With the release of the Better Care Reconciliation Act, the Senate has finally shown its cards. Much will and should be made of how the bill, if adopted, would deprive millions of people of health insurance in order to finance huge tax cuts for the wealthy.

But the bill is troubling in another respect: It affords states the immense, hidden power to eliminate some of the Affordable Care Act’s most critical financial protections.

How does it accomplish this? Nestled in Obamacare is a provision allowing states to seek waivers from some of the law’s most important rules, including rules requiring insurers to cover certain “essential” health benefits and imposing annual caps on out-of-pocket spending. The point was to allow states to experiment with alternative approaches to achieving the goals of health reform: improving coverage and lowering costs.

That’s why the waivers haven’t been given out like candy. Under the Affordable Care Act, a state has to show that its alternative plan would allow it to cover as many people, with coverage as generous, without increasing federal spending. Only if a state met that demanding triple standard could it take the money allocated for the ACA and chart its own course.

The Senate bill requires the secretary of health and human services to say yes to state experimentation of virtually any kind

The Senate bill retains this waiver provision — but removes the guardrails that ensured state-based alternatives would offer strong coverage. Under the Senate bill, to get a waiver, a state doesn’t have to demonstrate anything about coverage. Instead, it just has to show that the plan won’t “increase the federal deficit.” Once a state makes that showing, the bill is explicit: The secretary of health and human services “shall” approve the plan.

Not “may” approve the plan — “shall.” This is a crucial legal distinction. The Supreme Court has squarely held that this sort of mandatory language means what it says: If the condition is satisfied, the secretary has no choice but to give his approval.

That could lead to some bizarre consequences. What’s stopping a state from submitting a half-baked plan for a high-risk pool that will lead millions of people to lose coverage? Or, for that matter, from using Obamacare money to fund public schools or affordable housing? According to the Senate bill as written, the secretary would have to approve plans like that so long as they don’t increase the federal deficit.

True, the bill says that a state must submit a “description” of how its plan would “take the place” of the rules it waives, including the “alternative means” that the plan will use to increase access to coverage, reduce premiums, and increase enrollment. But the description has no legal bearing on whether to grant the waiver. Again, if the state’s plan doesn’t increase the deficit, the secretary has to approve it. Period.

The bill lets governors concoct new health care plans, without involving legislatures

The bill goes further to grease the wheels for waivers. Under Obamacare, a state had to pass a law in support of the proposed waiver, which meant legislatures had to give their approval before the state could experiment with novel approaches to health insurance. The Senate bill would cut legislatures out of the equation. Governors, together with their insurance commissioners, could devise new health care plans on their own. (They would only have to “certify” that they have the authority to implement the plan — a low hurdle.) The federal government must then review the waiver request on an expedited basis.

And once a waiver is granted, the Senate bill says that the federal government cannot terminate the waiver, no matter what. It is hard to overstate how unusual — even unique — this is. When the federal government offers money to states, it places conditions on how the states are to use that money — and reserves the right to cut off the states if they fail to adhere to those conditions. The cutoff threat is essential to prevent state abuse of federal funds.

The Senate bill removes that threat. It says that a waiver “may not be cancelled” before its expiration. If state officials blow the Obamacare money on cocaine and hookers, there’s apparently nothing the federal government can do about it. At the same time, the bill expands the duration of waivers from five years to eight years. The upshot, then, is that the next president won’t be able to renegotiate any waivers granted during the Trump administration, no matter how badly a given state might have abused its waiver.

So any state that wants a waiver can get one; federal funding can be used for virtually any purpose; and the federal government has no power to discipline states that abuse the federal money.

Which parts of the Affordable Care Act could be waived? The Senate bill would retain some limits. It wouldn’t, for example, allow states to waive the prohibition on discriminating on the basis of preexisting conditions. But it would allow states to remove caps on out-of-pocket spending for exchange plans. Under Obamacare, an individual with insurance can be asked to pay a maximum of $7,150 out of pocket for her health care; for families, the cap is $14,300. If those caps are eliminated, the sky is the limit on what you might be asked to pay.

