans take shape.

Thanks for your support,
Jeff

On September 30, a Senate committee voted 12–11 to advance President Donald Trump’s nomination of Dr. Nicole Saphier to be the next Surgeon General of the United States. Her nomination now moves to the full Senate.

By the time the Senate gets around to voting on her nomination, the U.S. will have gone nearly two years without a confirmed Surgeon General.

Somehow, we survived. Americans continued seeing their doctors, hospitals treated patients, and state and local health departments carried on. The CDC tracked infectious diseases. The FDA regulated drugs and medical devices. HHS oversaw federal health programs. Business continued as usual.

Which raises an obvious question: If the country can go this long without a confirmed Surgeon General while its health agencies continue operating, why do we need one? And why do we still maintain the roughly 6,000-member uniformed service the Surgeon General oversees?

Last year, I joined Akiva Malamet, Bautista Vivanco, and Michael Cannon to examine that question in a Cato Institute policy analysis titled “Unnecessary Relics: The Surgeon General and the Public Health Service Commissioned Corps.” We concluded that both institutions have outlived their original purposes. Congress should abolish them and transfer their necessary functions to existing civilian agencies.

The larger reform concerns the 6,000-person Public Health Service Commissioned Corps. Its officers work throughout the federal government, often performing the same duties as civilian employees while receiving compensation and benefits modeled after the military. Abolishing the Surgeon General’s Office would eliminate a largely symbolic position. Civilianizing the Corps would tackle a more consequential problem: a costly, separate personnel system for work the federal government already employs civilians to do.

The institutions share a history. In 1798, Congress created hospitals to care for sick and disabled merchant seamen. Congress consolidated them into the Marine Hospital Service in 1871. John Maynard Woodworth organized its physicians into a mobile, uniformed workforce, which Congress formalized as a Commissioned Corps in 1889. As federal responsibilities expanded, the service became the Public Health Service, and its officers eventually provided care in federal prisons and on Native American reservations.

The Surgeon General once oversaw much of the federal government’s health apparatus. The Public Health Service Act of 1944 placed the office in charge of the Public Health Service and its constituent agencies. But reorganizations in the 1960s transferred that authority to other officials. The Johnson Administration abolished the Office of the Surgeon General in 1968, but it was reestablished in 1979 — without its former authority over federal health agencies — when Congress split the Department of Health, Education, and Welfare into the Departments of Health and Human Services and Education."

Today, the CDC, FDA, National Institutes of Health, and other agencies have their own leadership and responsibilities. The Surgeon General largely serves as an adviser and spokesperson. The Commissioned Corps survived these reorganizations, retaining its separate personnel system long after the government developed civilian alternatives.

The Surgeon General’s Uniformed Corps

In the 19th century, the federal government needed physicians it could deploy to Marine Hospitals across the country. Today, Corps officers work throughout the federal government as physicians, nurses, pharmacists, dentists, therapists, veterinarians, scientists, engineers, administrators, social workers, and other professionals. Many perform roles virtually indistinguishable from those of civilian federal employees.

The federal government doesn’t need a separate uniformed service to employ these individuals. Agencies that need physicians, nurses, pharmacists, scientists, engineers, or other health professionals can hire them directly. They can also contract with private-sector professionals when that is more practical. Federal agencies already use civilian employees and contractors to perform many of the same functions that Corps officers perform.

Supporters argue that the Corps can mobilize health professionals rapidly during emergencies. Emergency readiness matters, but a uniform does not guarantee a faster response. An HHS analysis cited in our paper found that Commissioned Corps officers took 24 hours or more to deploy during emergencies, compared to less than 12 hours for personnel from the civilian National Disaster Medical System.

Those figures do not prove that civilian responders will outperform Corps officers in every emergency. Missions and deployment requirements may differ. They do, however, challenge the assumption that a separate uniformed service is essential to an effective disaster response.

The historical cost comparisons also favor civilian staffing. A 2010 HHS analysis estimated that the average Commissioned Corps officer cost 15 percent more than a civilian employee performing equivalent work and estimated savings of approximately $217 million from using civilians instead. A 1996 Government Accountability Office analysis estimated that converting Corps positions to civilian roles could reduce spending on those positions by 22 percent.

In our Cato paper, we estimated the Corps’ annual costs at approximately $1.3 billion in inflation-adjusted dollars, based on 2009 data. Not all of that represents potential savings: civilian replacements, transition costs, and retirement obligations would remain. Still, historical HHS and GAO analyses suggest meaningful savings are possible. Congress should commission an updated estimate of the net savings from civilianizing the Corps.

Part of the problem is that the Corps has a compensation system modeled after the military. Its officers receive benefits that include tax-free housing and subsistence allowances, TRICARE coverage, military shopping and recreation privileges, and eligibility for certain veterans’ benefits. Host agencies pay officers’ salaries and administrative expenses, while HHS covers health and retirement costs. That arrangement can make a Corps officer appear cheaper to the employing agency, even when the officer costs taxpayers more overall.

It also obscures accountability. A Corps officer belongs to the Commissioned Corps but works for another federal agency. In an emergency, that officer may be expected to deploy elsewhere unless the employing agency deems the officer essential to its mission.

There’s a simpler way: Congress could abolish the Commissioned Corps and let federal agencies determine which functions they need. The Bureau of Prisons could directly hire or contract with medical professionals to care for federal prisoners; the Indian Health Service could do the same. Agencies needing scientists, pharmacists, engineers, or other specialists could employ them directly. Emergency medical personnel could work through existing civilian disaster-response programs.

Current Corps officers whose services federal agencies need could continue working as civilian employees or contractors. Congress would need to protect accrued benefits, address retirement obligations, and maintain recruitment, emergency readiness, and continuity of care during the transition. These are practical problems to solve, not reasons to preserve a separate personnel system indefinitely.

We don’t need a 19th-century uniformed service to staff a 21st-century federal government. The office that oversees it deserves scrutiny as well.

From Administrator to “The Nation’s Doctor”

Today, HHS calls the Surgeon General “the nation’s doctor.” But the U.S. is home to more than 340 million people and hundreds of thousands of physicians. We don’t need — and shouldn’t want — a government-appointed doctor to tell all of us how to live.

As the Surgeon General’s administrative authority declined, the office increasingly became a bully pulpit. Surgeon Generals have weighed in on topics ranging from smoking, diet, exercise, and drug use to violent video games, pornography, guns, social media, loneliness, working conditions, housing, and parental stress.

Public health intervention has a strong justification when people impose risks on others without their consent, as with contagious diseases or pollution. Sanitation, vaccination, and infectious-disease control have saved countless lives. But not every personal health risk is a public health emergency, and medical expertise alone does not settle questions of individual liberty or public policy.

Autonomous adults should be free to decide how to live, even when their choices pose health risks. Physicians can advise them, and government agencies can provide reliable information. Neither role requires a national doctor whose prestige lends medical authority to political judgments.

Physicians and public health officials can explain the health consequences of different choices. But policies on guns, housing, labor, education, or social media also entail costs, competing interests, and individual liberty. Medical expertise can inform those judgments, but it does not settle them. A presidential appointee’s title does not confer special expertise across all of these fields.

The prestige of “the nation’s doctor” can blur the line between scientific evidence and an administration’s policy preferences. When government health officials enter contentious political debates, Americans who disagree may become less willing to trust their advice. That can make it harder for those officials to convey credible information during a public health emergency.

The government already employs specialists in infectious disease, toxicology, epidemiology, and other fields. These experts can communicate with the public about matters within their expertise. We do not need a separate, presidentially appointed office to speak on their behalf.

Retire the Institution, Preserve the Work

The next Surgeon General nomination offers Congress an opportunity to examine the institutions behind the title. Why do federal agencies need a separate uniformed service to employ medical and scientific professionals they can already hire as civilians?

The prolonged vacancy has not prevented federal health agencies from doing their work. The more consequential question is whether the Corps’ separate personnel system improves their performance enough to justify its complexity and expense.

Congress should commission an updated assessment of the Corps’ costs, deployment capabilities, and recruitment needs, then establish a planned transition to civilian staffing where those functions can be performed as effectively. The goal should be to preserve essential services while eliminating unnecessary institutional overhead. That means protecting accrued benefits, maintaining emergency readiness, and ensuring continuity of care.

The United States needs physicians to care for federal prisoners, medical professionals to serve Native American communities, and skilled responders to handle public health emergencies. The federal government needs public health professionals, not a separate uniformed bureaucracy to employ them.

Jeffrey A. Singer, MD, practices general surgery in Phoenix and is a senior fellow at the Reason Foundation. His latest book is Your Body, Your Health Care (Cato Institute, 2025)

Physicians form a “sovereign” profession, meant to direct patients rather than bargain with their willingness to pay. Ex ante price opacity protects physician sovereignty, and health “insurance” performs not only the proper insurance function of protecting consumers from rare catastrophic risks, but insulates consumers even from routine costs, so that price opacity is tolerable. Of course, all too often, that cost insulation proves illusory ex post, and patients get surprise bills, sometimes ruinous. The No Surprises Act, which took effect in 2022, curbs only some of them.

Physicians’ and other suppliers’ self-interest creates incentives toward overtreatment, high prices, and restriction of supply and competition. Checks and balances have emerged, but the trend is not towards consumer sovereignty. “Insurers,” pharmacies, drugmakers, and others hire lobbyists (the health sector spent about $752 million on federal lobbying in 2024 alone) in a tug-of-war for a $5.3 trillion prize, equal to 18% of GDP in 2024, and these fights shape the industry far more than consumer choice does. It’s the lack of price transparency and market competition that let the lobbyists dictate outcomes.

A bit of economics jargon is helpful in naming the result. An “economic rent” is the extra income thrown off by a position artificially protected from competition by regulatory barriers. “Rent-seeking” is the lobbying and positioning done to capture and retain rents. The best description of American healthcare is thus neither capitalism nor socialism but rent-seeking paternalism. The system curtails consumer agency and competition through regulation and obfuscation, in favor of elite guidance and professional sovereignty, worked out through a mix of adversarial advocacy and interest group bargaining that doesn’t acknowledge self-interest but speaks the language of patient welfare.

To be sure, there’s a case for paternalism in many sectors, and it may be particularly strong in healthcare. It can even be life-saving, as we’ll see. But the comprehensive regulatory machinery to which it has given rise — occupational licensing, professional norms, public subsidies, price opacity, and third-party payment — leaves little room for consumer sovereignty or the capitalist competition that innovates for quality and cheapness to meet consumer demand. Other industries have rent-seeking and paternalism, too, but in no other industry except possibly higher education have price transparency and consumer sovereignty been excluded from the central work of resource allocation to such an extent as in healthcare, so that regulatory control and price opacity eviscerate the basic principles of capitalism. We can do better.

Subscribe now

The Burden of Paternalism

Paternalism is coercion meant to protect you not from others but from yourself, and it saturates medicine as it does almost no other corner of adult life. Paternalistic coercion is frequently applied not directly to the object of paternalistic concern, but to someone who wants to do business with them. Other people and companies are forcibly prevented from making you offers that the government paternalistically thinks you shouldn’t accept. In healthcare, paternalism spans a spectrum from hard statute at one end to bedside custom at the other. Law decides paternalistically who may treat you and what you may buy. At the level of the office visit, the doctor paternalistically directs and expects patient compliance. People are expected to follow “doctor’s orders.”

Thus, for example, in more than twenty states, nurse practitioners still face limits on independent practice or prescribing, sometimes through required physician agreements. The cost of that ultimately lands on patients, deterring care. Empirically, for routine visits, refills, and tests, nurse-led and physician-led care yield comparable outcomes. Nurse practitioners earn about half what physicians do, and their care costs less. And they go where physicians won’t — rural areas, the uninsured, the Medicaid-covered — for whom the realistic alternative is not a physician but no one at all. Where nurse practitioners are allowed to practice on their own, checkups rise and emergency-room visits fall. In state legislatures, physicians’ and nurse practitioners’ rival lobbyists each argue for scope of practice rules that favor more market share for themselves. But while the empirical studies matter, the more fundamental question is why the government should be dictating who should be allowed to practice what kind of medicine at all.

Again, dental therapists — midlevel providers who fill cavities safely and cheaply across the developed world — remain unavailable in most states. You might prefer a dentist to do your diagnosing, but that should be your call, not the state’s. Why is the argument conducted in state legislatures, rather than in public advertisements and consumer decisions? Why isn’t medical scope of practice determined by market competition and consumer choice, instead of lobbying and legal coercion?

At the technological cutting edge, these rules get in the way of medical use of AI. You cannot simply hire an autonomous AI doctor to diagnose and treat you. You can ask a chatbot for advice, but it may refuse anything individualized. Startups are building systems that do a primary-care physician’s cognitive work — taking a history, managing a chronic disease, renewing a prescription. In January 2026 Utah became the first state to let an AI system legally prescribe routine refills under a regulatory sandbox, but the larger question is why law should be deciding how consumers use AI for health. Chatbots are far more cheap and convenient than a doctor’s appointment, and would be used in many situations where a doctor’s assistance would not be sought. Many people are getting savvy to the peculiar mix of brilliance and fallibility in AI tools, and could navigate the epistemic nuances well. It’s the sort of thing the market should be left to discover. Instead, under a healthcare regime of rent-seeking paternalism, the awkwardness of whether an AI doctor is a device the FDA approves or a worker that states license is hard to surmount to get to yes.

It’s hard even to get your own medical records to give Dr. Chatbot the context it needs. A quiet paternalism governs medical privacy. Financial data provides an instructive contrast. Lenders and credit bureaus trade repayment data freely, and the payoff for consumers is, first, a portable reputation that puts a loan within minutes’ reach, and second, that you can download a credit report for yourself. While patient data is supposed to be shareable among medical providers, interoperability problems get in the way of that, and it’s not clearly in anyone’s interest to solve them, because the difficulty of moving your data to the next provider is an advantage for your current provider in retaining your business. By contrast, insurers would have ample incentive to compile the data if they were allowed to seek out and compete for the most favorable insurance customers. Instead, paternalism and third-party payment create a universal fog.

Paternalistic laws also limit what you may buy. Many common medicines, like an albuterol rescue inhaler or a blood-pressure pill, require a prescription, even if you’ve taken them for decades. Continuing a medication you know that you need can hinge on a clinician’s renewal, sometimes via another paid visit. Again and again, a service you might reasonably want is withheld, ostensibly to protect you, but often in ways that hurt your access or increase your cost, while protecting the incomes of incumbents.

The standard justification for healthcare exceptionalism traces to Kenneth Arrow’s 1963 paper “Uncertainty and the Welfare Economics of Medical Care.” Patients are often sick, frightened, and poorly informed, unable to judge whether a diagnosis is right. This information asymmetry, Arrow argued, drives medicine to grow its own non-market institutions, and Arrow’s argument tracks the kind of reasoning and intuition that continues to support paternalism. The patient doesn’t know, so the doctor should decide. Yet markets routinely handle “credence goods” whose quality the seller grasps far better than the buyer: auto repair, tax accounting, litigation, real estate. Darby and Karni, who coined the term, observed that when the same expert both diagnoses a problem and sells the fix, the customer may never learn, even afterward, whether the work was needed, so a temptation to oversell is built into the trade; yet they found that reputation, repeat business, and competition among sellers hold such fraud to a tolerable equilibrium level rather than letting it run unchecked.

Surveying the literature a generation later, Dulleck and Kerschbamer sorted the dangers into overtreatment, undertreatment, and overcharging, and showed that when the work performed can be verified, or the seller can be held liable for failing to fix the problem, market forces go a long way toward disciplining all three, without the buyer needing to become an expert. Around each such market, reputation, reviews, competition, brokers and advisors evolve to help people buy what they cannot fully judge. The alternative to paternalism is not to leave patients unguided, but to let the market find better ways to guide them. In other credence goods industries, there are licensing and cartel arrangements to some extent, but none disempower consumer sovereignty to the extent of healthcare.

At a basic level, much of healthcare paternalism consists in causing people to get medical treatments for which doctors get paid, when people would rather just have the cash. The RAND Health Insurance Experiment that ran from 1974 to 1982 showed how this works1. It randomly assigned families to plans that made them bear more or less of their own costs. Those who paid more used less care and ended up, on average, no worse in health, which as a vindication of cost-sharing seems almost too good to be true. What they cut was poorly aligned with what the experts consider necessary, and among the sickest and poorest, the skipped care did show up as worse blood pressure and failing eyesight. People make mistakes. But they also learn, and that slow social learning is, in the long run, wiser than the verdicts of experts. In the RAND experiment, patients had to improvise without the advisory systems that would develop if healthcare went full-spectrum capitalist.

The trouble with paternalism is not only that doctors’ and patients’ interests don’t fully align, but also that in some ways, the patients know more. The patient knows less than his doctor about anatomy, but he knows his symptoms from the inside, and he knows his habits, priorities, and above all his finances as no physician can. “Doctor knows best” is too simple, even clinically, and especially when willingness to pay comes into play. That is the case for consumer sovereignty against the sovereign profession.

Consumer sovereignty need not mean pure caveat emptor, which would be a bit implausible for medicines. But a rule can soften caveat emptor without being paternalistic, so long as it lets informed adults knowingly accept a risk. This is an off-ramp that healthcare law largely refuses to take. It grants no general right to gamble on a drug that the FDA hasn’t approved, or that a doctor won’t prescribe. The federal Right to Try Act of 2018 is the exception that proves the rule: it opens a narrow path to experimental drugs only for patients with life-threatening conditions who have exhausted approved treatments and cannot enroll in a clinical trial2. Sometimes this operates through lawsuits rather than through statutes, and de facto paternalism is the shadow of Americans’ unfortunate litigiousness. The Lyme vaccine can serve as a parable here. Lyme strikes hundreds of thousands of Americans annually. LYMErix, a vaccine, was approved in 1998 and never shown to be unsafe. In its pivotal trial, apart from passing soreness and flu-like symptoms after the shot, vaccinated volunteers suffered no more serious or lasting adverse events than those given a placebo. After approval, a federal review of the adverse-event reports filed by doctors and patients found no pattern suggesting the vaccine was causing harm3. And even if there were risks, some who love walking the New England woods might gladly accept a vaccine’s side effects to de-risk for the disease it prevents. LYMErix was withdrawn in 2002 after sales collapsed amid public safety fears and litigation, despite the lack of adverse safety evidence. There’s no way for consumers to opt out of the false protection provided by other people’s litigiousness. A quarter century passed before a successor even reached regulators: only in 2026 did Pfizer and Valneva report that a new vaccine was more than 70% effective in a large trial and begin seeking approval in Europe. All that time, lovers of walking New England’s woods could not get protected against Lyme disease from a tick bite, though oddly enough, their dogs could.

If LYMErix shows how safetyism can make us less safe, the takeaway is not that the nanny state is always crying wolf. Freedom can be tragic. We let people wreck their lives by drink, promiscuity, and addiction to porn or video games. We protect the right to bear arms, and sometimes that enables suicides that wouldn’t have happened if a person didn’t have access to a gun at the wrong moment. We allow dangerous recreations like ATV riding and parachuting from cliffs. And the opioid epidemic that kills tens of thousands a year4 is another case where many fatalities were downstream of something the government allowed (OxyContin). Opioid deaths flowed from people’s own choices, but taking a few options off the table may pass the cost-benefit test. Yet alcohol and tobacco take far more lives than opioids, and we don’t ban smoking or (since 1933) beer. Most of the risks that paternalism forecloses are far milder than opioids or tobacco. Why not let people choose their own level of medical risk, as we already let them drive cars, smoke, and drink?

Our passivity in the face of medical paternalism may be a spillover from the ritual of the office visit. The doctor hands you no menu and only grudgingly volunteers the cheaper Plan B. Physician paternalism is tolerable when the ask is a daily pill or a single operation. “Doctor’s orders” carry that much authority easily enough. But people can’t be expected to tamely comply when a recommended course of treatment costs a serious slice of a patient’s annual, or lifetime, income. Then the conversation has to turn to money and trade-offs, and maybe to family, or even to the will to live and bequest motives. Realtors talk about those things. Life-insurance salesmen can talk about them. Physicians would rather not. It is not the bearing of a sovereign profession, and not what they trained for. That is probably the deep reason why American medicine is so extravagantly opaque about price. It protects clinical paternalism. And the institution built to keep the exam room insulated from financial arithmetic, so that the doctor can keep prescribing and the patient keep complying, is the one we so misleadingly call “insurance.”

Price Opacity and Pseudo-Insurance

What we call “dental insurance” isn’t insurance at all. Insurance smooths consumption by covering large costs. Dental plans cover small ones and quit precisely when a crown or root canal makes them large, capping out at a thousand or a few thousand dollars a year. So why does it exist? One reason is the tax treatment of employer health benefits. Pay a worker in dental coverage rather than wages, and the compensation escapes the federal income tax.

Major medical coverage, by contrast, does provide real insurance, but that isn’t its main job. Insurance, properly understood, is a service that smooths consumption between the lucky and unlucky scenarios, trading a little expected wealth for higher expected “utility,” as shown in Figure 1:

The chart shows how lifetime utility rises with wealth but flattens as it grows, since each extra dollar adds a little less happiness than the last. Happiness isn’t measurable, so this is plausible hedonic psychology rather than proof. But it’s the best general explanation for why people purchase insurance. Economists call it diminishing marginal utility. Someone facing a risky future — comfortable wealth if all goes well, along with potential setbacks or disasters — is worse off than they would be if they could remove the uncertainty and just get average or “expected” value. They’ll be willing to sacrifice not only the upside but some expected (average) wealth for more safety. The least sure sum that a person would accept in place of the gamble is his “certainty equivalent.” Because that certainty equivalent is less than average wealth, insurers can make money taking the premiums and paying the occasional claims.

The curvature of utility is needed to provide some margin, because insurance faces not only general administrative overhead, but also the classic insurance-specific problems of adverse selection and moral hazard. Adverse selection occurs when people know things about their risk profiles that insurers don’t. People whose unobservable risks are higher will opt into insurance, making the insurer’s customer pool more expensive to serve than its observable traits would suggest. Moral hazard occurs when people’s risk profiles depend to some extent on their own behavior, and the fact that they have insurance makes them behave in riskier ways. In the worst case, this involves actual fraud, e.g., a person might deliberately crash a car that’s insured for more than it’s worth. More often, people get careless or live a bit more dangerously, e.g., you go skiing because if you break your bones, your insurance company will foot the medical bill, not you. Where adverse selection and/or moral hazard run strong enough, a risk becomes all but uninsurable, which is why private unemployment insurance, for example, barely exists. Insurance works best for risks that are statistically predictable on readily observable traits, which don’t depend too much on consumer behavior; and where the risk is large compared to the insured’s wealth or income. Below a certain level, smoothing a small bump isn’t worth the premium, though insurance may exist for other reasons, such as legally mandated accident liability insurance for cars.

“Health insurance” doesn’t operate like that. It typically helps you pay for every visit with a small copay. It does provide real protection against catastrophe, but it’s mixed up with insulation against routine costs, in exchange for premiums that many families could not afford unaided by employers or the government. The average employer-sponsored family plan cost $26,993 in 2025, or about $2,250 a month, more than the median mortgaged homeowner’s entire monthly housing bill, with mortgage, property taxes, insurance, and utilities included5. “High deductible” plans are closer to real insurance, but even they kick in well before medical bills put a major dent in lifetime wealth.

Insurance as cost insulation instead of risk smoothing adds two new layers of social cost. First, patients lose any reason to shop or bargain for cheaper healthcare. Second, providers gain a perfect alibi for telling them nothing about cost.

Healthcare price opacity is extreme, and “secret shopper” studies have caught it for over a decade. In one, researchers posing as a self-paying patient called 122 hospitals (two in every state, plus the nation’s 20 top-ranked orthopedic hospitals) and asked for the complete price of an elective hip replacement. After as many as five calls, more than a third still could not give a complete price, and the estimates that did come back ranged from about $11,000 to $126,000 for the same procedure, more than a tenfold spread6. A decade later, after a federal rule began requiring hospitals to post cash prices online, another team checked whether 60 hospitals’ posted prices for two routine services, a vaginal childbirth and a brain MRI, matched what the same hospitals quoted secret shoppers by phone. The two numbers were poorly correlated, and the phone quote was not reliably the lower one: several hospitals posted childbirth prices above $20,000 online but quoted under $10,000 by phone, while two quoted more than $5,000 for an MRI they listed online at about $2,000.

Pricing a given patient’s care is hard, but other industries also have to deal with complex, contingent pricing, yet they manage to settle terms in advance. Medicine uniquely acts as if it has the right to perform a service and then bill whatever it likes, with no meaningful prior consent, even for charges that swing by multiples for the same procedure, and are often ruinous for patients who didn’t know ex ante what they would owe.

The perceived need for health insurance is driven not just by the risk of genuine catastrophe, but by the fact that if you go to the doctor, you face mysterious liabilities, and you need a big company in your corner to shield you from that. You don’t know if the doctor is billing you fairly, and you can’t fight back. Hopefully the insurance company does, and can. You buy insurance to keep from being held hostage to whatever bill the hospital sends. Health “insurance” became even more misnamed when the Affordable Care Act barred insurers from denying coverage or raising prices for pre-existing conditions. Pricing for risk is the heart of insurance.

Banning insurers from pricing properly for risk made it part of a larger system of transfers from the healthy to the sick. This involves the familiar trade-off between equity and efficiency, with inefficiency resulting from (a) reduction of the incentive for people to adopt healthy lifestyles to reduce medical costs, and (b) when people do get sick, reduction of the incentive to wait and see, self-medicate, or shop around for the cheapest care. But the other perversity here is that some of the sick are rich and some of the healthy are poor, so the transfers can run backwards. This happens a lot, because the young tend to be healthy with few financial assets, whereas older people are often affluent, but with more health problems. So a lot of transfers from the healthy to the sick are inevitably transfers from the economically struggling to the relatively well-off.

A lot of what’s wrong with healthcare is plausibly downstream of price opacity. Excessive occupational licensing and restrictive scope of practice rules, for example, are a distinct issue, but the benefits of looser scope of practice are masked by the fact that it usually isn’t a patient paying, even when they could easily afford to do so.

Diagnosing the Blight

Rent-seeking and paternalism are common across the economy, and they often travel together. In Bottleneckers, Mellor and Carpenter of the Institute for Justice catalogue, with grim hilarity, the absurd arguments by which lobbyists for casket-sellers, hair-braiders, and food-truck vendors dress protectionism as public safety. But in other industries, the bottleneckers have to argue against a default of consumer sovereignty which makes sense to people and can be articulately defended. Competition is the rule, paternalistic rents are the exception. In healthcare, consumer sovereignty is orphaned and marginal, surviving only in pockets like LASIK, an elective and almost never insured procedure bought and sold in an ordinary market, whose average price has barely risen in a quarter century, or the Surgery Center of Oklahoma, a rare medical provider that posts all-inclusive cash prices online. But such pockets of consumer sovereignty add up to a rounding error in a multi-trillion-dollar industry that rarely answers to consumers’ wallets. Opacity and paternalism are so entrenched that we can scarcely imagine what healthcare capitalism would even look like. When healthcare policy is debated in state legislatures, rent-seeking paternalism is the only game in town. Legislators who champion free market policy in other areas don’t know where to begin, and might not want to get on the wrong side of the lobbyists anyway.

Healthcare does have its peculiarities that make capitalism hard to implement, like the emergency that allows no shopping, the unconscious patient, and a fair amount of complex but urgent decision-making. But they are bounded, not pervasive, and the parallels to other credence-good markets give good reason to expect that opening the industry to consumer sovereignty would yield large and lasting gains. There would be losers, too. Some now sit on the receiving end of the status quo’s hidden redistribution. But what gives overhaul the most urgency is simply that the present system is unjust, because it cannot help sending people bills they never agreed to, because it could not tell them the prices in advance. Those bills are often very burdensome, but even when affordable, they lack the legitimacy of real consent. And the whole economy is increasingly hostage to healthcare dysfunction as Medicare grinds along its budget-busting path, with its hospital insurance trust fund projected to run dry in 2033. It is high time the American economy stopped following doctor’s orders.

Leave a comment

The economic challenge for our healthcare system is to align costs and benefits. The information that policy makers would need in order to do that is difficult to obtain.

Many Americans complain about our healthcare system. We experience health insurance as confusing and bureaucratic, with ever-rising premiums. The bills that we receive from healthcare providers seem ridiculous, although with insurance the prices are heavily discounted and the consumer actually pays little or nothing.

Healthcare spending in America has increased dramatically over the past 50 years. On a per capita basis, our spending is now much higher than that of any other nation. Are the benefits commensurate?

One way to measure benefits is to use the reduction in treatable mortality. Whereas overall longevity statistics include deaths from automobile accidents and homicides, estimates of treatable mortality more closely measure the effects of medical care. In 2022, the most recent data available from Health at a Glance, treatable mortality in the U.S. was 95 per 100,000, vs. 77 per 100,000 in the OECD average. On this measure America is worse in spite of higher spending.

Another way to try to assess whether higher spending on healthcare reduces mortality is to look at sub-populations in the United States with similar characteristics and different levels of spending. Many studies show little or no effect of higher spending in lowering mortality. This suggests that there is a problem with cost-effectiveness in our healthcare system.

Note, however, that not every study fails to show a reduction in mortality from more spending on healthcare. Note also that mortality is not the only healthcare outcome that we care about. Quality of life matters as well. Procedures such as knee replacements or in vitro fertilization are expensive, but patients value them highly.

Our healthcare institutions face trade-offs. Consider the regulation of who can practice medicine. Medical licensing systems and practice regulations govern entry into medical professions. The malpractice insurance system is one mechanism for forcing exit from medical practice. These systems are costly to operate, but we cannot do without some costly system for regulating entry and exit.

Consumers and the third parties that pay for healthcare services have to make two types of decisions: which medical procedures to undertake; and how much to pay for those procedures. When healthcare is paid for by insurance companies or government, that means that third parties become involved in making these decisions.

In order to hold down costs, health insurance companies deny coverage for some procedures. They also negotiate with doctors and hospitals for lower prices. But consumers resent being denied coverage for a procedure. And hospitals, doctors, and other healthcare providers resent what they see as inadequate compensation. Government appears to be in a better position than private insurance companies to tolerate such resentment. Consumers in countries where government pays for healthcare find that some services, such as colonoscopies to screen for colon cancer, are not available. Doctor salaries appear to be lower in other countries than in the United States, although international comparisons are difficult given differences in hours worked and degree of specialization.

Concerning the question of which medical procedures are covered, consider this thought experiment: Imagine that you were offered a health insurance policy that only covered medical procedures that were widely available 50 years ago. The premiums on such a policy would be much more affordable than current health insurance premiums. But you could not be reimbursed for these expensive and now-common services:

• Knee replacement

• Liver transplant, lung transplant, or reliable heart transplant

• MRI

• Various heart procedures, including balloon angioplasty or stents

• In vitro fertilization

• Cell therapies, gene therapies, other advanced cancer treatments

• Modern drugs, including biologics, proton pump inhibitors, GLP-1 agonists

These innovative treatments and procedures account for a lot of the increase in healthcare spending over the last fifty years. Health insurance premiums have skyrocketed as a result.

An increase in medical specialization over the past fifty years also has contributed to rising healthcare expenditures. Today, people are used to seeing neurologists, dermatologists, endocrinologists, as well as other specialists. For treatment, the general practitioner often serves in the healthcare system as a point of entry rather than the patient’s last stop.

The trade-off between overhead and fraud

At least some of the overhead in the health insurance industry is needed in order to prevent fraud. When a provider asks for reimbursement, the insurance company must screen for fraud. Was the procedure actually performed? The insurance company must spend resources to find out. If the industry were to reduce the resources used in deciding to approve and reimburse medical procedures, that would make it easier for providers to commit fraud.

Similarly, an insurance provider has to determine whether a procedure is necessary in order to know whether to provide reimbursement. When a procedure is covered by a policy and is necessary for the particular patient, it would be a mistake to deny coverage. But it also would be a mistake to provide coverage for a procedure that is not necessary for this particular patient. The insurance provider has to spend resources to screen out unnecessary procedures.

Procedures that are not cost-effective

But many medical procedures do not fall clearly into the binary of “necessary” or “unnecessary.” Instead, they fall in a gray area. Precautionary tests that may or may not reveal anything are one example. Treatments that have a low probability of working are another.

Colonoscopies to screen for colon cancer are an example. There is no doubt that these save lives relative to the alternatives, such as screening via fecal occult blood samples. But other countries rely on the less costly screening methods. They seem to have determined that the cost per life saved of routine colonoscopies is unacceptably high.

I believe that American culture, habits, and institutional arrangements lead to over-use of medical procedures that have high costs and low benefits. I doubt that this will be easy to change. What many doctors regard as futile care near the end of life is something that relatives often insist upon. And anything that smacks of healthcare rationing produces outrage. In 2009, for example, the U.S. Preventive Services Task Force recommended several changes that would have reduced the amount of breast cancer screening done in this country. The recommendations were denounced and rejected.

Pay for quality?

The fact that healthcare providers are compensated for performing procedures leads economists to suspect that providers are biased toward recommending more procedures and costlier procedures than are optimal. But any alternative method for compensating doctors poses problems.

If doctors are paid a straight salary, then they have an incentive to have only a small panel of patients. If they are paid by the number of patients on their panel, then they have an incentive to accept only the patients who require the least effort to see. If they are paid on the basis of the health outcomes of their patients, then they have an incentive to keep the sickest patients out of their panel.

One approach, championed by health economist David Cutler and others, is to pay physicians on the basis of how well they adhere to standards of care. For example, suppose that there are standards that say when a heart stress test is indicated or not. Doctors who either fail to give tests when indicated or give tests when not indicated will be compensated less than doctors who adhere to the standards.

This “pay for quality” approach imposes heavy information requirements on the third party setting the compensation rules. In effect, the standards based on generic cases are being used to over-ride the doctor’s judgment about the particular patient as seen in the office.

Pay less?

Another idea is to pay for procedures, as we do now, but pay less. The implicit assumption in this approach is that providers earn what economists call rents, meaning incomes in excess of what they require in order to supply their services. In the words of the title of a famous paper, “It’s the prices, stupid.”

The laws of supply and demand say that if the price is artificially high, then that should produce a surplus. But we do not observe an excess of doctors and nurses.

There could be a surplus of people who wish to become doctors, nurses, or other healthcare providers. They are prevented from doing so by medical licensing requirements. In my home state of Maryland, for example, one needs to obtain a doctorate in order to enter the physical therapy profession. Economists have long questioned whether licensing and practice regulations truly protect consumers, as opposed to incumbent healthcare providers.

If we believe that healthcare providers are paid too much, then one policy prescription for cracking down on excess rents is to increase the negotiating power of the third party payers—insurance companies and government. Doing so requires, at least implicitly, estimates of the elasticities of supply of various forms of healthcare provision. If supply is relatively inelastic, then lower prices will not cause providers to exit. But in fact there are doctors who refuse to take Medicaid or Medicare or even private insurance. Supply is at least somewhat elastic.

Conclusion

From an economic perspective, a healthcare system is a set of institutions that processes information and provides incentives to consumers and healthcare providers. What we have seen is that there are trade-offs involved, such as the trade-off between keeping administrative costs low and policing fraud. And the information to make good decisions about what procedures to provide and at what price is far from perfect.

It appears that America’s healthcare system has unusually high costs relative to benefits. Broadly speaking, there are two ways to attack this. One way is to try to reduce the use of medical procedures that have high costs and low benefits. The other way is to try to bring down the prices charged by healthcare providers.

How can we reduce Americans’ use of medical procedures that have high costs and low benefits? If consumers had to pay for more of their medical expenses out of pocket, then they would purchase fewer medical services. But given consumers’ imperfect knowledge about the likely benefits of procedures, they might make choices to forego beneficial care as well as procedures that are not cost-effective.

The alternative is to have insurance providers, both private and government, make the decision to deny coverage for procedures that are deemed to have high costs and low benefits. But with imperfect information, third parties will make mistakes. Nearly every procedure offers at least some benefit, if only a slim probability of helping the patient to feel better. Insurance providers could be more rigorous about denying coverage for procedures that do not provide enough expected benefit to justify the cost. To do this well, third parties would need very good information about the average outcomes of various procedures. And even so, they would be in the position of over-riding the judgment of doctors concerning particular patients. More rigorous decisions would provoke more resentment from patients and their doctors.

For reducing prices, we have seen that one approach would be to increase supply in healthcare by loosening restrictions on licensing and medical practice. Another approach would be for insurance providers to negotiate for lower prices from providers. With either approach, a lack of knowledge about elasticities of supply will hamper the ability of policy makers to achieve desired outcomes.

We ask a lot of the institutions that make up our healthcare system. These institutions must govern who can supply various medical services. They must channel resources for innovation. They must encourage the use of cost-effective medical procedures while discouraging the use of procedures that are not cost-effective. They must provide adequate and appropriate compensation for healthcare providers, taking into account their incentives to take advantage of any system, including by committing fraud. They have to take into account political constraints, including the resentment that Americans feel toward any institution that is perceived as restricting insurance coverage of medical services.

In all of these activities, our capabilities are limited by a lack of available information. We do not know who is qualified to be a healthcare provider, so we rely on the licensing system to control entry and on the legal malpractice system to influence exit. We do not know what procedures are cost-effective on average, and we give doctors imperfect guidelines. Even if guidelines apply on average, we do not have the information that doctors obtain about particular cases. We cannot look over doctors’ shoulders to observe their behavior, so that we cannot be confident that whatever compensation system we are using is producing the right responses to the right incentives.

It would be easy to fix America’s healthcare system if there were a set of clearly identifiable villains. But the problems that we face are a reflection of imperfect information and trade-offs. There are ways to improve our institutions, but there is no miracle cure.

Subscribe now

Leave a comment

Subscribe now

What is ‘universal healthcare’?

The World Health Organization defines universal health coverage as follows:

“All people have access to the full range of quality health services they need, when and where they need them, without financial hardship.”

The general idea is that everyone can access and afford the care that they need, though ‘need’ is subject to interpretation. No country provides its citizens with unlimited, unrestricted, free healthcare. Every country has a mechanism for rationing care and paying for it. There are a few main models for doing this:

• Public funding, public provision (socialized medicine). The United Kingdom is the most obvious example. The government owns the National Health Service (NHS) and finances it with tax revenues. The NHS rations care through queues, subject to government determination of clinical priority.

• Public funding, private provision (single payer). Canada is the classic example of single payer. Healthcare providers are private, but the government sets prices and pays for medically necessary care using tax revenues. Like the NHS, Canada rations care through queues.

• Compulsory private insurance (managed competition). Switzerland and the Netherlands have an individual mandate for their residents to carry health insurance. Government regulations stipulate that insurers must offer coverage to everyone (guaranteed issue) and cannot charge extra for preexisting conditions (community rating). The government also subsidizes lower-income enrollees. These systems use various rationing mechanisms including deductibles, cost-sharing, restricted networks, prior authorizations, primary-care gatekeeping, etc. 

• Hybrid approach. Not every country with ‘universal healthcare’ fits cleanly into the above categories. France, for example, has a universal public insurance layer that includes patient cost-sharing. Most citizens purchase supplementary private insurance to complement the public layer.

Why doesn’t the U.S. have universal healthcare?

Technically, the U.S. has a framework for universal coverage thanks to the Affordable Care Act. That consists of Medicare for those 65+, employer-sponsored insurance (ESI) for workers, Medicaid for those below 138% of the federal poverty line (FPL) in the 41 states (plus DC) that expanded it, and subsidized Marketplace coverage for those between roughly 100% and 400% of FPL without access to ESI.

The ACA Marketplace was modeled after the managed competition systems in Europe, though the individual mandate was weakly enforced initially, and Congress ultimately zeroed out the penalty in 2019. A handful of states (CA, NJ, RI, MA, DC) kept the individual mandate at the state level, but even these states don’t have universal coverage. Part of the reason is a lack of a defaulting mechanism: in Europe, noncompliant citizens are fined and eventually defaulted into an insurance plan. The ACA did not include any such provision.

Complexity is another major hurdle: according to KFF, roughly half of all uninsured Americans qualify for either Medicaid or subsidized Marketplace coverage but are not enrolled. Further fueling this complexity is significant churn between programs: many people cycle between eligibility for Medicaid and subsidized ACA coverage (based on the 138% FPL cutoff). People also typically lose their employer-sponsored insurance when they lose their job, creating further churn.

Why is the system so fragmented?

To answer that, we have to go all the way back to World War II. The government imposed wage and price controls at the time, which meant that employers couldn’t raise wages to compete for employees. So they found a workaround: offering fringe benefits like health insurance. Congress later decided to make those fringe benefits tax-exempt, which created a major incentive for employees to receive part of their compensation in the form of benefits like health insurance. This quickly became standard practice at most employers.

A rational health insurance product would allow an individual to stick with the same plan for the long term, perhaps via multiyear contracts or change of health status insurance. The problem with tying health insurance to employment is that you lose your health insurance when you lose your job. This creates some major gaps in the market for health insurance, most notably for retirees and the unemployed. It also means that people who get sick while employed and then lose their job (and insurance) will struggle to find insurance in the individual market because of their preexisting condition.

These gaps are the basis for the fragmented system we see today. Medicare filled the gap for retirees. Medicaid filled the gap for the unemployed, underemployed, and disabled. The ACA filled the gap for those with preexisting conditions. Among peer countries, the U.S. is a major outlier in having such a convoluted, patchwork system. Why didn’t the U.S. just implement a single, universal system? You’d have to ask Lyndon Johnson.

Switching to a universal system today would be extraordinarily difficult, as the transition would involve enormous disruption to a deeply entrenched, highly path-dependent system that now accounts for nearly one fifth of GDP.

For all the system’s faults, most people are satisfied with their current coverage arrangement. We see this with Medicare for All polling: the phrase polls well, but support plummets once people learn they would lose their private insurance. Insurers and hospitals would also fight tooth and nail against any change that threatens their viability.

How does the U.S. system stack up internationally?

Ranking healthcare systems is difficult and highly subjective, as different systems optimize for different metrics. The most commonly used metric in everyday parlance–life expectancy–is arguably one of the worst measurements to use. As we discuss here, life expectancy is heavily influenced by things the healthcare system can’t directly control: car crashes, gun deaths, drug overdoses, and poor metabolic health. For a useful illustration, consider this Johns Hopkins study examining the gap in life expectancy between the U.S. (78.6) and the U.K. (81.3):

The category most reflective of actual healthcare system quality — cancer treatment — actually reduces the gap in life expectancy. Drugs, guns, and cars have little relation to the healthcare system, while cardiovascular disease is more a function of lifestyle choices. From Hopkins:

“The eight key risk factors and behaviors for the development of cardiovascular disease are suboptimal diet, sedentary lifestyle, tobacco use, overweight/obesity, hyperlipidemia (such as high cholesterol), elevated blood pressure (hypertension), diabetes, and insufficient sleep. Of these, evidence indicates that the U.S. has a lower-quality diet, less physical activity, more obesity, more hypertension (but also a higher rate of blood pressure control), greater intake of sodium (linked to hypertension), and more diabetes.”

All that is to say — we probably shouldn’t use life expectancy as our barometer for assessing the quality of a healthcare system. Better measures to look at are comparative access, treatment outcomes, and spending.

Access to Healthcare

All healthcare systems have either price or nonprice barriers to accessing care. Socialized and single payer systems have nonprice barriers; the biggest obstacle tends to be long queues. Managed competition systems can have price and/or nonprice barriers, depending on the system and the service in question. The U.S. tends to perform very well on wait times, but leads peer countries (by a lot) in reported incidence of foregone care due to price:

It should be noted that access to healthcare varies wildly in the U.S. depending on what kind of insurance you have. People on Medicaid actually have an experience more closely resembling Canada or the U.K.; they are largely shielded from price barriers, but often face very long wait times due to narrow doctor networks. On the other hand, a Medicare fee-for-service (FFS) patient with a generous Medigap supplemental plan can receive first-dollar coverage (no cost-sharing) and has access to almost every doctor in the country. Private insurance tends to be somewhere in the middle, though with considerable variation: a heavily subsidized ACA plan can look a lot like Medicaid, while a generous PPO plan can resemble Medicare FFS with Medigap.