Goodbye, essential benefits?

What’s more, states could exempt themselves entirely from the rule requiring coverage of the essential health benefits — a change from the bill the House enacted last month, which required states to identify an alternative list of essential benefits. Without that rule in place, many insurers on the exchanges are likely to drop their coverage for maternity services and for the treatment of substance use disorders, among other benefits.

Waiving the essential health benefits rule won’t directly affect most of the health plans that people get as a fringe benefit of employment. Under Obamacare, the rule doesn’t apply to health plans from large employers. But a waiver would have enormous indirect effects on employer-sponsored coverage.

Here’s why: In one of its most popular provisions, the Affordable Care Act prohibited all health plans, including those offered by employers, from imposing annual and lifetime limits on coverage. These limits were common before the advent of health reform, and had especially harsh consequences for people, including children, with debilitating chronic illnesses, as Vox’s Sarah Kliff has movingly documented.

The ban on annual and lifetime limits was thus a boon even for the 159 million people who get their coverage through their employers. But the waivers contemplated under the Senate bill could wipe out the ban on such limits. Under the Affordable Care Act, the ban on lifetime and annual limits applies only to those benefits that are considered “essential.” If a state gets a waiver saying that no benefits would be considered essential — which was the status quo prior to Obamacare — insurers could impose lifetime and annual limits on all services.

If one state weakens health insurance rules, that could have ripple effects in every state

It gets worse. In general, federal law attempts to create a single, nationwide set of rules for large employers that offer health coverage. That’s why, as Matt Fiedler at Brookings has pointed out, “current regulations and guidance permit large employer plans to apply any state’s definition of essential health benefits for the purposes of determining the scope of the ban on annual and lifetime limits.”

So imagine that Texas were to get a waiver from the essential health benefits rule. Employers in every state could then adopt Texas’s definition of what counts as essential — which is to say, nothing. A single state could thus wipe out annual and lifetime limits across the whole country, making a mockery of federalism, where each state is supposed to be able to make decisions for itself.

Although the Trump administration could tighten its regulations to mean that Texas’s waiver would apply only to employers in Texas, it isn’t likely to do so. Even if it did, waivers would apply within any state that chose to grant them — meaning that the Senate’s bill would still expose millions of families that get coverage through their employers to financial catastrophe.

In short, the waivers available under the Senate bill are breathtaking in scope: not just “big waiver,” as some lawyers call the emerging system that lets states opt out of government rules, but “waiver unlimited.” It’s not clear that the Senate appreciates what it’s doing; indeed, it’s not even clear that rules governing waivers can be included in a reconciliation bill.

But if the Senate parliamentarian decides that they can be included, and the Better Care Reconciliation Act of 2017 becomes law, its waivers will be a backdoor method for undoing some of Obamacare’s most popular and significant protections.

Nicholas Bagley is a professor at the University of Michigan Law School. Find him on Twitter @nicholas_bagley.

The Big Idea is Vox’s home for smart discussion of the most important issues and ideas in politics, science, and culture — typically by outside contributors. If you have an idea for a piece, pitch us at thebigidea@vox.com.

You’ve heard it time and again: The Affordable Care Act was a federal takeover of the health care system, an overweening piece of legislation that President Barack Obama shoved down the throats of a balky public on a party-line vote. For years, Republicans have deployed themes of federal arrogance and overreach to underwrite their attacks on health reform.

Now the Republicans have an alternative, the American Health Care Act. Part of its appeal, they say, is that it returns authority to the states. Speaker of the House Paul Ryan’s earlier template for the legislation makes the case concisely: The states “should be empowered to make the right tradeoffs between consumer protections and individual choice, not regulators in Washington.”