The U.S. is also unique in having a considerable percentage of the population (~8%) with no health insurance. Uninsured patients face significant price barriers to obtaining outpatient care–and while hospitals will treat emergent conditions regardless of ability to pay, the subsequent bill can be financially devastating for the uninsured.

Treatment Outcomes

Cancer treatment is a notable bright spot in the U.S. healthcare system:

For acute-care outcomes, the U.S. is among the best of its peers at surviving a stroke, and about average after a heart attack:

However, the U.S. leads its rich-country peers in ‘treatable’ and ‘preventable’ deaths:

Importantly, poor population health affects both incidence and treatment outcomes for most diseases. It is difficult to isolate whether an individual treatable/preventable death was caused by poor health, poor healthcare access, or both.

Healthcare Spending

The U.S. has no peer when it comes to healthcare spending. We spend ~18% of our GDP on healthcare, roughly double the OECD average of ~9%. This astronomical spending is largely driven by higher prices, not by higher utilization. In other words: our dollars purchase fewer healthcare services than they would in other countries.

One rather startling way to visualize this is to compare public healthcare spending by country. Many healthcare pundits point to the government-run systems of Canada and Europe to make the case that the government can procure healthcare services more efficiently than the private market. Yet in the United States, despite public healthcare programs covering only ~40% of the population, public healthcare spending already rivals that of many countries with universal systems!

For another useful illustration, a JAMA study compared healthcare spending by age cohort relative to other countries. They found that all age groups vastly outspend their international peers–including the Medicare population.

Conclusion: There are no easy solutions in healthcare, only tradeoffs.

No single decision created the U.S. healthcare system as it exists today; rather, it developed through decades of path dependency. For all its faults, the U.S. system performs relatively well on timely access to care and certain treatment outcomes. The biggest challenge is runaway spending, which both the government and the private sector struggle to control.

Leave a comment

Let’s kick it off with life expectancy data.

Life expectancy at birth is mostly a statistic about children. In 1900 about 1/10th of American babies died before their first birthdays. By 1976 it was one out of every 65. Most of these gains occurred without dramatic increases in spending on healthcare. But since 1976, when life expectancy at birth was already 73, health spending as a share of GDP has roughly doubled alongside squeezing out life expectancy gains of only 6 years.

The United States reached a life expectancy of 66 in 1946, at an income of about $9,200 per head in 1990 dollars. China climbed to the same level on only about $1,400 per head, across a period that included the Cultural Revolution and a catastrophic self imposed famine. Still, the interventions such as sanitation campaigns, mass immunization, and the barefoot-doctor program generated a huge lift. In effect, China got healthy before it got rich.

Now, if you look at US age-specific death rates vs. Europe, the peak sits where almost nobody dies of disease — in the range of where spending is NOT concentrated — and collapses into the disease-ridden years of a person’s life.

So what’s actually driving the discrepancy?

A nontrivial factor is the guns, drugs, and cars — none of which are “healthcare” deaths per se. But together they are why American life expectancy fell three years in a row in the mid-2010s, the first sustained peacetime fall since the 1918 flu.

Interestingly, significant racial disparities exist in life expectancy. Depending on the sub-population, life expectancies in the US vary by a lot, even under the same health system. How? What separates the top from the bottom, in reality, is income, place — and importantly, deaths from guns, drugs, and cars, which fall hardest on the last three groups.

As for the older end of the spectrum, age-adjusted deaths from heart disease have fallen from 412 per 100,000 in 1980 to 161 in 2019, and cancer mortality, flat through 1990, has since fallen by a third. The decline in heart disease is especially important and revealing because it has coincided with a dramatic increase in obesity.

So, heart-disease deaths collapsed as one of its main risk factors (obesity) trended in the exact opposite direction — clear evidence of medical advancements.

Nevertheless, we still spend far more money on healthcare for the people who are disease-prone than on the ones who aren’t…

Personal health care spending per American in 2020 on people over 85 was 8.5x that of a child. An aging country that is no longer dying from it’s main disease naturally spends more, which explains why about a quarter of Medicare’s money goes to people in the last year of their lives (and this share has barely moved since the 1970s).

Zooming out, rich countries just spend more on healthcare, and above-trend US spending amounts to a few thousand dollars per year per capita.

Americans see a doctor less often than people in other major countries, but that says very little about what a health system costs. Furthermore the delta also isn’t explained by “administrative bloat.”

The viral chart showing a massive divergence — administrators up +3,000% since 1970 against doctors up +150% — is highly misleading. When counting actual dollars spent, admin-related spending is basically flat (or at least rangebound) since 2003.

The US is rich in machines and drugs, but poor in doctors and hospital beds — the read here being that the cost drivers have more to do with the fact that whatever a hospital can simply buy, America has a lot more of. And whatever a license or statute controls, it has much less of. Scanners and drugs can just be purchased fairly simply.

But beds are rationed by state laws, and doctors have also been “under-rationed” in a certain sense...

There was basically never a doctor shortage until after Congress cut funding. Throughout the 1990s, the Council on Graduate Medical Education warned Congress of a looming physician surplus. A couple highlights:

1994: “The reduced physician growth rate and increased proportion of generalist physicians attained by implementing COGME’s goals can be expected to substantially reduce health care expenditures.”

1996: “COGME recognizes that the nation’s most significant workforce problem is an increasing surplus of physicians, primarily of specialists.”

Congress took their advice and capped the number of federally funded residency slots in 1997. Whoops.

On the pharmaceutical drugs side of things, however, it’s certainly not much of a supply-side story. Americans have the most access to the latest and most expensive drugs, and they pay up for them. It is the very small sliver of expensive, non-generic drugs that drive the overall spending in America.

Zooming back out once more, here’s how much we’re paying, and who is actually paying for it…

Government healthcare financing alone accounts for nearly half of American healthcare spending — 8.5% of GDP — while private spending adds another 9.5%. Meanwhile, the average OECD country spends 9.3% of its GDP on healthcare in total.

American public health spending alone exceeds what Britain, Canada, France, or Japan spend on health in total. So, it’s pretty evident the US has an efficiency problem, not a generosity problem.

Healthcare took two cents of every federal dollar in 1962, but now takes thirty. And the “generosity” surrounding Medicare in particular has contributed heavily to the cost explosion.

Medicare has always relied on a mix of funding streams. Part A (hospital insurance) is funded by payroll taxes, while Part B (outpatient and physician coverage) and Part D (prescription drug coverage) are covered by a mix of beneficiary premiums and transfers from general tax revenues.

As Medicare’s costs have increased, general revenues (i.e., deficit spending) have picked up most of the tab — importantly, crowding out other budget priorities, as it now comprises a significant part of what is making the federal interest bill explode too.

Essentially, the lion’s share of growth in Medicare spending has not come from changing demographics. The big bar (61% of growth) is not so much price increases as it is medicine that barely existed in 1967 — dialysis, bypass, stents, joint replacements, statins, biologics, etc. In other words, the ballooning costs come from spending on new technologies in an environment where doctors and patients default to “yes,” and where price is an afterthought.

The share of the population receiving government-financed insurance has grown considerably since the 2008 recession. From 2008 to 2024, Medicaid grew from 13.4% of the population to 20.4%, while Medicare grew from 10.9% to 14.8% of the population. The individual market also became heavily subsidized post-ACA: 92% of Marketplace enrollees received a subsidy in 2024.

Today I’m excited to announce our next sprint, where we ask:

In particular we are interested in ideas that grapple with unleashing technology and markets to improve American healthcare.

The sprint will begin September 1st and will run through the end of the year.

In addition to the new sprint, we are also excited to talk about some of the changes coming to our format. Specifically, we will be shifting our content production, community, and essay contest with the intention of making them all more fun to participate in and better designed to elevate bold asymmetric solutions.

I’ll get into each of these changes in detail in the article, and if you have any questions please do not hesitate to reach out. But the spark notes version is this:

• Community involvement in writing: we’re opening paid article submissions to all of you, with essay contest winners first in line.

• Boyd Scouts: a Signal group chat for paid subscribers and contest winners, replacing the Discord.

• Essay contest: more prize categories — including a few we just think are funny.

Community involvement in writing

Concretely, this means it will no longer be Will and me producing most of the content. Instead, we hope to use our growing platform to help elevate the ideas and voices of you all, in particular those writers who have submitted to and won prizes in our essay contest.

We have already reached out to some of you with an invitation to participate and will be reaching out to more as we move along in the sprint. But this is not the only way you can participate. As a big believer in personal agency, I want to open the doors to all of you.

You’ll find a Google form HERE where you can submit any ideas you have for an article on the subject, and you’re always welcome to email me or Will directly if you’d rather chat first.

The goal here is not for this to replace our essay contest, and if you want to write for us in the future, submitting something to it and winning a prize is your best bet. But we also want to acknowledge the exceptional work of our previous winners and offer them a chance to participate on a go-forward basis in this project if they are interested.

We will be paying authors a base rate of $100 for their articles. But if you think you should earn more — because of the size of your audience or the quality of your writing — the form includes a section to name a price, and you are always welcome to negotiate.

Building “Boyd Scouts”

We’ve created a Signal group chat that is open to paid subscribers and essay contest winners. If that’s you, a separate email with the link to join will be on its way.

Part of upgrading Boyd from a (very high quality) Substack into a real community is building a town square — somewhere for people who buy into the LFG American framework to connect and talk through the sprint and everything adjacent to it. A couple of months ago we announced a Discord, and while it still exists, the platform never quite fit. As a result, we will be retiring it and replacing that platform with a group chat, which we think better matches the ~vibe~.

I invite you to join us if this sounds like fun. It is as easy as becoming a paid subscriber — all donations are tax deductible.

Essay Contest

Keep your eyes peeled for our essay contest announcement article, which will have all of the formal rules and submission form. But here is what we are thinking for upgrading the contest.

Over the last three sprints we’ve given roughly $7k a quarter in awards — over $20k total. It’s been a great bang for our buck, and the ideas and enthusiasm coming out of you all have been incredible.

Our model so far has been a top prize of $2,500, second and third place prizes of $1,000, and $300 honorariums. As always, we’re looking for the alpha, or basically the good ideas that we didn’t know we hadn’t thought of. But this time we’ve overhauled the prize structure, splitting the top prizes into two categories, and awarding superlatives instead of general honorariums.

Top prizes, $2,000 each:

• People’s Choice — decided by a March Madness–style bracket competition on Notes. We encourage you to campaign for your own article aggressively.

• Boyd’s Choice — internally judged by Boyd in a similar way as the previous top prizes: we’re looking for bold, asymmetric ideas premised on fundamental truths.

Five categories, $500 each:

• Most Engagement — measured by likes on Substack as of the end of the essay contest window, subject to our discretion.

• Most Feasible — the idea that could really work, and might actually have a chance of happening.

• Most Outlandish — the idea we couldn’t publish but couldn’t stop thinking about.

• Least Socialist — awarded with reverence and mild concern regarding the current political climate.

• Most Lobbied-Against in Washington — save your prize money for legal fees.

I’m excited to keep this going for years to come — and I look forward to tackling healthcare with you all.

Share

Fundamentally, America — really any nation for that matter — has a discrete set of non-destructive paths out of a fiscal bind. It can collect more, promise less, or produce more. Given the current political economy, the likelier path on “promising less” is that nothing structural happens on entitlements at all: nothing on Social Security until the retirement trust fund’s projected 2032 depletion forces the issue, nothing on Medicare until “waste, fraud, and abuse” becomes something more than a slopulist virtue signal. And aside from the swelling advocacy for a wealth tax, tax hikes with any potential for material fiscal impact — “collecting more” — are a non-starter heading into a new election cycle.

Over the years we’ve also seen a litany of other countries try to get cute and find alternative, more conspicuous escape hatches, often through monetary channels. These “fiscal dominance” tricks — namely inflating the debt away and/or financial repression —almost always fail. The US should not even consider them. The former would burn the very trust that makes our borrowing cheap in the first place, while the latter fundamentally requires a captive domestic bond market, a thing the US once had but no longer does.

So, if all the exits besides the “grow our way out” one are off the table, as is monetary sleight-of-hand, how do we craft policy that gives growth the best possible chance to materialize while not depending on first winning a national political argument? Well, what if I told you all that stands between America and a proper boom — a genuine, New Deal-scale building renaissance — is getting enough technocrats to read Ezra Klein’s book?

Kidding. But in all seriousness, cross-partisan, pro-growth policy preferences have gained very little traction outside the politically-online / econ-blogging milieus. Almost no ordinary voter understands “abundance” to mean anything in particular — a May 2026 Times/Siena poll found that 91 percent of Democratic voters had not even heard of it — so there’s basically no actual coalition organized around the label. And as an “elite persuasion” movement, outside a decent legislative push here and there, it’s arguably not doing much better.

Meanwhile in Washington, mercantilism (the populist-MAGA-right), degrowthism (Boomers, NIMBYs, and modernity-rejectors), and neo-Brandeisianism (the “big is bad,” DSA-left) are all eating mindshare in alarming fashion. Among voters, the two tribes that are most reliably pro-growth — “aggressive deployers” and “center-right abundance” — together comprise just 30% of the electorate.

But here’s the good news: neither Washington nor the online commentariat — nor a critical mass of voters — is needed to summon investment in the real economy into existence. Instead, private capital, of its own accord, is already attempting one of the largest sectoral investment buildouts in American history, all centered around data centers and AI. That is the genuine economic and fiscal gift of our time. Not because of what the data centers themselves will pay in taxes — modest at the federal level, even if substantial locally — but because of the production levels the rest of the economy might be empowered to reach once compute is cheap, capable, and widely accessible.

It would be an almost incomprehensible own-goal for public policy to strangle this outcome before it runs its course. The policy task is therefore, if not simple, at least known. That is: stop our inherited veto points from blocking the investment boom, and keep the financial system from buckling along the way, so that capital can become powered compute, compute can diffuse through existing workflows, and the resulting additional output can yield more domestic taxable income.

If “abundance” — abundant power, abundant access, abundant credit — cannot win as an electoral imperative, it has to at least become the institutional default.

How?

Reason from the principles of “abundance,” stop focusing on clever labeling and mimetics that clearly don’t work, ditch coalition-building and persuasion efforts, and just write the pro-growth solutions into the damn rules — for the systems that technocratic policy can make or break. For the electric grid that has to power the data centers, for the competitive interfaces that let ordinary firms use the compute, and for the dollar credit markets that must run smoothly to finance it all.

But before wonking out on policy moves, it’s worth wonking out on the data center-capex-growth boom already underway — to understand how it is distinct from a productivity boom, why the distinction matters, and why it’s the latter that’s required for the nation’s fiscal outlook to improve.

Subscribe now

Capex surge ≠ productivity boom.

A brief moment to reflect on just how remarkable AI-capex boom really is. Economists and Wall Street spent much of the fifteen years following the financial crisis scratching their heads, wondering why cheap money was failing to get firms to deploy capital in the real economy. Rates are much higher now, yet only now has the investment arrived. Boy has it arrived.1

The scale is historic: information processing equipment, software, and data center construction accounted one third of all US business investment by Q3 20252, the largest share since records began in 1947. Harvard professor Jason Furman ran the numbers and found that combined, investment in information processing equipment and software accounted for about 4% of GDP and 92% of GDP-growth in the first half of 2025. A year later now, and that 4% of GDP has expanded to nearly 5%, eclipsing the dot-com era high watermark:

The last time these categories ran as hot as they are now was in the heat of the 1990s IT buildout, a worthwhile comparison for two reasons. One, because the internet was, like AI, a general-purpose technology with the potential for revolutionizing whole industries and sectors the economy. And two, because the dot-com investment surge came first, while the subsequent productivity payoff took the better part of a decade to show up. Indeed, the current growth being fueled by capex says little on its own about whether — or when — an economy-wide productivity boom might arrive:

Think about it in b-school terms. If GDP is a country’s closest analogue to a company’s top-line (revenue), one could think of productivity as something like “operating leverage.” Whereas companies may use financial leverage to magnify returns by issuing debt to employ more capital, operating leverage speaks to a company’s ability to turn top-line growth into disproportionately higher earnings growth. This happens when a meaningful share of costs become fixed (vs. variable), such that total costs don't scale one-to-one with revenue.

Similarly, in macroeconomic terms, productivity is what determines whether output keeps compounding without inputs — labor and capital — increasing in lockstep. For the public balance sheet, that means higher wages, higher profits, and ultimately more taxable income against public spending levels and a debt stack that does not rise at the same rate. This is precisely where the fiscal case lives.

Why the capex-growth vs. productivity-growth distinction matters, fiscally.

First, if GDP growth is driven predominantly by heavy capex, the immediate result is a front-loaded surge in construction and hardware revenue. And that’s good. Nominally, near-term growth is good; more compute is good. But it does not yet imply a recurring productivity dividend.

Because after initial construction, data centers operate autonomously (not much ongoing human labor required), because of the way depreciation is treated in federal tax law (subsidized), and because the lionshare of maintenance capex is tied to replacing the semiconductor chips (which have short depreciattion cycles), data centers alone are not very tax-accretive. At least at the federal level, the payoff arrives only if abundant compute translates to increased output — higher wages and profits — across the wider economy, boosting federal tax receipts.3

Second, the expectation alone of a future productivity miracle can create a paradoxical demand shock today. Call it something like Greenspan’s famous hunch, but in reverse. In the ’90s, Greenspan’s bet that productivity had already accelerated became part of the Fed’s rationale for resisting calls to tighten. But now, the promise of future productivity — the economy-wide expectation of it — may be pulling spending (aggregate demand) forward today, keeping inflation higher than it otherwise would be. Businesses spend like profits will be higher tomorrow, while AI-driven stock wealth causes households to look at their 401(k)s and think, “yeah, I can afford this!”

That doesn't make AI mechanically inflationary, or the Fed mechanically hawkish. But if the buildout and the stock-wealth effect lift demand faster than new power and real adoption lift supply, the Fed has less room to ease — thus rates stay higher, and financial conditions tighter, for longer.

Which leads into the final point, on the public balance sheet impact. We are at a critical juncture. As of July 31, about a third of marketable Treasury debt is set to mature within twelve months — a $10.5 trillion maturity wall that must be refinanced. And the government's effective exposure is even greater than that number suggests. The Fed holds trillions in long-dated Treasuries financed by overnight borrowing from banks, which makes the consolidated public balance sheet more rate-sensitive still — because when rates go up, those overnight reserves reprice right away, hitting the government’s interest bill immediately. (The Fed’s profits and losses land on the Treasury’s books, which is to say, on the taxpayer.)

If one could wave a wand and make rates nil and inflation low again as was the case throughout the 2010s, the current capex boom would be considerably easier. But alas, if the buildout keeps demand hot while inflation remains above target, the Fed has less room to offset rising long-term rates without risking its inflation-fighting credibility. That means this enormous near-term exposure reprices at prevailing rates and the government’s interest bill rises quickly — potentially years before any productivity dividend could possibly show up in tax receipts.

Therefore the sooner the AI-productivity lift comes to fruition, the better the country’s fiscal outlook. That’s just the reality:

Like it or not, the country’s future is wagered on AI.

Not because Treasury borrowed to build data centers, or because David Sacks and Marc Andreessen are pals with Scott Bessent and advisors to the current administration, but because an already-deeply-leveraged fiscal regime has become unusually sensitive to whether taxable output rises faster than interest expense swells. It really is that simple. Stanford economist has called Treasuries a “long call option on AI”; Lazard CEO similarly referred to the US economy as a “levered bet on AI.” These characterizations may seem provocative, but the underlying dilemma is real.

The bad side of the coin, the nightmare scenario, is where we — Washington, Wall Street, Main Street — hum along expecting a productivity payoff, the fruits of which may or may not actually show up tomorrow. In other words, a scenario where we will have invested in top-line growth to bolster the bottom-line, only to find out that the “operating leverage” was never really there to begin with.

How is the bet going?

The country’s “top-line” is growing. Task-level gains are real. Some firms now report modest positive effects — labor productivity (output per hour) is up about 2.5% a year since late 2022 — but this could just be the result of handing workers more capital and running it hot (tokenmaxxing). What it’s not yet proof of is an economy producing more with less.4

Total factor productivity, the actual efficiency stat (though noisier), has shown little clear, AI-specific acceleration. The KC Fed finds AI adoption explains little of which industries even drove the pickup, and nearly nine in ten executives in a survey of almost 6,000 firms reported no realized own-firm productivity gain over the prior three years. The same respondents nevertheless expected AI to raise productivity 1.4 percent and output 0.8 percent over the next three years.

That gap is the wager: capital is being committed against expectations, but the conversion has not yet shown up broadly in realized output. Basically, task-level savings haven’t yet translated broadly into measurable output and earnings — i.e., an economy-wide AI-productivity lift — which means the new taxable income, and therefore the fiscal dividend, has not yet appeared at scale.5

In short, and simply put, for the fiscal payoff to actually come to fruition, the following still must happen:

1. Capital deployed → powered compute

2. Abundant compute → broad AI-adoption

3. AI-adoption → additional domestic output

4. Additional output → higher tax receipts

5. Higher tax receipts → outpace the associated increase interest costs and spending

How much of the debt path is riding on the AI boom?

It’s not simply a matter of whether AI raises productivity. It’s also about ~when~ the productivity gains arrive, how robust they are, and what happens to debt/credit markets along the way. The dynamic is easier to visualize than to describe:

Interactive debt path model →

Policy moves to make the AI-bet pay off, fiscally.

Five policy moves, none of them requiring a voter mandate, nor politicians in Washington picking winners.

1) Decide where America can build, and then just let it build.

Plain-vanilla, pro-growth policy, straight out of Abundance playbook — but worth doubling down on, because people and firms trying to build things in this country still face a gauntlet of political and legal hurdles (veto points). This is now especially true for data center projects, which have rapidly become a breeding ground for hyperlocalism. But it can be routed around.

The federal government should, ex-ante, publish the permitting requirements and let states and localities opt into certified “zones.” And then, once a project meets the rules, review periods should mostly just be audits of conformity.

2) Give faster access to power to projects that help build out the grid.

A project shouldn’t have to wait years — sometimes as many as eight!! — for a conventional grid connection. If it is willing to add enough power and flexibility to smooth over the strain it might be placing on the grid, it should be fast-tracked.