Ryan’s assertion raises an implicit challenge, one that Republicans themselves have not wholly thought through: Why bother with national health reform at all? At the core of our federal system is the principle that the states should retain control over their own affairs unless there’s good reason for Washington’s involvement. Yet the ACA is a federal statute, and the progressive push for health reform has had a doggedly national focus. Even the Republican “replacement” stops well short of total devolution to the states. Why?

The question is more challenging than health reform’s supporters generally acknowledge. There’s a kernel of truth to the claim that Obamacare needlessly impinged on state prerogatives. But the question does have an answer. Because of underappreciated legal constraints on their ability to tax and spend, the states can’t go it alone. The federal government really does have to take the lead.

A more refined understanding of the need for federal action serves as a rebuke to those who claim Congress can just wash its hands of health reform. And it offers a yardstick against which to measure the new American Health Care Act, which would leave intact the very obstacles that have long prevented the states from tackling reform on their own. Unless Republicans change their approach, their talk about federalism should be seen for what it is: empty rhetoric that masks a refusal to allow any level of government to achieve near-universal coverage.

State-by-state reform might have more effectively built public support for change

Strictly as a strategic matter, the campaign for national health reform needs some defending. On at least a half-dozen occasions during the 20th century, federal reform efforts had gone down in flames. Massachusetts’s coverage expansion in 2006 seemed to offer a more promising path: Maybe supporters of reform should rack up some wins in the states before tackling the issue at the national level.

By way of comparison, consider the march to same-sex marriage. When the Supreme Court invalidated state bans on same-sex marriage in Obergefell v. Hodges, fully 35 states already allowed it. Before Obamacare, in contrast, only two deep-blue states had significantly expanded coverage, Massachusetts and Hawaii. Judged solely on that metric, the progressive commitment to a national solution seemed premature.

Nor was there a self-evident functional case for federal intervention. It makes a lot of sense for the federal government to act, for example, when states can impose costs — “externalities,” in the economic jargon — onto other states. Federal environmental laws, for example, aim to prevent states from sending their air or water pollution to their neighbors.

But concerns about externalities can’t justify health reform. If New York declines to adopt near-universal coverage, that doesn’t hurt Connecticut or New Jersey. The country can tolerate a patchwork of state insurance rules. Indeed, because the federal government has generally stayed out of the business of regulating insurance, it already does.

Nor are states locked in a race to the bottom that prevents them from embracing health reform. We might see such a race develop if, for example, a state’s adoption of a coverage expansion led sick people to flock to the state. If that happened, no state would want to be the first mover, even if most of the states would prefer to expand coverage.

But the data tells us that sick people rarely move in response to coverage expansions. In a 2014 study, two Harvard researchers examined migration patterns in response to Medicaid expansions in four states. They found “no evidence of significant migration effects.” A similar 2016 study estimates “that the migration effect of Medicaid is very close to zero,” a finding echoed in other research.

Other arguments for federal intervention also fare poorly. In some circumstances, national action is thought necessary to prevent states from discriminating against historically disadvantaged minority groups. Think here of the Voting Rights Act, which is supposed to stop states from jerry-rigging their voting rules to dilute the political power of blacks or Latinos.

Is there a civil rights case for federal oversight of health care?

It’s hard — not impossible, but hard — to see health reform as analogous to a civil rights statute. Yes, members of minority groups benefit disproportionately from coverage expansions. And yes, states might resist expansions partly because of insensitivity to minorities’ interests. But the same can be said in many other policy domains. Take education, for example. Many states can (and do) rely on property taxes to finance local schools. Wealthier communities get well-funded schools; minority communities, not so much.

Appalling as that might be, it’s not generally thought sufficient reason for a federal takeover of public schools. Outside the enclave of federally protected rights, the risk that states might behave badly has generally not been thought an adequate justification for national action. If it were, lingering discrimination would be reason enough to oust the states of their authority to tax and spend, spelling the end of federalism in any meaningful sense of the word.

When it comes to financing health care, the states are hamstrung

So why do we need federal health reform? Here’s the answer: money.

During a recession, the federal government can run a budget deficit. Most economists think that’s a good thing because federal spending can stimulate the economy and bolster the safety net. States don’t have that luxury: They’ve all got to balance their budgets every year. (Vermont is the lone exception.)

When a recession comes, state tax revenues take a hit at precisely the moment that lots of people lose their jobs and their insurance. (Indeed, revenues take a hit because of the job losses.) To cover the newly uninsured, states will have to spend more, requiring the imposition of new taxes or cuts to other state spending, either of which will exacerbate the recession. Coverage expansions thus commit states to an economic policy that could harm them during downturns.

This is what I call the countercyclical trap, and it’s the single biggest obstacle to state-level reform. Here, Massachusetts is the exception that proves the rule. When it expanded coverage, it had one of the lowest rates of the uninsured in the country, which reduced the size of any new fiscal obligations. And, with Sen. Ted Kennedy’s help, Massachusetts cut a sweetheart deal with the Bush administration to get an extra $1 billion in Medicaid funding to support Gov. Mitt Romney’s vision for conservative health reform. Massachusetts could thus manage — barely — to expand coverage. Other states without its advantages can’t afford to.

The problem runs deeper. A state that’s looking to expand coverage can always ask its taxpayers to foot the bill. But many of those taxpayers will complain, with some justice, that it’s unfair to ask them to bear the whole burden. If you get health coverage through your job, you already face a reduction in take-home pay commensurate with the value of that coverage. Your friend who works at a similar job but doesn’t get health coverage is paid somewhat better. Should you both be taxed equally in order to subsidize insurance for the higher-income worker?

From the state’s perspective, it’s both easier and more equitable to adopt a law penalizing businesses that fail to offer insurance. These pay-or-play laws have a clear political logic: Employers should live up to their end of the social bargain. They have a certain economic logic, too — an employee who starts getting coverage because of a pay-or-play law will see an offsetting wage reduction. That employee will thus “pay” for his own coverage, reducing the need for broad-based tax increases.

The trouble is that the Employee Retirement Income Security Act of 1974, or ERISA, disallows state laws that “relate to” the design of employee-benefit plans, including health insurance. Although there’s some legal ambiguity, ERISA means that the states probably can’t adopt pay-or-play laws.

Here, Hawaii’s experience is instructive. Its scheme for near-universal coverage depends on a robust pay-or-play law. That law is still on the books only because Hawaii got Congress to adopt a special exemption from ERISA. Without a similar carve-out, the other 49 states will find it hard to achieve near-universal coverage.

The compromise: Washington provides the money, states the policies

Because of the countercyclical trap and ERISA, federal money must be the lifeblood of health reform. At the same time, however, the weakness of the other justifications for national intervention implies a rough division of responsibility: The feds should finance the bulk of any coverage expansion, but the states should be allowed to adopt their own distinctive approaches to health reform.

To a large extent, the ACA honors that division of responsibility. Its subsidies and the Medicaid expansion are financed almost entirely with federal money. The states retain authority to regulate their insurance markets and were given the option of establishing their own health care exchanges. The Obama administration also proved willing to grant broad waivers to states seeking exemptions from Medicaid’s rules.

In some respects, however, the ACA is stricter than it needs to be. Take, for example, the rule requiring insurance plans to cover all the “essential health benefits.” It sets a comprehensive baseline package protect consumers, but it also makes insurance more expensive, which can put it out of reach of some of the very people who need it most. The tradeoffs are genuinely hard.

Why not let states to choose their own essential benefits package? You shouldn’t reject the idea just because you think some states will make stupid choices. The point of federalism is that states get to make their own choices, stupid or not, unless there’s a good reason for federal intervention. As the Yale professors Jerry Mashaw and Ted Marmor argued 20 years ago, “[t]here is no agreed-upon ‘best’ health insurance (or medical care) system that a state could offer.” What if a policy that seems stupid turns out not to be so stupid after all?