Specifically, if a data center or any project requires new substations, transmission, or generating capacity, it should be able to sign a long-term take-or-pay contract, post collateral, and bear the stranded cost if it leaves. If it wants firm, round-the-clock power, it pays the full cost. If it can curtail during scarcity — or supply deliverable generation or storage — it should pay less. The theme here is just price the externalities.

Fortunately, FERC is already pushing regional grid operators in this direction — it’s pushing for flexible service for large projects willing to limit how much they tap the grid, paired with new generation slated to come online in that same constrained region. This is the right move, and a bargain that should be made national and enforceable.6

3) Make AI easy to try, easy to switch, and legal to use for ordinary firms.

The economic purpose here is diffusion of the general-purpose technology, and the barriers aren’t necessarily the technology itself, but permission structures and information security. It’s still very early-innings, but a few things are clear already.

First, a firm’s data should belong to it, and a firm should be allowed to point any AI tool at it, bearing the existing legal and business risks it already faces. There doesn’t need to be a policy for this; most companies already have internal AI committees and sometimes even entire departments taking these risks very seriously.

Second, with more than a fifth of American workers needing a government license to do their job, a safe harbor model should be adopted, allowing licensed professionals and firms to test and use AI for specified tasks so long as an accountable human remains in the loop and there are adequate disclosures and audit trails. Allowance for (and sandboxing of) iteration and tinkering is crucial for firms to find new technological use cases and deploy them safely. The regulatory approach should be in service of this, and the regulatory default should be permission with understood liability.

And finally, what should absolutely not happen is for anything resembling the EU’s growth-stifling tech-regulatory playbook (GDPR, Revised Product Liability Directive, Cyber Resilience Act, etc.) to even be considered in the US. These types of moves — such as the recently-passed watermarking rules under the EU AI Act, requiring providers of generative systems to embed machine-readable marks in their outputs — rarely lead to their intended outcomes, but instead impose undue costs on firms and consumers alike.

4) Stop manufacturing fiscal crises.

Kill the debt ceiling — it’s low-hanging fruit. I know this seems counterintuitive, and maybe even sounds orthogonal to the entire AI-fiscal thesis; it isn’t. A country asking global investors to absorb historic levels of debt issuance shouldn’t schedule recurring hostage negotiations with itself. Debt-ceiling impasses are entirely political theater. The ceiling never actually restrains spending, but the brinksmanship carries real, tangible costs. Shutdowns themselves have also been shown to reduce GDP by billions (permanent losses, not recovered).

However large or small the negative impact here, the point is that it’s all for exactly nothing. It’s completely fake, and avoidable, and we should just adopt an automatic continuing appropriation whenever Congress misses its funding deadline.

5) Don't let rising interest rates turn into a market meltdown.

In plain English: the Treasury market should be made more liquid, so inevitable bouts of volatility do not devolve into crises. A set of shock absorbers could be built-in. We already have circuit breakers in the stock market, and it turns out they work pretty well. Treasuries do not need identical trading halts, but they do need safeguards capable of stopping a temporary liquidity failure from cascading through a market that is orders of magnitude more important than stonks.

The policy moves here would materially improve both liquidity and information asymmetries in the most important market in the world. There is a caveat, which is that price discovery is incredibly important. Global bond yields are currently spiking for very real and fundamental reasons (not least the glut of government debt), and efforts by governments and central banks to fight this without getting their fiscal houses in order will have nasty second- and third-order consequences. So these moves aren’t about artificially lowering rates. Rather, they’re about insurance and information — and ultimately preventing legitimate market repricing episodes from becoming economic doom loops.7

First, complete SEC-mandated central clearing of cash Treasuries, tapering dependence on broker-dealer balance sheets in times of market panic. Second, make the Fed’s standing repo facility clearable, at a penalty price, so the backstop actually reaches the market instead of eating scarce dealer balance sheet capacity. Third, run a stress test America has weirdly never run: put banks, hedge funds, money funds, and the clearinghouse through the same shock, together, and measure the resulting leverage and liquidity in aggregate. Finally, give Treasury a standing way to swap fragmented old bonds for liquid benchmarks — the switch-style exchange operations it is actively studying now — and make the consolidated Treasury-Fed rate-reset picture a permanent and prominent part of quarterly refunding.

“But what about muh…”

Most of the pushback on accelerating the data center buildout and AI diffusion across the economy — water usage, noise pollution, aesthetics — is certifiably populist garbage. But some is legitimate. So I’ll address the three most legitimate counterarguments I’ve seen: jobs, the environment, and electricity prices.

Jobs?

This is the lousier of the objections. The premise is that economic policy should preserve the job market as it exists today. But when has that ever actually been a good outcome? Moreover, a serious question: why is it even a bad idea to vaporize all the bullshit tasks we currently do — e.g., email jobs, number-pushing, order-taking — such that humans may graduate to handling more meaningful types of work? Isn’t this exactly what has happened through every industrial-revolutionary phase?

Either way, so far the “jobs-pocalypse” is a dud. Now years into the AI era, still no data signalling any kind of meaningful disruption in the labor market. Most firms themselves, asked about their hiring outlook, say they’re actually discovering that deploying AI properly takes more people, not fewer.8

The broader point here though, irrespective of any near-term trend, is that creative destruction is good, actually. The economy already destroys and creates tens of millions of jobs every year underneath the headline numbers. And the world is as mobile as it’s ever been. Helping people traverse the job market and physically relocate are major modern problems we’ve gotten pretty good at solving.9 Being poorer and stagnant is not.

Environment?

To the tree huggers, with whom I genuinely sympathize: your concerns around the negative externalities imposed by data centers — especially emissions — are real. But direct your anger at the political, procedural, and vetocratic machinery that makes nuclear plants, transmission lines, solar farms, energy storage, and the rest take substantially longer than gas turbines to build and bring online.

What the data center moratorium crowd is seemingly confused about is that moratoria do functionally nothing to eliminate the demand for compute. Instead, these bans just determine where the next data center will not go, leaving the local tax windfall up for grabs for the next county or state over. The buildout pushes on regardless, and so long as clean power remains slower and more costly to get approved versus gas, the buildout will increasingly go fossil.

Texas is the counterfactual (notwithstanding Gov. Abbott’s recent degrowthism bent). Its permissive connect-and-manage grid enables generation to be built quickly — last year it passed California in utility-scale solar generation. Think about the irony: Texas, with its famously laid-back environmental regulatory regime, has emerged as the new leader in a major category of clean energy generation.

Electricity bills?

This objection is very real: recent Dallas Fed estimates point to a 3-5% increase in wholesale power prices nationwide, with larger effects concentrated in major data center corridors. But the answer is not hyper-regulating our way into lower demand for power. It’s exactly the opposite, which is government getting out of the way of supply.

The answer here is price — price the externalities, give permission to build, and ultimately leave it up to hyperscalers to face the true cost of the power they consume and determine whether that cost is worth building into anyway. Utilities, merchant generators, private microgrids, and even data center developers themselves, should be able to freely participate in an open marketplace to compete for who gets to serve the load.

If approached well from a regulatory standpoint, the hyperscaler buildout could serve as a strong catalyst for massively upgrading the grid both in terms of capacity and interconnectivity. In the long run, such infrastructure gains are shared by everyone, households and firms alike.

Conclusion: Growth = good.

What if the whole thing is a bubble? It might be! The AI-capex boom is disproportionately carrying growth right now, which cuts both ways. Absent it, inflation and electricity prices might be lower, thus lower interest rates, which would unlock higher growth potential in other sectors and potentially ease the government’s fiscal pinch. But rough estimates say that counterfactual growth rate would be half of what the rate of growth is now, at best. And all else equal, that would probably mean lower tax receipts for the government — higher deficits.

But let’s ditch the thought exercise and live in reality. The AI trade, if it goes tits-up, would mean the loss of the growth engine, it would mean household spending seizes up due to negative wealth effects, and it would mean AI-linked debt goes sour, which would potentially be a systemic event at this point. This is reality, and we should double-down on it.

In terms of policy moves, almost everything I laid out above is a no-regrets move, and would help, not hurt, in either the bull or bear case. Permitting reform, better grid rules, more seamless throughput in markets, less political theater in Washington — these things are worth pursuing in any event, but in the bear case especially. That’s literally the whole point of betting on sustainable institutional rule structures: the future is always uncertain, and popular opinion can remain irrational longer than we can remain solvent.10

Whether we can literally “grow our way out of the debt” is the wrong question. In the abstract? Yes. Within any plausible range of productivity gains? Almost certainly not. On current policy, it is a near-certainty that the debt will continue to swell decades into the future. So the real question concerns the second derivative — whether the debt trajectory inflects higher or lower, even if it’s still upward-sloping in either scenario. The fundamental economic question, then, is whether we can legitimately find ways to seamlessly diffuse the technology such that ordinary firms can produce more with less.

On a personal level, the growth-versus-degrowth tension really hits home. I am part of a unique generation that is old enough to remember AOL and dial-up internet, but young enough to see the value in a $200/month Claude Code subscription today. Unfortunately, I’m also part of the generation that has inhereted, with no real say, a growth-stymieing political, fiscal, and regulatory mess ushered in by a cohort of people who literally won’t live to know the difference. I don’t want my country to become EU-ified and sclerotic and dusty. France has wonderful cafés and a beautiful countryside… with the GDP-per-capita of, like, Mississippi. Yuck.

No, I want America to be modern, and to allow air conditioning without a permit. I want us to have dynamic institutions, and to maintain our technical supremacy. I want us to continue to show the world that leaving the private sector free to ‘just do things’ — crucially, at scale and for a profit — is the best formula for unlocking prosperity in the aggregate.

Growth itself is not a fiscal plan, but unnecessarily foreclosing the country’s largest growth option in generations is a fiscal choice.

Leave a comment

10

There is, of course, one other proven growth lever — letting more working-age people into the country — but that’s a different essay for a different day. However, ’s 3rd place submission in our recent essay contest is probably about the best solution we’ve seen on this!

Welcome the Stranger

Welcome and Tax Mass Immigration to Escape America's Fiscal Trap

This article is part of The Boyd Institute’s quarterly policy sprint on the debt and deficit. To learn more about them and the work they do or submit your own article click here. NOTE: After posting, I made a few last-minute edits just before the deadline…

Read more

5 months ago · 13 likes · 4 comments · Nathan Smith

1st Place —

Incentives, incentives, incentives.

That is what Ma’s grand-prize-winning essay flagged as the real root cause of the fiscal trap the US finds itself in. He argued that forty years of accumulated tax architecture has taught capital to pool idly in appreciating assets rather than circulate throughout the productive economy, where it would generate higher taxable wages and profits along the way.

The solution put forth is what Ma dubbed “productive deployment weighting” — taxing idle assets harder than working ones to produce real output gains by spurring capital reallocation. It’s a fairly novel idea. A deep survey of the economic literature yields just one formal cousin: a 2023 paper in the Quarterly Journal of Economics, "Use It or Lose It" by Guvenen et al., which largely corroborates Ma's conjecture.

Overall this was just an exceptional essay, and it was the only one in the field that truly answered the question of how America, the wealthiest country in the world with the deepest capital markets, could possibly have a debt problem:

"America is not running out of money. It is running out of circulation. Debt is a symptom."

The Marketplace by Quy Ma

The Velocity Problem: Why America’s Fiscal Trap Is a Circulation Failure and How to Fix It

This article is part of The Boyd Institute’s quarterly policy sprint on the debt and deficit. To learn more about them and the work they do, or submit your own article, click here…

Read more

4 months ago · 44 likes · 21 comments · Quy Ma

2nd Place —

We’ve argued throughout the sprint on debt and deficits that big spending cuts are (more or less) politically impossible in the US — at least with the current composite of electoral interests. This essay directly challenged that premise, positing that spending cuts are only impossible if those interests aren’t economically (or directly financially) placated.

While there were a few “buy our way out” submissions in the essay contest, Orson’s proposal was hands down the most cleverly constructed and completely worked-out. The logic runs like this:

The kicker here is that the bonds matter enormously. A promised benefit and a government bond are legally different animals — the former is a promise that can be reneged on while the latter cannot be cut, as it is a debt contract protected by the Constitution. And unlike the current system, which owes money forever, these bonds are paid off in 25 years and it's over.

Now, the biggest open question is how trillions in new bond supply gets plopped into the markets without rates blowing out. But as far as political-economic engineering goes, this essay was spectacular. Imagine the stump speech:

"The future is unknowable, I just want to put tens of thousands of dollars into your bank account before it all falls apart."

What if We Just —

What If We Just… Bought Our Way Out of America's Looming Fiscal Collapse?

This article is part of The Boyd Institute’s quarterly policy sprint on the debt and deficit. To learn more about them and the work they do or submit your own article click here…

Read more

3 months ago · 24 likes · 7 comments · S.B. Orson

3rd Place —

Another banger from Nathan, who took the top prize in our housing sprint, and who recently became a regular writer for Cato’s HumanProgress.org!

The longstanding caveat to the economic argument for pulling the immigration lever for growth is that immigration-driven systems can reach a Ponzi-esque failure mode when new entrants generate more benefit claims than government revenue. Milton Friedman warned of this half a century ago — "It's just obvious you can't have free immigration and a welfare state." — and America today is unequivocally a welfare state. Nathan’s proposal addresses this head-on.

Specifically, the argument made here was that immigration should be priced at the door. $5,000, roughly the “coyote” price, except it buys legal status — as well as licensing surcharges, extra property tax on non-citizen purchases, a substantial citizenship fee, and income surcharges of 30-50%, collectible because the penalty for evasion is deportation.

And for the econ-brained folks, Nathan showed his work with a model and data to boot. He even tilted it against himself — in his model immigrants never fully close the human-capital gap. Wages dip, then recover almost entirely as investment floods in behind the new labor; in practice the capital moves faster than the people, so natives likely never feel the dip.

Of course anti-immigration populists aren’t often keen on hearing the economic arguments, but this essay made the imperative clear:

"We do not have to accept a future of managed decline, crushing taxes, or a closed society. We can escape the fiscal trap. It is time to welcome the talent of the world, price the privilege of participation, and build a solvent, dynamic America."

Welcome the Stranger

Welcome and Tax Mass Immigration to Escape America's Fiscal Trap

This article is part of The Boyd Institute’s quarterly policy sprint on the debt and deficit. To learn more about them and the work they do or submit your own article click here. NOTE: After posting, I made a few last-minute edits just before the deadline…

Read more

4 months ago · 13 likes · 4 comments · Nathan Smith

Stay tuned for our announcement of the upcoming sprint topic and essay contest — some changes are brewing!

Leave a comment

The top three essays will be announced via cross-post to Boyd subscribers over the coming ~week. This post is for the eight honorariums, which took a lot of deliberating on to narrow down.

We want to again thank everyone who participated and encourage even more people to participate in our future essay contests!

Subscribe now

— “The Tax State Reads Wages. It Cannot Read Compute.“

The argument here was that AI is moving income into forms that legacy tax machinery was never built to observe — accruing “to capital, to equity, to the owners of compute, and increasingly to autonomous systems whose output has no wage line at all.”

The essay’s answer for commensurate tax revenue with AI-derived growth is Federal Compute Withholding, with everything raised used to cut payroll taxes on labor. The implementation was thoroughly thought through, down to a rate that only rises if labor’s share of income actually falls.

The design is revenue-neutral on purpose and closes none of the deficit. But we found the essay compelling because it flagged what seems to be a real risk — tax receipts actually falling as output grows — and provided a well-fleshed-out solution.

— “DOGE, Done Right“

This essay was just what the doctor ordered for us. We lamented DOGE several times throughout the sprint for taking the wrong approach and amounting to slopulist cope.

AP’s version of addressing “waste, fraud, and abuse” was to flip GAO recommendations from opt-in to opt-out, where agencies must implement anything with proven financial benefit within two years, and stonewalling becomes a fireable offense. Paired with an IRS funded to close the tax gap, he pegs it at $2-4 trillion in savings over ten years.

— “The Generational Deal“

One statement in this essay really stood out, and that is that anyone selling you a painless version of fiscal course-correction is lying. Then the author takes his own advice: the plan costs $17 trillion by his own math, and his own cohort pays.

What this would fund is the end of pay-as-you-go replaced by a sovereign wealth fund built to be politically untouchable, with its returns eventually seeding retirement accounts citizens own outright.

Eli Sheets — “Staying Is Nowhere”

This was among the least exciting yet most coherent essays.

The thesis is that there’s no single fix, only a portfolio. The compelling part was that every proposal was specified down to the level a staffer could draft from. Not “reform Medicare Advantage” but coding intensity, favorable selection and the spending gap MedPAC has documented.

It also ruled out holding rates down through monetary policy, because it would damage the dollar and breed the complacency that got us here — which seems unequivocally true based on the research we did over the course of the sprint.

Overall, just a really impressive policy paper.

Debt Writeup Elijah Sheets

338KB ∙ PDF file

Download

Download

— “Achieving fiscal solvency by not letting private equity bros torture grandma to death“

This was a banger. It said many of the quiet parts out loud, and ultimately landed on the rather unsavory prescription of assisted dying, incentivized and at scale.

But before the essay got dark, it priced out every popular fix: seizing all billionaire wealth covers a third of one year's deficit, and means-testing, corporate taxes, and retirement at 73 run about $300 billion each. It also shined a light on the abuses found in PE-owned healthcare facilities, along with hospices recruiting Alzheimer's patients because they tend to outlive the six-month billing window.

Though the prescription may be contestable on grounds of “social cohesion” and whatnot, we really appreciated this essay for its boldness and its research receipts.

— “Project Hail Mary Zero“

This essay went after healthcare costs upstream of the payment system, at metabolic disease, on the logic that the cheapest way to stop paying for something is for fewer people to need it. Around it he stacks the build-side agenda: barriers to energy generation removed, density-limiting zoning killed, defense procurement rebuilt around drones. Energy is productivity, productivity is tax revenue, revenue is how you avoid a crisis. Overall this was a very well-rounded buffet of proposals.

— “We’re not going to solve the debt crisis“

Reid had an idea nobody else had: CaaS — sell governance and security as a service, one country to another. The US supplies rule of law, contract enforcement and protection to willing developing nations in exchange for resource rights and a share of the growth that follows.

His claim is that cheap surveillance has collapsed the manpower cost that made this unprofitable historically. But the demand is real — five West African heads of state were at the White House last summer pitching something close to it.

— “Twenty-Five (Mostly) Terrible Ideas to Fix America’s Debt Problem“

This one got high marks for creativity. The meta-insight here is that when you completely reframe the approach, unironically, you start to think outside the box and may actually come up with an interesting solution or two.

The most serious one in the pile, premised on the notion that neither party can cut alone without getting slaughtered for it, and that the only fix that survives politics is one where both share the blame: a bigger, uglier Simpson-Bowles.

Leave a comment

Congratulations to all eight, and thanks to everyone who entered. 31 essays is a lot of unpaid, careful work on a problem most people would rather not look at or talk about.

The spending side

, Project Hail Mary Zero — The federal government is a health insurance plan that happens to own an army, and type 2 diabetes and its downstream diseases are the largest item. So metabolic disease is fiscal policy: free GLP-1 drugs for anyone diabetic or pre-diabetic, an end to corn and soy subsidies, and hospice good enough to choose.

, Achieving fiscal solvency by not letting private equity bros torture grandma to death — Roughly 70% of federal spending on identifiable groups goes to the elderly, concentrated in the final years of life, and private-equity-owned nursing homes already deliver worse care at higher cost. He proposes two voluntary exits after 65: one trading a do-not-resuscitate order for unrestricted access to drugs and risk, the other a paid assisted-dying program on the Canadian model.

, The Generational Deal — Medicare Advantage plans are paid more for sicker patients, giving them a reason to record patients as sicker than they are. Ending that overpayment closes most of Medicare’s $193 billion hospital-fund shortfall with nobody losing coverage, buying time for the larger trade: a sovereign wealth fund paying out as citizen accounts, and Medicaid devolved to the states.

, What If We Just… Bought Our Way Out? — Nobody votes to lose a benefit, but people will sell one. Pay every American 30% of the present value of their lifetime benefits in Treasury bonds, abolish Social Security, Medicare, and FICA, and replace delivery with at-cost government clinics, catastrophic Medicaid, and a civilian medic corps.

, Bariatric surgery for Leviathan, STAT! — Congress will not diet, so remove its capacity to overeat: repealing the Sixteenth Amendment eliminates 51% of federal revenue, enough for discretionary spending and interest but not entitlements, forcing cuts no legislature would pass.

, How Does America Escape Its Fiscal Trap? — Federal money props up the price of whatever it buys, and Medicare’s open-ended design turns every advance in medicine into a larger bill, so phase the benefits down and move people onto 401(k)s and HSAs.

, A Social Security Severance for Workers in the age of an AI-Bolstered Workforce — A firm can deduct the software that replaces a worker while the worker’s payroll taxes disappear, so tax AI productivity gains to backfill Social Security, and let under-40s cash out at $15,000 to $90,000.

, Punt to the People on Social Security Reform — Congress cannot cast a vote that raises taxes or cuts benefits, but a referendum can settle it without producing a losing side. Put both on a national ballot and close the 75-year gap in proportion to the result: 40% for higher payroll taxes closes 40% that way, slower benefit growth the rest.

, Gerontocracy and Its Discontents — A generation that had few children now draws on a smaller working cohort, and eldercare has become a leading growth industry. Coverage starting at 65 with nothing before it leaves the young carrying untreated illness for decades.

Eli Sheets, Staying Is Nowhere (submitted as PDF1) — No single measure closes the gap, so stack the ones that survive a vote: Social Security reshaped along the EITC’s phase-out curve, a smaller federal share of the Medicaid expansion, a credit paying seniors to leave Medicare Advantage, restored IRS auditing. He catches what such lists miss — those seniors cannot buy Medigap after 65, so the credit needs an enrollment window attached.

The revenue side

, Welcome and Tax Mass Immigration to Escape America’s Fiscal Trap — Admit roughly 200 million immigrant workers and charge for entry through visa fees, licensing surcharges, and income surtaxes, with benefits restricted to citizens. The added labor roughly doubles GDP and the migration taxes compensate native workers in cash, so the wage effects get paid off rather than denied.

, Fixing the Deficit without Bankrupting the Young — A VAT taxes spending, which falls hardest on the young; a land value tax falls on landowners, who are disproportionately the retirees drawing the benefits. A 5.6% tax on the $32 trillion of non-federal land covers the deficit, phased in over twelve years, with payment deferrable until sale.