The same goes for many of the ACA’s insurance reforms: age banding, coverage of preventive services, even the individual mandate. To be sure, the federal government must establish guardrails to prevent the sale of junk insurance and to assure that states have a plan for achieving near-universal coverage. But there’s a lot to like about giving states more flexibility.

The ACHA doesn’t do the job right

Republicans may talk the talk of devolving health care policy to the states, but that’s not what the American Health Care Act does. Instead, it starves health reform of the funding upon which it depends.

Most significantly, Republicans intend to phase out the Medicaid expansion and to impose a hard cap on federal contributions. If a recession forces a state to exceed its cap in a given year, any overruns will come out of its Medicaid payments the following year. With that kind of shortfall, the states will have to make savage Medicaid cuts to make ends meet.

Republicans also want to slash the subsidies that make insurance affordable in the private market. Under the ACA, no one making less than four times the poverty level has to devote more than 10 percent of her income toward private coverage; most pay much less. The American Health Care Act would erase that affordability guarantee and, instead, extend age-based subsidies that would be much too meager for most people to afford coverage.

If federal money is withdrawn, states will be stuck. Because of the countercyclical trap and ERISA, they won’t be able to enact and sustain coverage expansions on their own. The end result will not be the diversity that federalism celebrates. It will be a uniformly crappy system that leaves millions of the sick and poor without coverage.

It doesn’t have to be this way. A group of Republican senators led by Bill Cassidy (R-LA) and Susan Collins (R-ME) has floated an alternative, the Patient Freedom Act of 2017, that retains the ACA’s funding streams while giving the states more room to choose how to use that money. That’s a model that deserves serious attention from both Republicans and Democrats. It might enable partisans on both sides move past the rancorous debate over the ACA.

For now, however, the Republicans seem intent on dismantling coverage gains across the entire United States. Their proposals trade on the rhetoric of states’ rights, but they would have the perverse effect of inhibiting state power. That’s bad for federalism — and bad for the country.

Nicholas Bagley is a professor at the University of Michigan Law School. This piece was adapted from his essay, “Federalism and the End of Obamacare,” in the Yale Law Journal Forum.

The Big Idea is Vox’s home for smart, often scholarly excursions into the most important issues and ideas in politics, science, and culture — typically written by outside contributors. If you have an idea for a piece, pitch us at thebigidea@vox.com.

This month, Kellyanne Conway, senior adviser to President Donald Trump, was asked whether the administration would refuse to enforce the Affordable Care Act’s individual mandate — the requirement that people get health insurance or pay a penalty. “He may,” she said, instantly sending a shiver of fear down the spines of health reform’s supporters. Without the mandate, insurance markets in many states will teeter; some will probably collapse.

Would it be legal for Trump to decline to enforce the mandate?

The short answer is no. The longer answer is more complicated, but it’s also instructive. At key points, President Barack Obama delayed aspects of the ACA in an effort to put health reform on a sound footing. The delays were classic examples of executive overreach; they never should have happened. The Republican-led House of Representatives even sued the president over them.

And now the shoe is on the other foot. With Trump in office, some of Obamacare’s fiercest critics seem almost giddy at the prospect that he might use the same weapon against the act. Pick your favorite tagline: payback is a bitch, what goes around comes around, “I learned it from watching you, Dad!”

What Trump has hinted at, however, would be a far greater overreach than Obama ever attempted. But Obamacare’s critics are unlikely to care. There’s an important lesson here about the accretion of executive power in the 21st century, how law is enforced outside the courts, and what presidential power might look like in an age of Trump.

Obama used executive power to delay an unpopular part of the ACA

The story begins in the summer of 2013. Three years earlier, Democrats had pushed the ACA through on a party-line vote, only to watch Republicans retake the House of Representatives in a midterm election that President Obama called a “shellacking.” The reform bill was slated for full implementation in 2014, but public support was eroding in the face of unremitting Republican hostility.

To dampen the opposition, the Obama administration decided to delay the implementation of the employer mandate for one year. This mandate, which penalizes employers that don’t provide their employees with health insurance, along with the individual mandate, was among the most vilified provisions of Obamacare. “In our ongoing discussions with businesses,” wrote Valerie Jarrett, senior adviser to the president, in announcing the delay, “we have heard that you need the time to get this right. We are listening.”

The decision was front-page news. It was also controversial. Under the Constitution, the president has a duty to “take Care that the Laws be faithfully executed.” Could the president really disregard the ACA’s instruction about when the mandate would take effect?

To defend the delay, the administration argued that the Internal Revenue Service had made a habit, over three presidential administrations, of extending “transition relief” from new tax statutes. That relief gave taxpayers a chance to make the business adjustments necessary to calculate and pay the tax. Since Congress had never objected to the practice, the argument ran, Congress must have thought the delays didn’t constitute the faithless execution of the law; its acquiescence blessed the practice.

This wasn’t a strong argument. The previous delays were usually for days or weeks, at most six months, and concerned narrowly targeted taxes without a lot of revenue at stake. Sure, Congress didn’t object. But that’s probably because it didn’t notice or didn’t care enough to make a stink about it. It was fiction to think that Congress had approved the open-ended power to delay tax statutes. If there were any doubt about that, its fury over the delay of the employer mandate should have dispelled it.

But the Obama administration went ahead anyway. No one immediately sued, probably because no one had standing to object that someone else wasn’t being taxed. (When the House of Representatives later attempted to challenge the delay in court, its challenge was promptly dismissed for want of standing.) Emboldened, the administration delayed the mandate for a second year for midsize employers.

A second aggressive executive move by Obama: fixing the “if you like your plan, you can keep it” mess

Then came a second crisis, and a second presidential intervention. In the fall of 2013, people started getting insurance-cancellation notices, contradicting President Obama’s oft-repeated promise that “if you like your plan, you can keep it.” The ACA had established baseline rules about what counted as minimal coverage, and some of the old plans were now illegal. So insurers had to drop them.

Under intense pressure from an outraged public — and from Republicans making hay of the broken pledge — the Obama administration came up with an “administrative fix.” State insurance commissioners, who have front-line responsibility for overseeing health insurance, were told that they didn’t need to apply the ACA’s new rules to old health plans. At first, the fix delayed the ACA for 10 months; the administration later extended that by another two years.

The law was clear that these plans could not be sold. To defend this delay, the administration invoked its enforcement discretion. Because the executive branch can’t enforce every law to the hilt, it argued, it’s okay under the Take Care Clause to temporarily suspend enforcement in order to smooth a new program’s implementation.

This argument never saw the inside of a courtroom — again, standing was a roadblock. But it also wasn’t a good argument. Yes, the executive branch must make hard choices about how to marshal its limited resources. We don’t want FBI agents ticketing jaywalkers if it means they can’t stop the next terrorist attack. There’s a big difference, however, between setting priorities and suspending a law outright.

After all, the Obama administration could have quietly deprioritized ACA enforcement. Had it done so, however, insurers would still have felt obliged to follow the law — and plans would still have been canceled. Fixing the “like it, keep it” problem required the administration to publicly commit to non-enforcement. In so doing, the administration licensed the regulated community to ignore a congressional statute, a move that was inimical to the Take Care Clause.

It also set a worrisome precedent. Although I’m a supporter of Obamacare, I raised alarms about the delays in the New England Journal of Medicine:

A future administration that is less sympathetic to the ACA could invoke the delays as precedent for declining to enforce other provisions that it dislikes, including provisions that are essential to the proper functioning of the law. The delays could therefore undermine the very statute they were meant to protect—and perhaps imperil the ACA’s effort to extend coverage to tens of millions of people.

A Republican president’s refusal to enforce the individual mandate was exactly what I feared most.