, How America Escapes Its Fiscal Trap: Fifteen Doors Out — Foreigners pay large sums for American degrees, visas, and market access, and almost none of it reaches the Treasury. Capture it through auctioned residency slots, visa surcharges, a paid immigration fast-track, and a remittance tax, with every dollar locked in a debt-retirement fund.

, The Velocity Problem — Tax revenue comes from transactions, and forty years of capital gains preferences and stepped-up basis have taught capital to sit still. Tax gains by whether the asset was put to work rather than how long it was held.

, The Tax State Reads Wages. It Cannot Read Compute. — The government funds itself through payroll withholding, which works only because wages are visible and countable, and AI shifts production into forms with no wage line to read. The state is going blind. Levy a small charge on AI training and inference, where output is already metered, and return it as payroll tax cuts.

, Stocks, Flows and Fiscal Madness — Government deficits are matched dollar for dollar by surpluses elsewhere, so cutting the deficit takes money off someone’s balance sheet — take it from foreign holders and corporate cash rather than households, through a wealth tax and buyback taxation.

, How to balance the federal budget — Restoring pre-1997 tax rates and applying FICA to all income, investment income included, closes the gap arithmetically. He traces where it came from — Social Security up 2.4% of GDP since 1966, Medicare up 3.5%, Part D and Medicare Advantage passed without revenue — then retires interest with platinum coins.

Build a national asset

, A US Sovereign Wealth Fund as a Foil to US Debt — An asset can be accumulated faster than $39 trillion can be paid down: seed a Norway-style fund with repatriated tax-haven assets and a 1% financial transactions tax, spend 4% a year, and cover Social Security by year fifteen, Medicare by year thirty.

Abdulganiyy Abdulwasiu Olanrewaju, The Growth Covenant (submitted as PDF2) — The worker-to-retiree ratio has fallen from 5:1 in 1960 toward 2.5:1 by 2040, so raise the number of workers rather than cut benefits. Entrench three things constitutionally: skills-weighted immigration that expands with the fiscal gap, a fiscal commission that enforces rather than recommends, and conversion of debt above 70% of GDP into growth-linked securities.

Change who decides

, How Do We Solve a Crisis Very Few People Care About — Voters have no incentive to inform themselves about the debt and politicians have every incentive to add to it, so change who holds the veto: a fourth branch of bondholders able to block any bill that worsens it, or elections weighted toward informed voters.

, Sovereign Solvency — Spending expands to fill whatever revenue is available, so tax increases fund more spending rather than closing the gap, and deadweight loss rises with the square of the rate. That makes this a spending problem. Bind the next Congress the way Ulysses bound himself: a constitutional amendment capping federal spending at 20–22% of prior-year GDP, with the upper band available by majority vote in a recession but never two years running.

, The Fiscal Cliff Crisis Isn’t An Accounting Problem — Every available fix imposes concentrated losses on an organized group — AARP’s 38 million members, the drug and hospital lobbies, bondholders — and representative government exists to prevent that. The reforms that did happen ran three things at once: a frontman with standing to absorb blame, a technocratic team working out of view, and enough spectacle to hold attention elsewhere while it lands. He grades Reagan, Thatcher, and the city receiverships on it, and reads 2028 as a casting problem.

, We’re not going to solve the debt crisis — Nothing gets fixed before the crisis arrives, so the useful work now is putting a capable president in office ready to act when it does. His revenue proposal is a neoprotectorate: a country cedes resource rights and a share of future growth for American protection and contract enforcement.

, Unleash the Soldier of Fortune and Slash the Deficit — Cuts are blocked and tax increases suppress the growth needed to outrun the debt, so the remaining revenue lies in what is currently prohibited: “innovation exceptions” letting a business pay a higher rate for exemption from a named regulation, contractors licensed to run badly governed resource states, and asteroid mining.

The money itself

, Fifty-Five Years and Zero Reversals — Productivity is up 178% since 1971 and real wages are up 10%, and much of the difference went into health premiums rather than paychecks: per-person health spending rose from about 100 hours of labor in 1970 to nearly 500 today. Cutting spending only moves the gap onto households.

, Will Deglobalization, Reciprocal Tariffs and Sanctions Resolve U.S. Payment Imbalances? — America’s external deficit comes from saving less than it invests and from supplying the reserve currency, so tariffs relocate it without shrinking it. Closing it takes coordination: surplus countries appreciate on a schedule and expand health and pension provision, while the US saves more at home.

, The Despicable Fix for the Fiscal Cliff — Shadow banking already turns $1 of collateral into $10 of credit by reusing it, and the only thing capping the loop is the haircut taken at each step. So stop issuing debt and start leasing it: Treasury creates a non-deliverable "Super-Bond," leases it to dealers who pledge it for cash, and drives the haircut to zero, since a sovereign that cannot default can simply forbid the margin call.

Grace M. Lo Porto, The Dollar, the Petrodollar, Gold, Tariffs, and How To Escape The Fiscal Trap (submitted as PDF3) — The debt is larger than the entire money supply, so no schedule of taxation retires it and inflation is the only instrument left. Issue a second electronic currency denominated in grams of gold, so the dollar can be inflated to clear the debt while savings hold value.

Outside the categories

, Twenty-Five (Mostly) Terrible Ideas to Fix America’s Debt Problem — Neither party will cut spending unilaterally without losing the next election, which leaves a bipartisan Simpson-Bowles-style commission as the only mechanism that survives contact with democracy. He gets there through twenty-five worse options.

, Deficits and Moonbeams — Any reform should leave the country no worse off if it fails or gets reversed, which disqualifies most of the sharper proposals on offer. Stop freezing foreign assets so the dollar keeps its role, halve the military, protect Social Security by taxing a wider band of income, and pass Medicare for All, since employer-based insurance was a wartime accident and much of its cost is billing rather than care.

, DOGE, Done Right — Both goals everyone assumes are wrong: the debt never needs paying off and the budget never needs balancing, only deficits held below nominal growth, around 3% of GDP. Getting a third of the way there takes no new law — quadruple the GAO's budget and make its recommendations binding within two years unless an agency formally challenges them, and restore the IRS to 1990s funding and audit rates to collect the roughly $600 billion a year already owed.

Subscribe now

Comment and let us know which ones are your favorites! And stay tuned for the announcement of the winners around the end of the month.

Leave a comment

1

“Staying Is Nowhere”

338KB ∙ PDF file

Download

Download

2

The Growth Covenant: A Structural Reimagining of America's Fiscal Architecture

79.7KB ∙ PDF file

Download

Download

3

The Dollar, the Petrodollar, Gold, Tariffs, and How To Escape The Fiscal Trap

40.4KB ∙ PDF file

Download

Download

And yet... I still think Social Security is a vital national program. Because America needs an old-age support system of some kind to support the many citizens who reach senility without sufficient savings, whether through negligence or through circumstances beyond their control. Even the payroll tax strikes me as broadly reasonable — 12.4% capped at $184,500 of income (although if I was designing it from first principles I would probably eliminate the cap altogether).

Not only do I think a program “like” Social Security is essential for “other” people; it also gives me more confidence in the future and increases my willingness to take risks, because I know that even if I stumble, I will not starve in my eighties. And — without revealing too much — one set of my grandparents has only been able to survive at all because of Social Security. In my own case, family could probably have intervened without it, but it would have been an extremely messy and financially difficult situation for everyone, one I am frankly glad Social Security foreclosed. It is to my memories of them, as much as to the thoughts of myself as an old man, that I find myself turning, half-involuntarily, whenever I get too resentful about payroll taxes. In short, I’ve never had an issue with a public pension system.

No — my problem with Social Security is that we have never approached it with either the maturity or the self-awareness it deserves. Figuring out how to balance the social contract between working-age people — who will produce the goods the elderly consume no matter how we structure society, since no one is buying cars built during the Carter administration — against the genuine claims of the retired is an extremely important question, and one we as a country have essentially ignored. The few times we have tried to openly discuss improving the system, as in 1999 or 2005, our republic did everything but cover itself in honor.

Even now, as we approach the program’s “insolvency” — the 2026 Trustees Report puts the date at the fourth quarter of 2032 — the vast majority of Americans don’t know even the basic facts of how our current social contract works. Just look at these key points from a recent Cato poll!

So this article is an attempt to think honestly about our generational contract: how we can fix it, and how we can make it more sustainable, so that when I retire some fifty-odd years from now I can do so in confidence, without the worming thought that I am robbing the future. My proposal is a bold, asymmetric one — I think that if it were implemented it would radically improve our long-run trajectory, but it is also necessarily early, and so what I want more than anything is for you to poke a million holes in it in search of the best possible solution.

We have known the solution for 50 years.

Since fertility dropped below replacement in 1972 — and arguably earlier because of how it structures payouts proportionally to income rather than prices1 — we have known that the Social Security system, as constituted, is not sustainable. This has been a mathematical certainty for over fifty years at this point, and although it has generated enormous amounts of discourse, it has yet to produce substantial action. You can see the degradation of the program in the simple ratio of workers to retirees, which has fallen from 5 in 1960 to only 2.9 today.

No one has ever thought the problem of solving this pension dynamic is, in some way, mysterious, which can make talking about it a little boring.

“Yes, we know it isn’t balanced!”

But it is worth talking about until we solve it! Because there has been a cornucopia of potential solutions. For example, we could have embraced the pay-as-you-go Italian model and just said that the plan for Social Security is to pay out what we can afford. In 1972, when fertility was below replacement but we still had a ton of young people, that meant a relatively large amount of money. Whereas in 2026, when there are fewer young people, it means a relatively smaller amount.

Or we could have invested the surplus in growth-linked assets. For essentially every year until 2010, Social Security brought in more in taxes than it paid out to recipients, but it was statutorily constrained to “invest” that money in Treasuries, effectively parking the money in the general fund. This meant we could spend the money on other programs, like Medicaid or the military, but it also hamstrung the returns the fund fetched.

For a while there was hope that immigration would make the system sustainable. And I do think it postponed the deadline — without the massive migration of the late twentieth and early twenty-first centuries we would probably already have crossed into insolvency. But it didn’t solve anything on a structural level. In fact, it turned the program into a Ponzi scheme of sorts that required perpetually rising absolute numbers of immigrants just to keep the system afloat.

We have tried to implement these fixes before. First, in 1999, Clinton proposed committing budget surpluses to Social Security with a slice invested in equities, through an independent, non-political board contracting private index-fund managers by competitive bid. But pushback from economists like Alan Greenspan over the risk of the government holding private assets killed it. Next, Bush ran for reelection in 2004 on a Social Security reform built around voluntary personal accounts, but the plan died in Congress in 2005, as is so often the case with reforms.

In other words, the problem has never been one of understanding, or lack of academic consensus, or lack of anything like that. The problem has always been the lack of anyone having the power and agency, or, to put it another way, responsibility, to act.

Don’t let academics run firms.

Broadly, I think that we can split most government functions into two categories. The first is engineering tasks, where the objective function is known and we’re optimizing for it under uncertainty. The second is truth-seeking, where the objective function itself is what we’re trying to discover, and there will always be room for fundamental disagreement.

Dan Wang’s Breakneck — which we read for our sprint on problem-solving — basically makes this same argument with his conception of China as the engineering state and America as the lawyerly state. And once you take that presupposition seriously, a lot falls out of it, since these types of organizations are simply good at solving different types of problems.

The two task-types have each evolved their own machinery, designed for their comparative environments. For truth-seeking, the most effective form of social organization looks something like the Talmudic tradition: constant open discussion, everyone sort of pseudo-equals, everything perpetually open to debate and input, where even the “losers’” objections are recorded for posterity. In other words, we never come to a definitive conclusion and everything is always open to discussion.

Engineering objectives instead demand an alternative way of structuring an organization, and you can see it most clearly in the American corporation. What has come out of centuries of Darwinian selection in the corporate environment is something that basically approximates elective monarchy. There is a goal — profit — that is ambiguous and complicated to pursue but easily definable, easily measurable, and hard to game. That last property is the key one. Without committing fraud, a CEO can’t perpetually fake profit, meaning there is basically no way to honestly (or sustainably) misrepresent the outcome, aka “teach to the test,” or fall victim to Goodhart’s Law.

Nothing is legally preventing shareholders from making all the decisions in a company — every choice that crosses the CEO’s desk could be put instead to a majority vote. We don’t do this because it is an obviously absurd system. Some amount of delegation is required, and we have found that finding the best officer, putting them in trust, giving them broad objectives, letting them cook, and assessing them only on outcomes works best.

You can take monetary policy as another example of an engineering task done well. There is a broad consensus on what “good” outcomes from monetary policy are — stable growth and low inflation — but how you actually achieve it in any given country, or any given year, is ambiguous and hard to plan for in advance. So we entrust a single body to manage it independently, subject to review.

There is a large body of academic work supporting everything I’m saying above. The question of why some tasks get open-ended negotiation while others get a hierarchy and a boss is one of the oldest in economics, and the answer that eventually won the Nobel is that tasks get matched to the governance structures suited to their attributes. The same test has been run on government agencies directly: where outcomes are observable, an agency can operate like a firm; but where they aren’t, it can’t.

And so we should be willing to say something that people like me find uncomfortable: I don’t think the lawyers and the academics and the journalists should be in charge of everything. There are areas of human life that are better run by people who are not like Peter Banks. I’d be a decent academic in another universe. But temperamentally, I am not the kind of person who should run ExxonMobil. I would spend all my time theorizing, having open discussions, never really coming to a final conclusion, forever refining and integrating everybody’s perspective to surface as much of the Truth as possible. That is simply not what you need to do to run ExxonMobil. Much of America looks to me like what would happen if people like myself were in charge of everything.

The proposal

Take pensions away from the lawyers and give them to the engineers by chartering an independent Social Security Trust under two plain mandates: keep the system solvent for the long run, and keep retirees’ standard of living as high as that solvency allows.

Here is how it would actually work. Just as now, the Trust collects the payroll taxes — but rather than being forced to park its reserves in Treasuries, the Trust has full agency over how it invests. Critically, it also holds the payout parameters: benefit levels, the retirement age, the formulas.

Congress and the President would provide oversight by appointing the Trust’s board and retaining removal powers, while the Trust would manage all aspects of Social Security finance and benefits other than the payroll tax rate. In governance terms, it would resemble something like the Federal Reserve’s framework — an independent body operating under a statute that sets clear objectives. Similar to the Fed’s dual mandate of maximum employment and stable prices2, the Trust should face a real tradeoff between two competing goals: namely, 1) long-run solvency, and 2) benefit adequacy. This dynamic would optimize for sustainability because a board that maximized solvency alone would simply cut benefits, while a board that maximized benefits without regard to financing would risk insolvency. The advantage of such a framework is that both goals could be measured, tracked, and audited publicly using standard actuarial and program‑performance metrics.

But for any of this to work, the Trust would need to have a surplus, meaning that it is bringing in more in taxes than it pays out — a precondition for building a reserve. Which naturally raises the question of “how?” Ultimately the options would be either to raise taxes, cut benefits, or some combination of both.

If I was in charge, and in this hypothetical I am, I would remove the payroll tax cap, turning it into a flat tax on all payroll income. Or even just raise it to cover ~90% of national wages as it did in 1983, rather than ~83% today.

In addition to this, I would follow the technocratic consensus and cut payouts for high-income recipients by changing the payout formulas. At the end of the day someone has to foot the bill, and wealthy retirees are the ones most poised to foot it. These people are some of the richest in the country and don’t “need” Social Security to survive — I’m thinking here of my other set of grandparents.

It is true that my generation will have to pay twice — covering today’s retirees while pre-funding our own retirement. But alas this is the heavy burden created by postponing a known problem. A burden which will only grow heavier with every year that we delay.

Corporate governance is a solved field.

In America we are culturally primed to prefer the lawyerly arrangement in our government but I would claim that the question of how to run large organizations well has largely been “solved”, at least when outcomes are measurable.

To see why, the place to start is “moral hazard,” because Social Security is itself a textbook case. When people form contracts, the contract can change how the person behaves after signing. If Social Security did not exist, people would save more — this is trivially obvious. Since if you thought that failing to save meant dying in a ditch, or your kids taking care of you, saving would be worth more to you than it would be in a world with Social Security. The classic estimate of the savings displacement was enormous, but probably overstated. The true size is still argued over, but the direction follows from the logic, and even a small effect is reason enough to design the contract well.

In order to address this problem, the solution is writing a better contract. We approximate this already with things like people who earn more, get more; or if you retire early you get less. The whole idea of this is to try and reduce some of the “perverse incentive” created by a public pension. As I said this is a well trodden area.

My PhD program was, nominally, Accounting — “information economics” — and most of what I studied was corporate governance. The running joke in that field is that it is de facto solved. Being into it is a little like arriving in a physics program really excited about Newtonian mechanics. If you want the whole subdiscipline in great detail, the book Corporate Governance Matters is the place to begin.

What that solved field has found is that when a principal must delegate execution under ambiguity, the best possible design is to entrust a single agent with simultaneous power and simultaneous responsibility — full control rights over the organization — while the principals retain complete and total impeachment control, and nothing else. Then the principal should assess the agent exclusively on outcomes. You don’t go to the agent and say: man, how are you doing? You’re a pretty sick guy, you have some interesting ideas. That is the philosopher-academic-lawyer approach to life, and it is precisely how you must not assess an executive. Instead shareholders should be laser focused on if outcomes improved or not, and nothing else.

Two design details fall straight out of that literature. The Trust must never control its own revenue — an agent that sets its own budget maximizes itself — which is why the tax rate stays with Congress. And the theory also explains, with no appeal to stupidity or villainy, why the current owner, Congress, never acts. Since the cost of fixing the pension is concentrated, immediate, and easily traceable to them, while the benefit is diffuse, decades away, and credited to no one, deferral is the only rational strategy today — as it was in 1972, and as it will be right up to the point of disaster.

To reiterate, monetary policy is the clearest case of this model already working in practice. It has been shown again and again that when politicians run the state's printing press, they run the currency into the ground — you can watch it happen wherever they lean on the central bank, as in Turkey, where a stretch of pliant policy bought a short-term sugar high and then inflation touching 85%. Handing that lever to an independent body of technocrats instead — still under the careful supervision of our elected representatives in Congress — is a big part of what has kept inflation stable across the developed world without sacrificing real growth.3

Since pensions have the identical disease, they need a similar cure. Fine for money, you might say. Has anyone actually done it to a pension?

Subscribe now

Canada already did it.

In 1996 the Chief Actuary projected the Canadian pension fund would be exhausted by 2015, and pay-as-you-go costs were marching toward 14% of wages. To avoid this cliff, Ottawa and the provinces struck the deal in February 1997 which effectively solved the problem we are now facing.

Their fix had three parts. First, they raised rates from 5.6% to 9.9% by 2003 — deliberately above pay-as-you-go cost, to pre-fund the plan — alongside modest benefit trims. Second, they set up the CPP Investment Board, which is an arm’s-length Crown corporation with an investment-only mandate — maximum return without undue risk of loss. Third, they made it so that the independent Chief Actuary has to re-certify the plan, and if the numbers slip below an acceptable threshold while politicians can’t agree on a fix, a statutory default fires raising contributions by half the gap, freezing benefit indexation until the next review.

The result has been unambiguously good. With the fund growing from a C$12 million initial transfer in 1999 to C$793 billion in fiscal 2026, earning 8.8% annualized over the past decade; the Chief Actuary has certified the plan is sustainable for the next 75 years; and the contribution rate actually required is now below the rate on the books because of capital gains compounding. From the 2030s onward, investment income will carry a growing share of every pension cheque, lifting that burden off taxpayers. Already this June with the fund running so far ahead of its obligations, Canada cut the base contribution rate from 9.9% to 9.5%, effective 2027.

Canada has split my Trust into three organs, rather than consolidating this into a single agent with comprehensive responsibility, which falls short of my principal-agent ideal but is still much better than what we have today in America.

But won’t it create new problems?

Yes. And it would be naive to pretend otherwise. The question in policy debates is always about balancing tradeoffs — but in this case I think the tradeoffs are visible, and so as a result easily monitorable, in a way that slow-moving solvency dynamics are not. Most objections I can think of essentially collapse into the same point. You cannot depoliticize trillions of dollars in investments.

Here are two versions of the same basic argument.

Politicians will abuse it.

This is Greenspan’s 1999 objection: government ownership of private assets, he warned, would risk sub-optimal capital markets and political direction of investment. He was of course right that the risk is real. But two counter points of sorts. First, we already face this problem on a smaller scale with the federal government’s own Thrift Savings Plan which holds ~$1.1 trillion of federal workers’ retirement money, as well as state and local pensions. Broadly this has produced only limited distortions, to the point it may come as a surprise to you it exists already. Second, the risk of political abuse is a reason for, not against, taking control away from Congress. Since although the Trust has the risk of being politically captured, Congress is by construction a political body. Ultimately if you want to invest a public pension in capital markets, an obvious improvement, any system will face this same problem.

The payout function is too important to leave to technocrats.

Perhaps. Most technocrats like to avoid redistribution questions, instead leaving them to politicians. But Congress has had an exclusive monopoly over this value question for ninety years and has, simply, refused to answer it. An automatic system like the Canadian one would also be fine but in my opinion it would needlessly constrain the Trust’s ability to act. By construction something like the Canadian system still takes power from representatives and gives it to technocrats, but in this case it assigns that power to only the technocrats of 1997, rather than the ones of today.

However, even if they make the best possible set of choices, it is quite possible people will simply not trust them. Personally, I think a Trust that clearly communicates its objectives, and both has and takes real responsibility, could genuinely win the public’s trust. It would, of course, require Congress and the President standing behind it during the initial difficult years, as with any pension reform — but legitimacy of this kind is earned in arrears, accruing to any body that spends long enough proving it can be handed something important and not wreck it. In brief, I suspect it would have the same trust arc as the Fed, which is also forced to make important distributional determinations without regard for political economy.

Conclusion: a calm decade, or the blistering sprint

Like I’ve said, this is an asymmetric idea, though it is mostly a synthesis of programs already running elsewhere. And ultimately, we will do something — because when a pension system finally hits the wall, the options left are ugly: outright insolvency, or hauling in a team of technocratic economists to rebuild the thing in a hurry. It has happened over and over. Italy in 1995, and again in 2011. Brazil in 2019. Greece after the financial crisis. In every case, a neutral, expert body ends up holding the lever anyway — just at the last possible moment, with no room left for anything but austerity.

We are on the same track. If we do nothing, we will hand the redesign to the same people I’m saying should be in charge anyway. The only real choice we actually get is whether we do it calmly, over a decade, with the full menu of options still open — or in the blistering, histrionic sprint, after the cuts have already landed.

I know which option I prefer.

Leave a comment

The honest answer is that almost nothing that would meaningfully bend the debt curve is politically survivable in anything like its effective form. This isn’t a failure of will so much as a structural feature of democratic politics. The costs of serious reform are concentrated and immediate: a farmer loses a subsidy check, a retiree sees a smaller cost-of-living adjustment, a community loses the economic activity generated by an obsolete military base. The benefits are diffuse and delayed: an averted, unseen fiscal crisis, more fiscal room to respond to the next emergency, modestly lower interest rates a decade hence that no voter will ever trace back to the vote that produced them.

Concentrated, immediate losses generate organized, well-funded opposition; diffuse, delayed gains generate approximately no political constituency at all. Layer onto that the fact that politicians operate on the timescale of the next election cycle: they bear the short-term electoral risk of deficit reduction without ever collecting the reward for its long-term benefits.

No wonder our debt-to-GDP ratio has grown on autopilot for the past half-century and now sits at its highest level since World War II. If we want to solve our debt crisis, we must find a way to override the political incentives that produced it and make fiscal responsibility electorally viable — ideally, electorally beneficial.

One caveat before the cutting starts. Some readers will object that the composition of what follows — heavy on spending restraint, light on rate increases — smuggles in a contested premise, namely that consolidating through the spending side is less damaging to growth than consolidating through taxes. If you would like a deep dive into the impacts of spending vs. tax-based austerity, I would recommend our recent article on austerity.

Subscribe now

Assume No Political Constraints

An old joke goes as follows: an engineer and an economist fall into a hole in the ground. The engineer can’t figure out how to escape, but the economist says, “Easy, just assume a ladder.” For the rest of this piece, I’ll do the equivalent of assuming a ladder out by acting as if I’ve been made dictator for a day. In this exercise, I have unlimited authority to raise taxes, cut spending, and pass legislation with no need to win a vote, survive a primary, or ever face the electorate again.

This isn’t an endorsement of dictatorship as a policy-making method; it’s a deliberate thought experiment. If we strip away the political constraints entirely and ask what a purely technocratic fix would look like, we get a useful benchmark. The gap between that benchmark and what’s actually achievable is itself the measure of how hard this problem is.

First Order of Business: Eliminate Corporate and Agricultural Subsidies

As a staunch capitalist, I believe that firms either don’t need subsidies (because they would survive under market competition) or don’t deserve them (because they would not survive under market competition). Therefore, my first act as fiscal dictator is to zero out the $181 billion the federal government spends on corporate welfare each year.

Second Order: Reform and Cut Entitlements

It’s popular (if politically difficult) to get big business off Uncle Sam’s payroll, but it comes nowhere close to closing the gap between spending and tax receipts. The real money is in entitlements, so the second order of business targets the three largest line items in the federal budget directly: Social Security, Medicare, and Medicaid, which together account for close to half of all federal spending and are the single largest driver of the long-run debt trajectory.

Social Security

Three changes, layered together, do most of the work — and it’s worth being clear about which one is the workhorse. As the decomposition in our ground-truths piece showed, the indexing mechanics swamp the demographic lever: how initial benefits grow matters far more to the long-run cost curve than when people start collecting them.

So lead with the formula.

First, then, switch the initial benefit formula from wage indexing to a progressive price index: a blend that indexes benefits for lower earners to wage growth (as today) but indexes benefits for higher earners to the slower-growing CPI instead. Wages tend to outpace prices over time, so this alone bends the long-run cost curve substantially while leaving benefits for the bottom of the income distribution untouched. This is the heavy lifting.

Second — the minor knob, but still worth turning — raise the retirement age and index it to life expectancy going forward, so the program doesn’t have to be re-litigated every time Americans gain another few years of life. Social Security’s full retirement age was set when a 65-year-old had a fraction of the remaining lifespan they have today; letting the age drift upward automatically, the same way we’d want any other actuarial assumption to update, removes one recurring political fight from the calendar even if it does comparatively little of the fiscal work.

Third, layer in means testing at the top: high-lifetime-earners see their benefits taper down enough so that Social Security stops functioning as a middle-and-upper-class annuity subsidized by everyone else’s payroll taxes. None of this touches the program’s core promise for low- and middle-income retirees; it asks the people least dependent on the benefit to shoulder more of the adjustment.

Medicare and Medicaid

The core problem with these programs is that they are open-ended, cost-based reimbursements, which means the incentive on both sides of the transaction points toward more spending, not better outcomes. For Medicaid — the one program actually administered by the states — a block grant flips that dynamic: it caps the federal contribution and pushes states to genuinely manage costs and quality, rather than passing every marginal expense straight through to the federal balance sheet under the open-ended matching formula.

Medicare presents the same underlying pathology in a different institutional wrapper. Because the program is run directly by the federal government, there is no state intermediary to discipline. Moreover, while Medicare Advantage has already moved a majority of beneficiaries into capitated private plans, the federal contribution to those plans remains benchmarked to costs in traditional Medicare, so the open-ended spending commitment persists beneath the capitated surface. The analog to a block grant here is premium support: the federal government makes a fixed, defined contribution toward each beneficiary’s coverage; plans bid competitively against that amount; and beneficiaries who want richer coverage pay the difference. That converts Medicare from an open-ended entitlement into a defined contribution, putting plans — and to a degree beneficiaries — on the hook for marginal costs rather than the federal balance sheet.

Third Order: Tax Shift

Replace the existing income tax system with an X-tax. The X-tax, developed by economist David Bradford, is a form of progressive consumption tax implemented through two linked pieces: a flat-rate cash-flow tax on businesses (revenue minus wages and investment) and a progressive wage tax on individuals’ labor income (with exemptions for income from savings and investments).

Because investment is expensed immediately rather than depreciated over years, and because the tax falls on consumption rather than on the return to saving and investment, the code stops penalizing the choice to save and invest relative to the choice to consume. Because the business side has few deductions to fight over, most of the corporate tax code’s lobbying surface — the accelerated depreciation schedules, the interest deductions, the industry-specific credits — disappears along with the subsidies from the first order of business.

The individual side keeps the progressivity that makes an income tax politically and morally defensible; you’re not moving to a flat consumption tax, you’re moving to one with graduated rates on wages. The appeal of the X-tax is efficiency without regressivity: a flatter, simpler, more growth-friendly structure that doesn’t ask low earners to shoulder proportionally more of the burden.

Fourth Order: Pro-Growth Reforms

Eliminate all trade protections (except those for national security), massively expand migration, and slash regulatory bloat. None of these are deficit-reduction measures in the narrow accounting sense: they don’t show up as a line item cut or a new revenue source.

But debt sustainability isn’t just a function of the numerator (how much we owe); it’s a function of the denominator too (how fast the economy underlying our capacity to service that debt is growing). Tariffs raise consumer prices and invite retaliation while undermining the productivity gains from specialization. A more open migration system expands the labor force and the tax base at a moment when an aging native-born population is already straining the ratio of workers to retirees.1

And regulatory reform lowers the cost of building things — homes, energy infrastructure, factories — which is disinflationary and growth-enhancing in a way that shows up, eventually, in higher tax revenue without a single rate increase.

Final Order: Fiscal Rules

I am only dictator for a day, which means the entire point of this exercise evaporates the moment I hand power back. Cuts, reforms, and a new tax code mean nothing if the next Congress reverses them within a term.

The American track record here is a graveyard of paper promises. PAYGO has been waived whenever it threatened to bind. The Budget Control Act’s spending caps were blown past by bipartisan agreement, repeatedly. Automatic cuts get switched off more or less every time they actually bite. A statutory fiscal rule is exactly as durable as the next majority’s willingness to keep it, which is to say: not durable at all. This is the trap the opening of this piece described, and a clever statute doesn’t escape it: it just relocates it.

The model that has escaped it is Switzerland’s debt brake, which I covered in detail in our eight-countries piece, so I’ll only sketch the mechanics here: a cyclically adjusted ceiling on expenditure — deficits allowed below trend, surpluses required above it — with deviations tracked in a compensation account that must be worked off through lower future ceilings, automatically, with no committee empowered to waive the correction. In force since 2003, it has driven Swiss federal net debt down to roughly 16 percent of GDP, among the lowest in the developed world, with some of the cheapest sovereign borrowing anywhere.

But here’s the thing to take from the Swiss case (and the point the eight-countries piece made that bears repeating): the compensation account is clever engineering, and it is not why the rule holds. The rule holds because it’s written into the Swiss constitution, and because it got there by referendum with roughly 85 percent of the vote.

A politician who wants to bust the Swiss brake isn’t quietly amending a statute; he’s asking voters to repeal something they overwhelmingly and personally ratified. The bindingness is in the mandate, not the formula.

So the last act of my dictatorship is to entrench the rule constitutionally by decree: a cyclically adjusted expenditure ceiling, an automatic compensation account, and a supermajority — not simple-majority — exception clause for genuine emergencies. I’m dictator for a day; I don’t need a referendum, and within the four corners of the thought experiment, decreeing the amendment is enough.

But notice what the decree cannot manufacture. The Swiss brake doesn’t hold because the text happens to sit in the constitution; it holds because 85 percent of voters put it there themselves. A rule imposed by fiat has the words without the mandate, and the mandate is the enforcement: it’s what turns repeal from a routine act of amending a statute into an act of defying the public. That’s the ingredient no dictator can decree, and it’s what the opening of this piece said we were missing: a constituency for solvency. The real-world version of this program has to be built the hard way by persuading an electorate to ratify the constraint, converting the diffuse, delayed benefit of fiscal responsibility into an explicit public commitment that politicians defy at their peril.

Which is the honest place to end the exercise. The economist in the joke assumes a ladder; the dictator in this one can decree everything except the thing that matters most. The real ladder out of the hole is a popular mandate for a binding rule and unlike the assumed one, it has actually been built.

Leave a comment

And where Washington reaches for a reassuring analogy we have pushed back — on the country everyone calls the counterexample, the empire that debt actually dissolved, and the few nations that engineered real fiscal strength — before starting on the harder half ourselves, in what a disciplined cut would actually require and where the defense budget quietly bleeds. But most of that is diagnosis, carried about as far as we can carry it on our own. The way out is the part we have left for you. We've already got a strong batch of submissions, and we're looking forward to reading many more before the July 15 deadline!

As a reminder, the question we are asking for this sprint’s contest is:

Prizes and Recognition

• Grand Prize: USD $2,500 cash, re-publication and promotion on our Substack, and the honorary title of Boyd Fellow for one year.

• Second Prize (Runner-Up): USD $1,000 and re-publication/promotion on our Substack.

• Third Prize (Runner-Up): USD $1,000 and re-publication/promotion on our Substack.

• Additional Publication: We may offer a USD $300 honorarium to authors whose essays we choose to re-publish beyond the top three.

How to Submit

1. Prepare your essay following the guidelines above.

2. Either publish the article on your Substack, or prepare a PDF.

3. Complete the online submission form HERE.

4. Provide your contact details.

5. Link to your Substack post or attach your PDF.

6. Confirm that you have read and agree to the Terms and Conditions.

Full details here:

Also, feel free to reach out via DMs with any questions:

We want Boyd’s funding to come from core supporters who want to be part of the conversation. For that to work, we need people like you to step up now, while it’s still early.

Here’s what we’re offering:

• Supporter ($10/month) and Annual ($120/year): Access to our private Discord community and a quarterly call, where you can talk directly with us and other members

• Founding Partner ($1,000/year): Everything above, plus a monthly one-on-one call with us.

The Boyd Institute is a 501(c)(3). All contributions are tax-deductible. If you’re able to give through a Donor Advised Fund or make a direct donation beyond a subscription, it would meaningfully extend what we can do. Reach out to peter@boydinstitute.org.

Subscribe now

The BRICS nations occupy a genuine middle tier in the global financial system — better insulated than most developing economies, but considerably more exposed than advanced ones to shifts in growth, interest rates, and investor sentiment. A drop in global risk appetite, a domestic slowdown, or geopolitical isolation (as with Russia) can tighten the constraints on any of these countries even when their debt ratios look unremarkable. What determines which side of that line a country falls on is whether sound public finances and durable growth are in place to support it.

Subscribe now

Brazil: Deep Markets, Commodity Exposure, and an Unhedged Windfall

Brazil's debt-to-GDP ratio, which its central bank puts at roughly 80% (IMF says 87%), places it closer to the advanced-economy range than to its developing-country peers. It finances the overwhelming majority of that debt domestically and in its own currency which insulates it from the classic emerging-market trap of foreign-currency debt turning unpayable after a depreciation.

But Brazil’s fiscal story is inseparable from commodities. The early 2010s commodity boom — driven heavily by Chinese demand for iron ore, soybeans, and oil — flooded Brazil with export revenue and fueled a period of fiscal optimism, expanded social spending, and confidence that growth would outrun debt. When prices crashed in the mid-to-late 2010s, that revenue evaporated just as spending commitments remained, contributing to the deep recession and fiscal crisis of 2015–2016.

Part of what made the boom-bust cycle so consequential was Brazil’s approach to its major export industries. Brazil has state control over strategic resources — most visibly through Petrobras’s dominant, state-directed role in oil, including pre-salt reserves discovered in the 2000s — rather than channeling windfall revenue into an insulated savings vehicle. The contrast with Norway is instructive: Norway routes its oil revenue through the Government Pension Fund Global, a sovereign wealth fund that deliberately separates resource income from annual budgets and smooths spending across commodity cycles. Brazil never built an equivalent buffer at the same scale, so commodity windfalls flowed more directly into the budget and into Petrobras’s own investment and debt decisions, leaving fiscal policy more exposed to the next downturn rather than insulated from it.

The result is a country with real strengths — a sophisticated domestic investor base and strong debt management — but high domestic borrowing costs and a fiscal position with comparatively little room for error.

Russia: Low Debt, Foreclosed Options

At roughly 18% of GDP, Russia’s sovereign debt burden is among the lowest of any major economy, financed overwhelmingly through domestic sources, with meaningful fiscal reserves built up over years of energy exports. Yet Russia is the clearest illustration that a low debt ratio doesn’t equal financial strength.

Two forces compound Russia’s actual vulnerability. First, its budget remains heavily dependent on oil and gas revenue, so swings in global energy prices translate directly into swings in fiscal capacity.1 Second, the invasion of Ukraine triggered sweeping Western sanctions that cut off much of Russia’s access to international capital markets, froze a large share of its foreign reserves, and isolated its financial system from the channels it would normally use to borrow or hedge abroad.

A government can have plenty of room to borrow on paper and still find itself constrained by an inability to access the broader financial system. Russia underscores a theme that runs through all of these cases: the relevant question isn’t only “how much debt” but “what options does this debt leave you” — and a combination of commodity dependence and geopolitical isolation can foreclose options just as effectively as overborrowing.

India: Long-Dated Debt Built on a 1991 Foundation

India’s public debt sits at roughly 83% of GDP, financed almost entirely through domestic channels, with very little external sovereign debt. It benefits from a large pool of domestic savings, a deepening bond market, and long average debt maturities that reduce refinancing risk.

Much of this financial architecture traces back to the 1991 balance-of-payments crisis and the reforms that followed. Facing a near-default and dwindling foreign reserves, India dismantled large parts of its license-raj industrial controls, opened its economy to foreign investment and trade, devalued the rupee, and began liberalizing its financial sector. Those reforms set off three decades of faster growth and, over time, the development of a much deeper domestic bond market and a broader base of institutional savers — the very infrastructure that now lets India fund large deficits domestically and at long maturities (weighted average of nearly 13 years) rather than relying on volatile external borrowing.

That history helps explain why India’s debt picture today is one of manageable but not costless debt: a larger, faster-growing GDP base makes debt easier to carry over time, but persistent fiscal deficits and high interest payments continue to consume a significant share of government spending, leaving the trajectory dependent on continued strong growth.

China: A Headline Number That Hides the Real Fiscal System

China’s case is the most complicated in the group because its headline sovereign debt-to-GDP figure (88.3%) captures only a fraction of the relevant fiscal picture. The IMF’s “augmented” measure, which folds in local government financing vehicle debt, puts the real figure closer to 117% of GDP as of 2024 and projects it stabilizing near 150% in the coming years. That gap — nearly 30 points of GDP hidden from the headline number — is the story.

To understand why, it helps to look at how China’s fiscal system actually evolved. The 1994 tax-sharing reform raised the central government’s share of tax revenue from about a fifth to over half almost overnight, mainly through the VAT, while leaving expenditure responsibilities largely unchanged. Local governments still finance roughly 70% of public spending: a structural vertical fiscal gap that every pillar since has been improvised to fill.2

Land sales became the substitute tax system, booked off-budget in separate “government funds” largely outside the headline books. Land-use-rights sales peaked at 38% of local revenue in 2021, then fell by a third by 2023 as the property sector cracked — exposing local budgets in ways the headline debt figure was never built to show.

State-owned enterprises, despite their political prominence, are a surprisingly minor fiscal pillar: SOE dividends to the state were only introduced in 2007, remain low by global standards, and as much as 90% is recycled back into the enterprises rather than into public budgets.

That leaves debt — particularly through local-government financing vehicles, or LGFVs — as the fourth and most opaque revenue source. Local governments use these off-balance-sheet companies to borrow for projects that mostly don’t pay for themselves, with the median LGFV earning about 1% on its assets in 2022 while borrowing at over 5%, kept alive only by an implicit central backstop.3

The whole system was built for an era of high growth, optimized around investment — VAT, land sales, credit — rather than recurring local taxation. It boomed with the hot real estate market but is now buckling under a slower-growth era, with structural deficits already near 6-7% of GDP (what Rhodium Group calls “the myth of China’s fiscal space”) and falling land revenue. The obvious fixes — a property tax (blocked for years), higher SOE dividends, greater central assumption of local costs — are politically foreclosed. The lesson is that sovereign-debt statistics can badly understate real fiscal risk when large quasi-fiscal and sub-national liabilities sit outside the headline number — in China’s case, by roughly 36 points of GDP, and rising.

South Africa: Institutional Strength Built After Apartheid, Tested by Stagnation

South Africa’s debt-to-GDP ratio of approximately 79% is financed mostly domestically, supported by an independent central bank and a genuinely deep local capital market.

Those institutions are themselves a legacy of the post-apartheid transition. The African National Congress government that took power in 1994 inherited an economy isolated by decades of sanctions and underinvestment, alongside enormous pressure to fund redress for apartheid-era inequality in housing, education, and basic services. Rather than running large deficits to finance that redress, post-1994 governments pursued relatively orthodox macroeconomic policy — fiscal discipline, an independent reserve bank, and capital-market liberalization — partly to win the credibility and investor confidence needed to integrate into global markets after decades of isolation. That choice built the deep, credible domestic debt market South Africa relies on today.

But that credibility didn’t hold in a straight line. Looking across the three decades since 1994, South Africa’s fiscal trajectory runs through four distinct phases. The first, from 1994 to 1999, was one of successful fiscal consolidation, as the new government moved to allay earlier fears of populist overspending and got the public finances onto a sound footing. The second, from 2000 to 2008, built on that base with a genuinely healthy fiscal track record, coinciding with the country’s longest economic upswing since the Second World War.

The third phase, from 2009 to 2020, was effectively a fiscal storm: the global financial crisis hit revenue just as “micro fiscal populism,” the state capture era under the Zuma administration, and eventually the Covid-19 pandemic pushed spending and borrowing sharply higher. It was during this period that South Africa lost the investment-grade credit rating it had spent years earning, as its foreign-currency government bonds were downgraded to junk status by Moody’s in 2020.

This was followed by a fourth phase of renewed consolidation, marked by a shift away from relying on growth alone to cure fiscal imbalances and toward direct fiscal reform.

The Common Thread

Across all five countries, debt composition matters more than the debt ratio itself, and history shapes that composition as much as current policy does. Brazil’s commodity windfalls went largely unhedged rather than into a Norway-style buffer. Russia’s energy dependence and post-2022 isolation matter more than its low debt ratio. India’s domestic financial depth traces back to the reforms forced by its 1991 crisis. China’s local governments inherited a structural revenue gap from 1994 that land sales and LGFV debt have only papered over. South Africa’s institutional credibility was built deliberately after 1994, and is now being tested by the growth it needs to sustain it.

None of the five faces the classic developing-country crisis triggers — heavy external borrowing, foreign-currency liabilities, weak revenue systems. But none has the unquestioned credibility that lets Japan or the United States carry far higher debt loads without investor anxiety. They sit in the middle, and the details of how each got there are the real story.

What separates the more resilient cases from the more fragile ones, in the end, comes back to the same two fundamentals: sound public finances and the growth to sustain them. Where both hold — as in India’s deepening bond market — debt becomes easier to carry over time. Where either falters — Brazil’s unhedged commodity exposure, Russia’s foreclosed financial options, China’s hidden local-government liabilities, and South Africa’s economic stagnation — even a modest debt ratio can mask real fragility.

Leave a comment

The first inflation target in the world came out of New Zealand at the end of the 1980s. After the Reserve Bank of New Zealand Act took effect in 1990, New Zealand became the first country to write an explicit inflation target into law. The early agreements between the finance minister and the central bank governor set a band around 0-2 percent, later widened to 0-3 and eventually 1-3 with a 2 percent midpoint, meant to represent “true” price stability plus or minus about one percentage point to account for the fact that measured inflation runs a bit hot. Those ranges reflected a desire for low inflation and some allowance for index mismeasurement, but they were not the output of a carefully solved social‑welfare optimization problem, and subsequent changes were driven as much by politics as technocracy. That alone is reason enough to question the way the number has hardened into an object of faith.

And when you interrogate it honestly, the case against an evergreen two percent target turns out to be fairly strong. In fact, within the monetary policy literature the center of gravity has been drifting toward the view that there are much better ways to run policy than by fixing a single low number and treating it as sacred rule. The unfortunate reality, however, is that at the current juncture, the United States is ill-positioned to switch.