Fairly or not, what went around is coming around

And guess what’s happening now? Trump got the message that he doesn’t need Congress to change the ACA. His very first executive order told his subordinates to consider delaying the implementation of any and all of the ACA’s taxes. And remember: The individual mandate, the Supreme Court famously concluded, is a tax.

But if it wasn’t legal for the Obama administration to delay the law, it wouldn’t be legal for the Trump administration to do so, either. Last I checked, two wrongs don’t make a right.

Trump is also contemplating going further than Obama ever did. Obama’s delays offered transition relief from laws that hadn’t yet taken effect. They didn’t suspend laws that were already in force. And it’s not as if Congress acquiesced to the practice of delaying statutes. To the contrary, congressional Republicans objected repeatedly and angrily, opening themselves to the charge of hypocrisy if they defend Trump’s illegal actions.

I doubt any of this will trouble Trump, however, and, again, experience with the Obama delays suggests why. At the bottom, law is a set of social practices that tell us what’s permissible in a given community. That’s what makes the rule of law robust: the widespread belief that certain things are just not done means the president pays a political price for transgressing legal limits. But it’s also what makes the rule of law so fragile. If the social practice of condemning violations decays, the rule of law decays with it.

During the Obama era, however, almost no left-leaning lawyers spoke out against the delays. Some even defended them: Tim Jost and Simon Lazarus argued, for example, that the “postponements are not refusals to enforce the ACA but temporary course corrections in the interest of effective implementation.” Never mind that this set the stage for a future president to suspend other parts of Obamacare by claiming, however speciously, that it’s the only way to prevent it from collapsing.

At the same time, right-leaning lawyers spun the delays into a larger narrative about Obama’s perfidy and lawlessness. That narrative is exaggerated: As I’ve argued, the record, taken as a whole, “refutes the claim that President Obama has systematically disregarded the ACA’s text or displayed contempt for legal constraints.” But Republicans ran with it anyhow.

I fear we’re about to witness a jarring role reversal. Leftie lawyers will scream about legal violations, both when they happen and when they don’t. Rightie lawyers will offer defenses or stay mute. Lacking trusted mediators, the public will dismiss all this law talk as so much partisan bickering. And Trump will be Trump, law be damned.

Unlike Obama, Trump doesn’t face a hostile Congress

Recent press reports have suggested that Trump “plans to use his administrative powers to dismantle large swaths of the health law well before Congress passes a replacement package.” But there are also reasons Trump might hold off.

Obama tested the limits of executive power because he faced a hostile Congress bent on sabotaging the ACA. Trump doesn’t have that problem: He and Congress are on the same team. Why delay the individual mandate when Congress can end it?

Other parts of the ACA may be more vulnerable to delays. Through the reconciliation process, congressional Republicans can repeal taxes, including the individual mandate, without worrying about the Senate filibuster. But they can’t repeal rules that directly regulate health insurance. One of those rules, for example, requires health plans to cover the “essential health benefits.” Republicans hate that rule — making it a prime target for an administrative attack by Trump.

That said, I worry less about the illegal things Trump might do than all the legal things he can do. To implement health reform, federal agencies wrote hundreds of rules, issued reams of guidance, and made millions of decisions. All of those efforts were undertaken in service of covering the uninsured.

The Trump administration doesn’t appear to share that goal. Dismantling the ACA through the executive branch couldn’t happen overnight: Lots of ACA rules can be revoked or revised only through a lengthy, complex process.

But Trump has four years. He can do a lot of damage in that time, all within the bounds of the law. If he breaks the law, he can do even more. And, in a small way, President Obama made it easier for him to do so.

Nicholas Bagley is a professor at the University of Michigan Law School

The Big Idea is Vox’s home for smart, often scholarly excursions into the most important issues and ideas in politics, science, and culture — typically written by outside contributors. If you have an idea for a piece, pitch us at thebigidea@vox.com