Subscribe now

The origins of, objections to, and alternatives for, two percent inflation targeting

Canada, Britain, and Sweden followed New Zealand within a few years of its inflation target taking effect. The Federal Reserve was the great holdout: Alan Greenspan governed by instinct and constrained discretion, implicitly aiming at “low and stable inflation” but refusing to nail down a figure. It was only in 2012, under Ben Bernanke, that the Fed finally committed in writing to a two percent target “as measured by the annual change in the PCE price index,” making the anchor that disciplines the most important monetary authority on earth younger than the smartphone.1

To be sure, the reasons offered for two percent are real. A small inflationary cushion keeps interest rates far enough above zero that the central bank has room to cut in a downturn. Measured inflation runs a little hot because price indexes cannot fully capture how people substitute, trade up, and benefit from quality improvements, so true price stability sits slightly below the measured figure. And a bit of inflation greases the labor market because it lets real wages fall without anyone having to cut a nominal paycheck.

Each of these rationales is sound, but they are more so general reasons for a low positive number rather than for two in particular — as opposed to, say, one and a half or three. In reality, the number is a rudimentary convention (dressed as a calculation) that we’ve grown to accept. But because the world has genuinely shifted underneath it, the convention is aging poorly.

Here’s how. The neutral rate of interest (r*) — i.e., the rate at which monetary policy is neither contractionary nor expansionary, i.e., the short-term interest rate that would prevail when the economy is at full employment and stable inflation — has fallen for decades in advanced economies, dragged steadily down as the population has aged, fertility rates have plummeted, and productivity growth has stayed somewhat sluggish.

The problem is that a fixed two percent inflation target, which might have made sense when neutral rates were in the low‑to‑mid single digits, now guarantees more frequent collisions with the floor, meaning that the Fed starts every downturn closer to the zero lower bound (ZLB) and runs out of conventional policy-response room faster.2

It is on this basis that economists like Olivier Blanchard and Larry Summers have been arguing (for years now) that four percent is simply the better number. And more recent formal work has shown that in many mainstream models, as the neutral rate falls, the optimal inflation target tends to rise with it, recovering most of the decline, though not all of it. This is true because the cost of repeatedly hitting the ZLB rises while the marginal cost of somewhat higher steady‑state inflation remains comparatively small.3

There are also cleaner designs than an inflation target at all. The most elegant is to target the level of spending in the whole economy rather than the change in a price index — nominal GDP targeting, argued for years as a favorable option by and the market monetarists, and developed in serious detail by economists at Mercatus.

NGDP-targeting fixes a single deep flaw in price targeting: an inflation-targeter reacts to prices without asking what is moving them. When prices move for reasons that have nothing to do with demand, it pushes policy the wrong way — and which wrong way depends on the shock. A productivity boom that drives prices down reads as below-target inflation, so the inflation-targeter eases, loosening into a bubble.4 A supply shock that drives prices up while output collapses reads as above-target inflation, so it tightens, deepening the slump. A spending-targeter is fooled by neither, because it tracks total spending rather than prices: it does nothing in the first case and holds steady in the second.

In short, the spending-targeter stops fighting the real economy and starts accommodating it — and on the merits alone, NGDP‑targeting is probably the better system.5

One could go further still and tie the goal to something structural like the growth of the working population or the growth of productivity so that the target being optimized for maps to the economy in its current state rather than the state it was in a generation ago. I have not seen anyone build a rule that writes those structural trends into the target itself, and it would be something new. In effect it would bake the determinants of the neutral rate directly into the policy rule instead of leaving them as background estimates. But the instinct behind it seems sound (to me at least).

At any rate, on the technical merits, the critics of two percent inflation targeting are right: a static number governing a moving economy is a strange and suboptimal way to run the most important price in the world. It is largely arbitrary within a narrow band of “low positive numbers,” too low for a world of chronically depressed neutral rates, and too rigid to adapt to persistent changes in demographics and productivity.

So if this were only a question of monetary engineering, it is rather apparent that the answer would be to retire the existing target and build something smarter — be it higher long‑run inflation targets or some alternative convention.

But is it simply a question of monetary engineering?

No… at least not entirely.

There is also a serious opposing camp, including former Fed chairs Bernanke and Yellen, the economist Frederic Mishkin, and others, who worry that higher inflation targets can pull expectations loose from their mooring. Central banks have spent decades teaching the public that “two percent” essentially just means price stability; move the goalposts and you risk losing the price anchor itself.

The more basic worry is that inflation is not just a statistical outcome but a behavioral one. So long as expectations stay anchored — so long as the public believes next year’s prices will behave roughly like this year’s — a central bank can carry a little inflation without it compounding. But expectations are fragile. Let them come un‑anchored and a modest, well‑meant adjustment to the target gets read as permission for something larger: if households expect prices to be higher tomorrow, they spend sooner today, and at the macro level, the price-expectation loop begins to feed on itself. In other words, the hazard of moving the goalposts is that the act of moving them signals to everyone that the number can move.

That is why the Fed’s pandemic-era framework change was so risky even though it was modest on paper. In August 2020 it rewrote its own framework and promised to let inflation run above two percent for a while to make up for the years it had run below. This came to be known as flexible average inflation targeting, or FAIT. The idea was defensible on paper: after a decade of undershooting and a sunken neutral rate, a planned overshoot of two percent to achieve a longer-run average made sense empirically. But within a year inflation had surged well beyond what the new framework had contemplated, driven by a combination of supply shocks, fiscal stimulus, and accommodative monetary policy all at once. The promise to tolerate overshoot then became a serious credibility problem for the Fed (and frankly an embarrassment it spent years living down).

Formally, the language of FAIT survived into the mid‑2020s. But in practice, by the time inflation had been wrestled back down, the Fed returned back to something closer to plain-vanilla inflation targeting and symmetric concern for too much or too little. Overall there was at least one lesson from this episode that is pretty straightforward: even if the old framework was not perfect, a new rule is easy to announce but very hard to keep control of once shocks arrive and expectations start to move. And the part that should worry us is that FAIT was a modest reform attempted in a storm that turned out to be passing. The next reform attempt, if tried now, would be amidst a storm that is not so — dare we say — “transient.”

Political incentives make changing the Fed’s targeting framework right now risky

Consider the structural and rapidly-worsening situation the US finds itself in now. The country is not in a war, not in a deep recession, and unemployment remains low. Yet it is running a deficit of roughly six percent of national income, debt held by the public is closing in on the record it set at the end of WWII as a share of GDP on current projections, and net interest is approaching the trillion‑dollar mark — already larger than defense and Medicare — leaving only Social Security above it in the budget. This is an almost unprecedented combination.

Furthermore Treasury data show that on the order of one‑third of marketable debt is due to mature within a single year and well over half within three, meaning that every sustained climb in yields feeds straight back into the deficit that produced it.

Why the looming debt‑and‑deficits storm reshapes the inflation‑targeting reform calculus comes down to political forces. A government borrowing — more specifically, borrowing to refinance — at this scale has an unspoken but powerful interest in cheaper money: a few years of rates held lower than they should be, and thus a little more inflation, means that the real weight of the debt eases on its own.

To capture this opportunity would mean the inflation target stops being a technical dial the Fed turns at will and instead becomes something the Treasury has a stake in. That is the edge of fiscal dominance: a regime in which the fiscal authority, not the central bank, sets the path of inflation — a regime where “monetary policy serves fiscal ends.” So the risk to reopening the framework right now would be handing a fiscally cornered government say over the one number it most wants to move, and for reasons that have nothing to do with sound monetary policy. The unsettling question is whether we are already drifting toward it.

Despite abounding incentives, the answer is that we are not there yet, at least according to the market’s reading. Treasury yields, term premia, and inflation breakevens may be showing slight signs of distress over the past few years, but they are still consistent with a world in which the central bank retains the upper hand — and where inflation expectaions remain relatively anchored.

The Fed still sets policy to hit its target and the bond market has not begun to price anything like a fiscal-dominance regime. If Fed independence were genuinely broken, the signals would likely show up in a much sharper backup in bond yields, an unusually high term premium, and a generally far more disorderly Treasury market than anything we have seen recently.

None of this means the country is out of the woods regarding fiscal capture of monetary policy. According to the logic Sargent and Wallace set out in 1981, fiscal dominance happens less through a watershed moment than through a gradual regime shift a nation drifts into. When deficits run large enough for long enough, natural forces nudge the central bank toward no longer acting as if the budget belongs to the fiscal authority. And the clearer warning is not the headline debt ratio but the share of the budget swallowed by interest. As that share climbs, so too does the risk of higher rates doing damage and thus the political payoff of paring Fed independence — potentially setting off a harmful, self-reinforcing cycle. These leanings, it should be said, may already be underway.

The president has demanded deep rate cuts and has been aggressive in trying to impose costs for Fed noncompliance. In August 2025 the administration moved to remove a sitting Fed governor, Lisa Cook, over allegations about mortgage paperwork from before she joined the Board — conduct with no bearing on her FOMC vote, and for which she has not been charged. The Supreme Court heard the case in January 2026 and signaled it was unlikely to let the removal stand. Trump’s Justice Department also went as far as opening a criminal investigation into former Fed chair Powell himself in early 2026 (over his congressional testimony about the Fed’s headquarters renovation) before dropping it months later without bringing any charges. While these are not the symptoms of a regime that has already flipped, they are the political pressures that, if maintained over time, do flip it.

It is also worth being precise about this channel, because explicit adjustments to the targeting framework are not the only door in. Political capture of monetary policy does not require reopening the target at all. Along these lines — “personnel is policy” — adjustments to the Fed’s policy frameworks can run straight through appointments and pressure, which is what the Cook and Powell fights were really about.6

And there is another door, still blunter than either of the above. Congress could simply rewrite the Fed’s mandate. It is reaching for that lever already through the Price Stability Act of 2025 (H.R. 5396), which is now moving through the House and would strip the Fed’s employment goal outright, making its inflation target even more central to monetary policy. Granted, under current conditions, this particular bill points the hawkish way. But the same power that can narrow the mandate can loosen it, and a Congress cornered by its own financing needs is the most direct fiscal-dominance vector of all — one that routes around the Fed completely.

Underneath it all ultimately sits the age-old temptation a government deep in debt faces. Nixon succumbed to it, or at least tried to. In lieu of defaulting, cutting spending, or tax hikes, a government, once fiscally-dominated, can simply let inflation run a little hot, hold rates a little low, and watch the real value of the debt effectively melt away. Economists including call this financial repression:

“[A] catch-all term for all of the things governments, central banks, and financial regulators can do to lower the government’s cost of funding below the market rate. You can think of it as an informal tax on bond holders and savers. Over the past few decades, we’ve seen governments in advanced economies reach aggressively for financial repression tools in the face of increasing fiscal pressures as a result of demographic changes.”

Japan is the most prominent recent example, but it is also basically how the US took its debt down from over 100% of GDP after the war to under a quarter of it by the end of the 1970s (repression coming via surprise inflation plus pegged and regulated rates).7

But that escape required a captive bond market: capital controls, long-dated debt, and savers with nowhere else to put their money. None of those requisites hold today. Roughly a third of the debt is held by foreigners who cannot be repressed, about seven percent is indexed to inflation outright through TIPS, and the average bond now matures in about six years — which means the debt reprices almost as fast as it could be inflated away. To move the debt now, inflation would need to be high enough and sustained enough to outrun a market that virtually reprices in real time.8

So the reality is that the great liquidation of the postwar decades is no longer on the menu, even if the temptation to follow that playbook remains. Furthermore none of this is to mention that, given the hyper-salience of “inflation” and “affordability” as public issues in the current climate, it is just not a very politically rational way out.

Conclusion: the right reform, the wrong moment

The question of whether a better target than two percent inflation exists is fairly settled (it does). The unsettled question is whether the US has earned the right, and whether it is fiscally positioned, to reach for it now.

The “Volcker Shock” of the early 80s — when then-chairman Paul Volcker changed the Fed’s policy stance in the middle of a double-digit inflation crisis — offers a clean test. It worked because instead of reformulating the Fed’s mandate, he adjusted its policy regime to deliver the prevailing mandate even harder. He was willing to go as far as tipping the economy into recession by driving interest rates toward twenty percent, all to prove the anchor would hold, bucking what was politically expedient in service of reasserting institutional credibility.

So it is not simply that reform from weakness is impossible. Another example is Brazil, which adopted inflation targeting in 1999 in the teeth of a currency crisis and made it hold, pairing it with a hard fiscal-responsibility law the next year. But Brazil reformed because catastrophe left it no choice and forced discipline alongside the new rule. Today the US has the opposite problem: Fed credibility is often nowadays taken as a given, the dollar’s reserve currency status buys the government enough rope to reform badly, and no external creditor is forcing it to pair a new target with restraint.

Every framework change that strengthened the anchor was, at bottom, a move toward discipline. The world adopted inflation targets in the calm of the 1990s for this very underlying reason: to lock credibility in. What is being entertained now — loosening the anchor to relieve a financing problem — is the mirror image. Rewriting the rule to make a promise harder to break is fine and warranted; rewriting it to make the debt easier to carry would come with harsh second-order effects. And a reform undertaken in this fiscal weather would be the second kind, no matter how carefully it was dressed up as the first.

There is, to be fair, a real truth the reformers are reaching for, and it deserves to be granted. The inflation an economy can comfortably bear really does depend on how fast it grows, so the instinct to tie the rule to nominal output or something macroeconomically similar is not foolish. But you do not get the inflation a fast economy can carry by editing the target. You get it by… growing fast. Thus productivity is the one thing that actually changes what is sustainable, and no act of the central bank can conjure that into being.

None of this is to pretend that waiting is free. Leaving the target where it is for longer carries a real cost, and it is precisely the cost the first half of this essay laid out: a two percent anchor in a low‑r* world means more frequent collisions with the zero lower bound, weaker recoveries, and a central bank too often out of room — a failure of the employment half of its mandate as much as the inflation half (the Japan case is instructive here). To wait is to keep paying that price. And however nontrivial that price is, the alternative price — handing a cornered Treasury a credible-sounding reason to inflate — is larger, and far harder to reverse. A lost recovery may eventually be regained while a debased anchor may not.

So the honest conclusion is the uncomfortable one. Two percent is arbitrary, a better system exists, we know its name, and yet we should still leave it alone. A country in command of its budget could rewrite the rule and be the better for it. But the US — a country running massive deficits at full employment and above-target inflation — has lost the standing to act on it.

Upending monetary policy frameworks now would be like choosing a new referee in the middle of the game we are losing: in all likelihood, it would turn a bad situation worse. This is not an argument for never, but it is an argument for sequence. The moment to reach for a better target is when the conditions that make reaching for it dangerous have eased — e.g., when the deficits have normalized, when debt has stopped climbing as a share of the economy, when the interest bill no longer crowds out everything around it. Inflation-targeting reform belongs to that moment, not to this one.

Leave a comment

Norway and Singapore derive their fiscal strength from accumulated wealth — oil in Norway’s case, decades of compulsory savings and sovereign investment in Singapore’s. Denmark and Sweden demonstrate that generous welfare states can be compatible with fiscal discipline, but only after rounds of market reform and pension restructuring, and only when financed by broad-based taxation rather than redistribution alone.

Luxembourg and Ireland are cautionary as much as exemplary: their headline finances look impressive, but both live off a concentrated windfall, and unlike Norway and Singapore they spend it rather than save it, building permanent commitments on revenue that a change in tax policy or corporate strategy could redirect. Switzerland and Germany are instructive cases for large advanced economies, though for different reasons and through meaningfully different paths. Switzerland benefits from structural economic advantages that, while not as dramatic as North Sea oil, are nonetheless distinctive. Germany is the more instructive example of fiscal discipline achieved with fewer natural tailwinds.

Governments with high levels of asset holdings, such as Norway and Singapore, can be much more solvent than their headline debt-to-GDP figures suggest. At the same time, windfalls — whether from oil, favorable tax geography, or financial services — should be treated as temporary and saved rather than spent. And fiscal rules, when credibly designed and politically sustained, can secure a country’s fiscal solvency for the long term — though, as Germany’s 2025 constitutional exemption illustrates, even the most durable rules can bend under sufficient investment and geopolitical pressures.

Subscribe now

Norway

Norway has one of the largest and most comprehensive welfare systems in the world, but also a dynamic economy and one of the best fiscal positions of any nation on Earth. Many commentators attribute Norway’s ability to fund a welfare state without sacrificing growth or fiscal discipline to the nation’s immense oil wealth.

Indeed, while Norway’s gross government debt is 54% of GDP, its net public wealth remains positive because its financial assets dwarf GDP. What is unique about Norway among oil economies is not that it gives oil wealth to its citizens1, but that it places a strict 3% ceiling on how much of the oil fund can be withdrawn each year and invests the rest in a well-managed sovereign wealth fund.2

Singapore

On paper, Singapore carries a surprisingly high debt-to-GDP ratio for a country synonymous with fiscal conservatism and technocratic governance. In reality, while Singapore’s debt-to-GDP ratio is 171%, most is held to manage reserves and fund compulsory savings rather than to finance deficits.. Like Norway’s, Singapore’s government holds assets well in excess of public debt and other liabilities.

The engine behind Singapore’s reserve accumulation is the Central Provident Fund (CPF), a mandatory savings scheme that functions as the country’s primary social security system. All Singaporean workers and their employers are required to contribute a portion of wages into individual CPF accounts. These funds are then used by the government, which issues Special Singapore Government Securities (SSGS) to the CPF at guaranteed rates, effectively borrowing the population’s forced savings. The proceeds are passed to GIC, one of Singapore’s sovereign wealth funds, for investment in global markets.

This mechanism is a large part of why Singapore’s gross debt figure appears elevated: the debt is not a sign of fiscal distress but rather the liability side of a balance sheet that is more than offset by investment assets. In this sense, Singapore’s “natural resource” is not oil or minerals but the institutionally captured savings of its workforce, intermediated through the state and deployed into global capital markets.3

Singapore’s constitution allows up to 50% of expected long-term real investment returns from reserves to be spent through the Net Investment Returns Contribution, and the state uses these revenues to fund approximately 20% of annual government expenditure. This allows Singapore to maintain competitive tax rates while still financing world-class public services.

Denmark and Sweden

Like their oil-rich Scandinavian neighbor, Denmark and Sweden combine generous welfare states with strong growth and fiscal solvency. But this was not always the case. During the 1970s, both countries expanded public spending beyond what their economies could sustainably support, accumulating deficits. To change course, both nations (Denmark in the 1980s and Sweden in the 1990s) implemented pro-market reforms, spending restraint, and switched to defined-contribution pension plans.

The breadth of their tax bases is the part Americans miss. Denmark and Sweden do not finance their welfare states by taxing the rich — by most measures their tax systems are less progressive than America’s. Both lean on 25% value-added taxes that fall across the income distribution, and their top income-tax rates bite at little more than the average wage compared to the US, where the top bracket doesn’t begin until roughly eight times average earnings.

The ordinary Scandinavian worker hands over a far larger share of his paycheck than his American counterpart. The equality Americans admire is real, but it is produced on the spending side — through universal benefits and services — not by soaking high earners on the tax side. That is the uncomfortable lesson: Americans want Scandinavian services and have shown no appetite for the broad-based taxation, consumption taxes above all, that pays for them. The Nordic bargain is simply “tax everyone, and spend it well,” not “tax the rich.”

Luxembourg and Ireland

Norway and Singapore are not the only countries living off a concentrated windfall. Luxembourg and Ireland are too — the difference is what they do with it.

Luxembourg’s exceptionally strong sovereign balance sheet — with government debt of about 30% of GDP — rests on one of the world's most concentrated financial-services sectors, an industry that exists in large part because of its favorable tax and regulatory regime. Ireland's case is starker: it ran a budget surplus of approximately €12.4 billion in 2025 on corporation-tax receipts of about €33 billion, the bulk paid by a handful of large American multinationals booking global profits through Irish subsidiaries.

But where Norway and Singapore bank their windfalls and spend only the return, these two spend the windfall itself. Luxembourg runs its financial-sector revenue through the general budget, and Ireland — despite launching a Future Ireland Fund in 2024 — routes the large majority of its corporate-tax take, by its own fiscal watchdog’s estimate nearly 90%, into permanent commitments. Building lasting obligations on a windfall is a defensible bet only if the revenue is broad and durable. Ireland’s is neither — nearly half rests on the decisions of just three major foreign firms.

Germany

Germany represents perhaps the most instructive case for large advanced economies precisely because its fiscal strength was built not on resource windfalls or favorable tax geography, but on institutional self-discipline.

The rules and their origins

Germany’s Schuldenbremse, adopted in 2009 and fully binding on the federal government from 2016, caps the federal structural deficit at 0.35% of GDP. Alongside the “black zero” political commitment and a strong economy, it underwrote budget surpluses from 2014 through 2019 and cut public debt from roughly 82% of GDP to just below 60% before the pandemic intervened.

A nontrivial result of this is that German government bonds remain the euro area’s preeminent safe asset, a status that reflects deep market confidence in the country’s fiscal institutions.

Germany’s path to fiscal rules was shaped by reunification4, which required an estimated €2 trillion in fiscal transfers from west to east over three decades and structurally elevated debt through the 1990s. Combined with the low growth of that era — Germany was widely called the sick man of Europe by the late 1990s — these pressures produced the deficit overruns that became a political embarrassment: Germany breached the EU’s 3% ceiling in 2002, 2003, and 2004, making it one of the first prominent violators of the Maastricht rules it had designed. The Schuldenbremse was the response, intended to make recurrence constitutionally impossible.

The 2025 break

In early 2025, Germany’s newly formed government secured a constitutional exemption allowing substantially higher borrowing for infrastructure and defense. A €500 billion off-balance-sheet infrastructure fund was established alongside a carve-out exempting defense spending above 1% of GDP from the Schuldenbremse’s constraints — a response to years of underinvestment and the changed security environment following Russia’s invasion of Ukraine. The Schuldenbremse held for fifteen years and demonstrably shaped German fiscal behavior, but it bent under the combined pressure of geopolitical necessity and accumulated investment deficits. Whether 2025 proves a one-time correction or the beginning of a looser fiscal regime is the central open question in German public finance.

Switzerland

Switzerland's fiscal strength rests on both structural advantages and institutional discipline, making it a more complex case than it first appears, and one that sits somewhere between the pure rule-based model of Germany and the favorable-geography models of Luxembourg and Ireland.

Tax geography and the financial services sector play a meaningful role. Switzerland’s cantonal tax system creates internal competition that keeps rates low and has historically attracted both corporate headquarters and ultra-high-net-worth individuals. The country’s financial services industry — anchored by global private banking, asset management, and insurance — is outsized relative to the population and contributes substantially to government revenue.5

Equally significant is Switzerland’s long-standing role as a destination for wealthy European expatriates seeking relief from the higher tax regimes of neighboring countries. The lump-sum taxation regime (forfait fiscal) lets wealthy foreigners who do not work in Switzerland be taxed on notional living expenses rather than their actual worldwide income or assets — a provision that has drawn billionaires, executives, and high-net-worth retirees from France, Germany, Italy, and the UK.6 It is neither universal nor untouched — five cantons, including Zurich, have abolished it and the federal regime was tightened in 2016 — but it survives federally and in most cantons, giving Switzerland a revenue base that countries inside the EU's tax-harmonization framework cannot easily replicate.

None of this is the whole story, though, because Switzerland also bound itself with a rule. A property-bubble collapse in the early 1990s dragged the economy into prolonged stagnation and pushed federal debt sharply higher; the resulting pressure of a stagnant economy and ballooning debt7 summoned a broad reform coalition, and voters ratified a constitutional debt brake by referendum in 2001. In force since 2003, it has driven federal net debt down to roughly 16% of GDP (among the lowest in the developed world) and delivered some of the cheapest sovereign borrowing anywhere.

So Switzerland is neither a pure story of self-discipline nor a pure story of luck. The structural advantages are real, even if less dramatic than North Sea oil; but the debt brake extended the discipline well beyond what those tailwinds alone would have produced.

Takeaways

Several takeaways emerge from this survey. First, headline debt statistics are frequently misleading. Norway’s gross debt understates its strength; Singapore’s dramatically overstates its weakness. What matters is the full balance sheet — assets as well as liabilities — and the quality and sustainability of the revenues that service them.

Second, windfalls should be treated as temporary until proven otherwise. Norway, Ireland, and Luxembourg all benefit from revenue streams that could diminish or disappear. The countries that have managed such windfalls most successfully are those that saved them rather than built permanent spending commitments around them.

Third, and most importantly for policymakers in larger democracies, fiscal discipline is primarily an institutional challenge, not a technical one. Switzerland and Germany did not stumble into strong public finances; they engineered them through constitutional rules that made profligacy harder and created political cover for restraint.

Similarly, Denmark and Sweden chose fiscal sustainability, at some political cost, after experiencing the consequences of not doing so.

The United States faces none of the insurmountable obstacles that might make these lessons inapplicable. It has no shortage of administrative capacity, no absence of credible institutions, and no lack of economists who understand what a sustainable fiscal trajectory looks like. What it has lacked, at least so far, is the political will to impose constraints on itself before markets demand it.

Of course, none of these analogues map cleanly onto the US — in terms of real economy and political economy — but the lessons from the countries examined here suggest that certain constraints, once adopted and sustained, can compound over time into something extremely valuable: fiscal space that provides security in crises, credibility in markets, and genuine fiscal optionality that heavily indebted nations do not tend to have.

Leave a comment

The evidence, in fairness, has looked pretty compelling. A sovereign that borrows in a currency it prints took gross debt past 240% of GDP and held it there while suffering functionally none of the predicted consequences. There was no funding crisis, no default, no inflation spiral, and ten-year yields on government bonds floated persistently near zero. Blind faith in monetary orthodoxy led many traders to short JGBs in what became known as the “widowmaker” trade. Politically, Japan’s record has made debt hawks in Washington’s hand-wringing over a debt ratio half Japan’s size look like pure theater.

We have also been critical of MMT, even if it’s impossible not to concede much of the theory’s core premises. The central descriptive claim — that a monetary sovereign does not face the hard external-funding ceiling that constrains a household, a firm, or even a euro-zone member — is not wrong. Unlike the Soviet Union or the graveyard of other sovereigns whose fiscal crises precipitated their nations’ undoing, Japan’s debt is almost entirely held by its own citizens and domestic institutions. This means that the risk of a classic sovereign default, where a country runs out of foreign currency to pay its foreign creditors, is effectively nil.

But sovereign default is not the only risk vector and Japan’s record was never evidence of deficits being costless. In this piece, we argue instead that it was evidence of a specific machine running under specific conditions, that these rather goldilocks-esque conditions have meaningfully shifted, and that Japan isn’t as much the counterexample to fiscal anxiety as it is a fairly strong case for it.

Subscribe now

The Machine

The conventional fiscal-anxiety framework begins with Sargent and Wallace's 1981 “Some Unpleasant Monetarist Arithmetic,” which showed that monetary financing eventually becomes a tax on bondholders via inflation. Japan long looked like the exception. Headline and core inflation ran near zero or below 2% for most of the period after the 1990s asset bust, and even after the BoJ formally adopted a 2% CPI target in 2013 it repeatedly undershot that goal — the expected tax never showed up. Yet former IIF chief economist has characterized Japan as “where MMT goes to die.” Exactly how it is dying is an instructive case for the developed world's fiscal hawks.

Let's start with the underlying "machine," which involved three moving parts:

1. A central bank with an inflation target it couldn’t hit;

2. A fiscal authority that took advantage of the resulting low‑rate environment to borrow heavily at near‑zero nominal yields; and

3. A captive, predominantly domestic, investor base — banks, insurers, pension funds, and eventually the BoJ itself — that absorbed most of the issuance.

An aging population and chronically weak demand kept underlying inflation pressures subdued, giving the BoJ room to expand its balance sheet aggressively. Rather than demonstrating that debt and deficits are irrelevant, under the cover of chronic deflation, Japan’s experiment amounted to a prolonged attempt by the BoJ to lift inflation to target via large‑scale asset purchases (QQE and yield‑curve control) that made its balance sheet, as a share of GDP, far larger than the Fed’s at its QE peak.

This was highly convenient for the fiscal side. As long as inflation stayed below target and long‑term yields remained compressed, the government could roll over and expand its debt stock at minimal interest cost, while the central bank absorbed a growing share of outstanding JGBs through secondary‑market purchases. Put simply, the government issued while simultaneously the central bank absorbed.

What the BoJ didn't absorb directly, the captive domestic savings pool did. Regulations had channeled household deposits into low-yield accounts and the postal system: upwards of 90% of JGBs have been held by domestic institutions and households for most of the past two decades.1

makes the sharpest version of the reading. In the market-monetarist frame, the stance of monetary policy is defined by outcomes-versus-target, not by instrument settings themselves. Perhaps counterintuitively, low rates and a massive balance sheet are actually consistent with a tight monetary stance in a situation where the central bank keeps missing its inflation target to the downside.2

By that measure, the BoJ ran a relatively hawkish policy regime for three decades — whatever the face value of its instruments might have indicated, and however stable markets remained.

No Crisis Is Not The Same As No Cost

What the dovish reading misses, even on its own evidence, is what the dormancy actually cost. Japan didn’t disprove the inflation constraint so much as find the one macroeconomic environment in which it lies dormant — chronic, structural deflation — and live there for thirty years. The cost of doing so was real, even if it wasn’t paid in the way the dovish framework would predict.

The fiscal expansion that ran alongside the monetary one — in Sumner’s own words, “one of the most reckless in human history” — was reckless precisely because it failed to stoke aggregate demand. Monetary offset kept demand neutralized in real time, so the deficits bought very little in macro terms, not even inflation.3 For an MMT framework that locates demand in fiscal policy, this must constitute the model failing on its own metric. Indeed, the one period of decent NGDP growth — Abenomics — came when Japan combined aggressive monetary easing with a shift back toward consolidation, including consumption-tax hikes in 2014 and 2019, rather than the ever‑looser fiscal stance MMT advocates.

What the deficits bought instead was time. Japan paid continuously in foregone growth: for much of the period since the mid‑1990s, its nominal GDP has been essentially stagnant, with the level in the early 2000s about the same as in 1995 and even in 2025 still below its mid‑1990s peak in dollar terms.

So the honest framing here is that the debt ratio climbed less because the numerator exploded than because the denominator refused to grow. Three decades of calm bond markets were three decades of monetary policy successfully suppressing what fiscal policy was trying to do — while the debt stack piled up. As Sumner has put it:

“I was told ‘deficits don’t matter’ because interest rates were low. I pointed out that rates might not stay low forever, and that these debts would still be on the books when interest rates started rising.”

The familiar “Lost Decade” story does a good job of naming the symptoms, but the underlying dynamic gets much less attention.

Such an environment produced an economy where capital found no productive call worth answering, and what got starved was the productive base. Real wages went sideways for decades; corporate Japan, sitting on the deepest cash piles in the developed world, rationally declined to deploy them. Unsurprisingly, the Nikkei spent thirty-four years climbing back to its 1989 peak, creating a generation of equity investors who got less than nothing in real terms. Meanwhile balance sheets stayed pristine and markets non-volatile, while the central bank was left without the capacity to answer shocks and household purchasing power eroded under a currency the authorities had limited room to defend and ultimately allowed to slide.4

Now, none of this argues away demographic forces. An aging, shrinking workforce was never going to support 4% growth no matter what the BoJ targeted, and any honest account has to start there.

But the argument here is incremental: conditional on those structural drags, the policy arrangement let the productive base hollow much further than it had to — by suppressing the price signals (rising yields, a weakening currency, modest inflation) that would have pushed capital into more productive use. Former IMF chief economist Ken Rogoff puts the counterfactual — Japan without the bubble, the messy crisis aftermath, the policy errors — at roughly 25-50% richer per capita than the actual one. So yes, demographics did a lot of the work. But policy probably did a lot of it too.

Crucially, what began as the BoJ’s failure to hit its target became, over time, the architecture the state built around. Chien, Du, and Lustig (2025) gives the cleanest recent statement of what that architecture was: a sovereign wealth fund built on cheap, borrowed money. Yen-denominated household deposits, effectively held captive, were what shouldered the costs of exorbitant JGB issuance. In other words, Japan’s private households and financial institutions financed a leveraged equity portfolio — a hedge fund of sorts — running a carry trade that belonged entirely to the state. Another way to think of this is that Japanese domestic savings were effectively channeled into funding the productive economic bases of other countries, while Japan’s own went unfunded (or at least, underfunded).

By the mid-2020s, Japan’s government-owned stocks and foreign investments topped 100% of GDP. Less than half of that portfolio was in domestic equities, with the rest in foreign assets — US stocks, Treasuries, and other sovereign bonds.5

The financial-repression literature (Carmen Reinhart’s work is canonical) is pretty straightforward on this. Sustained negative real rates against captive demand quietly liquidates the national debt. Ken Rogoff, Reinhart’s frequent co-author on this work, calculates these “very large losses on domestic savers through a combination of low real yields, currency weakness, and financial repression” to be on the order of 3–4% of GDP per year in many historical episodes.

To be sure, this liquidation cost is ultimately being paid by Japanese households, whose stock‑and‑bond participation still hovers around the low‑20s percent, versus roughly 60% of US households owning stock in some form.6 And so the much-admired price stability came in the form of an expensive tradeoff. A generation of suppressed dynamism was facilitated in exchange for basically never having a single bad week in the bond market. This tradeoff has played out even as the currency's safe-haven status meaningfully eroded, which is where we turn next.

The Cover Is Gone

After thirty years of fighting falling prices, the conditions that made the machine work have now reversed. What Japan (theoretically) wanted — for inflation to normalize — has arrived: wage settlements above 5% for the third consecutive year, the strongest run since the early 1990s, and headline CPI that peaked above 4% in early 2023 and hovered above 2% for much of 2023–25 — partly imported through the weak yen coupled with rising global energy prices, partly a result of the domestic wage-price dynamic the BoJ spent two decades trying to summon.

Without the deflationary cover, the BoJ cannot print to suppress yields without feeding the above-target inflation it must now instead contain. So it has been forced to pivot: it ended negative rates and abolished yield-curve control in March 2024, raised its policy rate to 0.75% by December 2025, then to 1.0% just this week — the highest since 1995 — and began tapering its large-scale asset purchases.

The instinct here would be to take cues from the bond market — from JGBs — as to the manner in which the monetary pivot is being digested by market participants. But that instinct is wrong (or at least incomplete). JGB yields, as a market signal, are a noisy and probably muted read on Japan‑specific distress, because the BoJ still holds just under half the JGB stock and remains an important buyer at the long end.7

So, if the inflation constraint does not bite via higher yields on government debt, where exactly is it binding?

Robin Brooks’s central insight — that the constraint may alternatively take the form of currency depreciation, which can run out of control when debt gets too high — is useful here. He has adamantly maintained that a sovereign’s risk premium does not vanish when its central bank caps yields. Instead, the risk premium relocates, because yield caps plug just one of several release valves, “merely transform[ing] what would have been a bond market crisis into a currency crisis.”

The higher-signal thing to watch here is therefore the currency, and Japan’s currency is indeed behaving as though a crisis is already underway.8

Importantly, a weaker yen is a major source of anxiety for the Japanese consumer, who imports a great deal of everyday goods like energy and food. As the yen has cratered post-Covid, consumer spending in Japan has only just recently clawed back to around its pre‑Covid level in real terms:

Worse yet, the bond market is still leaking because, muffled as the market may be, price discovery persists.9 The forty-year yield punched above 4% in January 2026, the first 4%+ print on any super-long JGB in roughly three decades, and the May 2025 twenty-year auction drew the weakest demand since 2012.

The Ministry of Finance has thus responded by trimming super-long issuance and shifting toward shorter tenors, the classic move of a sovereign under funding stress when the pain of higher yields starts to flow through.

Nevertheless, on the fiscal side the Takaichi government — in office since October 2025 having run on subordinating the primary-balance target to growth — is pressing the accelerator, with a record ¥122.3 trillion budget financed by ¥29.6 trillion in fresh issuance. The fiscal-2026 budget assumes a 3% funding rate, the highest in twenty-nine years, with debt-servicing costs at a record ¥31.3 trillion, over a quarter of total expenditure. (Market practitioners named the ‘Takaichi trade’ months ago: stronger Nikkei, weaker JGBs and yen.)

But what the MMT-friendly “decades of calm” reading blatantly misses about exorbitant debt and duration is that the pain also arrives even when holding yields constant. As the ever-rising debt stock reprices, each maturing low-coupon bond is replaced by a higher-coupon one causing debt-servicing costs (interest payments) to rise even if yields stop climbing.10

It is telling that the central bank has stopped pretending. In a February 2026 speech, BoJ Policy Board member Hajime Takata attributed the rise in long-end yields to “a rise in the risk premium, known as the term premium.” This is, in no uncertain terms, the monetary authority acknowledging that the jig is up the market is now pricing fiscal risk into Japanese debt.11

Honest engagement with the dovish rebuttal survives all of this. An economist at Dai-ichi Life puts the rebuttal cleanly: in an era of inflation, rising debt-service costs come bundled with rising nominal GDP and tax revenue, so a bigger interest bill is not automatically a solvency problem. The IMF makes the same point structurally — nominal growth that exceeds the effective interest rate is precisely what has pulled Japan’s ratio down from its 258% pandemic peak toward 230%, even as gross debt stays the highest among major economies. This is also the Blanchard point, and it is correct as arithmetic.

A parallel rebuttal runs on the debt itself: net of the government's vast financial-asset holdings, Japan's net debt drops to around 130% of GDP. Consolidate the BoJ's JGB holdings as the intragovernmental claim they economically are, and the figure falls to on the order of three-quarters of GDP — lower than the US on the same measure.

Furthermore, Sumner’s “never reason from a price change“ rule says that rising stocks beside rising yields signal stronger expected growth, not panic, and Japanese thirty-year yields still sit more than a hundred basis points below the US counterpart. On the narrow question of whether Japan is experiencing something like the 2022 gilt rout, he is right — it unequivocally is not.

But the favorable arithmetic is a harder bet than it sounds. It requires nominal growth to durably exceed the marginal funding rate as the entire debt stock reprices, and Japan’s structural real growth runs near half a percent against a long-end now yielding 3 to 4%. Brooks’s framing cuts underneath the arithmetic entirely: a monetary sovereign can run out of fiscal space long before inflation forces the issue, because the constraint shows up first in the currency and the term premium rather than in a missed payment.

Markets, as Brooks wrote in January, are therefore telling Japan that it has run out of fiscal space. And the cost is very real, even if it gets paid through subtler channels than a sovereign-default headline would suggest — through captive savings, an eroded currency, the financial repression of negative real returns on deposits. Others have described this kind of erosion of creditors’ claims as a “stealth default.”

So while the dovish case on debt-and-deficits isn’t necessarily foolish or misinformed, it is unresolved. Japan, given the ambitiousness of its monetary experimentation, is doubtlessly where it gets resolved over the next several years. The harder question, and the one that matters for Washington, is whether the structural conditions that let Japan run the experiment in the first place are at all replicable.

Why Washington Can’t Copy Tokyo

Strip out the short-term rates set by central banks, and the long end of the yield curve is telling a similar story across the developed world at once: the same term-premium widening shows up in the United States, the United Kingdom, France, even Germany, the safe haven of old. None of this is unique to Japan in direction, only in degree.

Brooks’s frame is that this is a single synchronized event — the world used up its fiscal space during the pandemic, and markets now demand to be paid for holding long-dated debt everywhere. Japan is simply the leading edge, with a central bank that ran the suppression experiment longest and is unwinding it first, which makes it the diagnostic case. The question for Washington is not whether it faces the same debt-repricing pressures — it does — but whether it faces those pressures in the same way as Japan does.

The conditions, however, diverge sharply, for three main reasons.

1) What made Japan’s debt sustainable (for as long as it was) doesn’t translate to the US.

Japan financed thirty years of deficits out of a vast domestic institutional balance sheet: on the order of 90% of its government debt has been held at home, by patient, price-insensitive institutions — the BOJ, life insurers, banks, the postal system — with yen liabilities and nowhere to flee. Formerly the world’s largest net creditor for most of the period, running persistent current-account surpluses throughout, Japan lost its top-creditor spot in 2024.12

Now, price-sensitive marginal buyers exist where there predominantly used to be captive domestic ones: foreigners are now roughly 65% of monthly cash JGB turnover. In short, the conditions that made Japan’s exception possible — including a thirty-year deflationary window — are rapidly closing.

Crucially, the captive domestic creditor base is the single most important fact about what made Japan’s tolerance for debt so high for so long, and it’s just not something the US can manufacture. The US is the mirror image: the world’s largest net debtor, running structural current-account deficits, financing itself from abroad with foreigners holding a substantial share (albeit down to nearly a third) of marketable Treasuries.

2) The standard rejoinder is that the US has something better than domestic savers: the dollar.

The reserve currency commands a deeper, more durable bond‑demand base than the yen ever did: the dollar still accounts for close to 60% of global FX reserves versus only a few percent for the yen. The dollar’s “exorbitant privilege” is real, but it doesn’t rest only on market depth or network effects. A large majority of official dollar reserves and other “safe” U.S. assets are held by countries inside the U.S. security perimeter, and Eichengreen, Mehl, and Chiţu (2019) shows that military alliances boost the dollar share of a partner's reserves by about 30 percentage points. In a 2025 follow-up, the same authors estimate that credibly withdrawing those alliances would cause roughly 6% of US marketable public debt to be liquidated and long-term rates to rise by up to 80 basis points — adding around $115 billion a year to Treasury interest costs."

Thus a significant part of the reserve premium — over and above depth and network effects — is the alliance system. Pulling on the security architecture is pulling on the ballast of America’s borrowing position. The dollar’s role also brings its own asymmetric failure mode. Because it anchors global reserves, trade invoicing, and cross‑border funding, a disorderly dollar adjustment would not only hit US consumers via higher import prices, it would also propagate as a global shock through funding and credit channels.

Japan, by contrast, can endure a slow, grinding, decades‑long yen depreciation (painful for its own households but far less systemic) precisely because the yen is a much smaller reserve and funding currency.

3) Interest on the debt in the US isn’t merely a “distributional issue.”

Let’s assume that Mosler is right and that rising interest costs on Japan’s debt are merely a distributional issue — meaning that those higher interest payments amount to government transfers to domestic asset-holders (who skew higher in wealth and income). Whereas the vast majority of Japan’s public debt is held by domestic creditors, this would not hold true for the US, whose public debt is held by a significant contingent of foreign investors. In the US, instead of merely being a distributional issue, a meaningful portion of such debt‑servicing costs are transferred to bondholders abroad rather than recycled within the domestic private sector.

Final Takeaways From Japan

The right lesson from Japan is neither the dovish one nor its hawkish caricature — not “we have decades, relax,” nor “we are one auction from a Greek crisis.” Japan financed its reckoning in foregone growth, a captured monetary policy, and a currency now doing the adjusting, and it is making the payments in a compressed window with a government still pressing the accelerator.

Sumner makes the monetarist version of the warning — that the recent run of 6%-of-GDP US deficits arrived in unusually good years, with no recession, war, or pandemic. To read that calm as sustainable is to “miss the dog that didn’t bark.”

That leaves a narrow menu. Monetization is what Japan exhausted; it works until inflation returns, then inverts into the currency crisis now playing out. Consolidation exists on paper and dies in politics (in Tokyo and Washington alike) for reasons Buchanan and Wagner identified: democracies overspend on benefits that accrue to identifiable voters while the taxes are diffuse and deferred. Cutting those benefits concentrates pain on groups that vote, so the rational political move is to just avoid the pain to ultimately avoid the blame. In other words, the off-ramp exists but the politics necessary to take the off-ramp don't.

What then remains is the only exit that is neither austerity nor a currency-and-inflation event. And that is growing the denominator (GDP) faster than the bond market reprices the numerator (the debt). Productivity, as we’ve argued previously, is the only plausible path that makes the fiscal arithmetic survivable. Importantly, it requires investing in productive capacity underneath the debt stock rather than leaning further into mandatory spending.

Leave a comment

8

Brooks highlighted this interesting comparison: in real effective terms, the yen fell below the Turkish lira in April 2026. That is a developed creditor nation’s currency — considered for decades by global investors as the safe-haven currency — trading, in real terms, below a currency tied to one of the world’s most fiscally disastrous regimes in recent history.