Donald Kochan has set forth a concise and persuasive account of congressional delegation of broad swaths of lawmaking authority to administrative agencies, which may be why his article has attracted such attention from the usual suspects. Kochan is, as so many are, pessimistic in his conclusions: short of a deus ex machina like a revivified non-delegation doctrine, the pathologies of our institutions will continue to run their course. But there is, I think, more to the story.

Kochan’s account is that Congress is engaged in “unilateral disengagement.” To avoid accountability for unpopular policies, it legislates broadly, vaguely, or not all, and leaves all the details, often billion-dollar details, to the agencies. And the agencies, in turn, are only too happy to seize that power, for reasons of self-aggrandizement and lack of popular accountability. All of this is true, and it shouldn’t be controversial: it’s what the political scientists and public choice economists have been saying for decades.

The real debate is over whether this is a good or a bad thing. Kochan’s position is in the latter camp. The other side, meanwhile, argues that we need a vigorous regulatory state to confront the complexity of the modern world. Perhaps Congress could never, institutionally, handle all the details, even the billion-dollar ones. And perhaps attenuated judicial review of the all the agencies’ actions is all that’s needed to keep everything on track.

But I think the people who tell that rather Panglossian story are increasingly dissatisfied. Because there is, or was, something deeper going on in Congress than “unilateral disengagement.” And that deeper strategy isn’t working for them anymore.

Let me share an anecdote. It was 2008, I was 28 years old and I was a “senior policy analyst,” if you can believe that, at the Heritage Foundation. Congress was working on amendments to the Americans with Disabilities Act. Everyone knew that the legislation was going to move. In June, the House had passed its version. As relevant here, the ADA had always defined the term “disability” to include “a physical or mental impairment that substantially limits one or more major life activities of such individual.” The House bill expanded the definition, including by adding a new definition for the term “substantially limits,” defining it to mean “materially restricts.” The idea was that this would work with the other broad provisions in the bill to knock down limits on who could claim the protections of the Act. But the Senate was facing pressure from the business community, which feared that the Act’s new definitions would open the floodgates to litigation. So several members of the relevant Senate Committee asked for help in redrafting the House’s language. I gave them a laundry list of options to clarify the text and help ease the burden on employers trying to figure out their legal obligations. Ultimately, the Senate made a single change to the definition: it struck out the definition of “substantially limits” as meaning “materially restricts.” And then everyone on both sides of the aisle declared this to be a stunning compromise, a real victory for the business community and for the disabled.

Of course, here on planet earth, there was no compromise: Congress ducked the issue entirely—it literally deleted a definition and replaced it with nothing. The reason it did that wasn’t solely “unilateral disengagement.” The Republicans had little power in the 110th Congress and would take whatever they could get and claim a win. Meanwhile, Ted Kennedy, the Senate Committee chairman, could be confident that the advocacy groups, the agencies, and the courts would take the new text and run with it, achieving the same result whether or not Congress bothered to give a definition to the term “substantially limits.”

That is, or was, the deeper game: Congress could punt on making decisions in legislation, and those favoring stricter regulation could rest secure that the administrative agencies and the courts would get the right idea going forward. So why fight the tough fights in Congress when you can win a year or two later through regulation or in court? And once a given issue was decided in a regulation or in court, that was it, forever—a one-way ratchet just as good as putting the words themselves in the U.S. Code.

Undergirding this strategy were two assumptions. The first was that the courts would grease the skids to advance a pro-regulatory view of legislative purpose, without concerning themselves overly much with minor things like statutory text. And that was a really good assumption! Back in the late 1970s, the courts directed EPA to create a whole new billion-dollar Clean Air Act program out of whole cloth, what’s now known as “Prevention of Significant Deterioration.” The relevant court decision relies entirely on a single statutory provision, the non-operative statement of statutory purpose, and then quotes a couple paragraphs of the legislative history to the effect that Congress’s purpose was to “protect and enhance the quality of the Nation’s air resources.” And that was enough to force the EPA to impose federal preconstruction permitting for the entire country. At the end of the day, the EPA was only too happy to comply.

But leisure suits and platform shoes are no longer in fashion, and neither is the freewheeling approach to statutory interpretation of the 1970s. As Justice Kagan famously observed, “We’re all textualists now.” Even the reliably corrupt canon of statutory interpretation that remedial statutes must be liberally construed—effectively, a brick on the scale in favor of administrative power—has gone extinct in recent years. And it’s fair to say that the courts have become increasingly skeptical of reading into statutes broad grants of administrative authority over important economic and social authority—in fact, the Supreme Court, according to some views, is now mounting an “attack” on the administrative state because it sometimes refuses to conjure major agency powers out of statutes that don’t quite appear to convey them. The bottom line is that today’s Ted Kennedys can’t count on the courts to reliably carry out their regulatory agenda.

The second assumption was that there would rarely, if ever, be regulatory reversals. Once a major policy was enshrined in regulation, that would be the end of it. And, in practice, “deregulation” over the years has largely meant, in practice, slowing the growth of new regulations, with little pruning of the old ones. But that assumption also appears to have broken down. For every major policy willed into being by the Obama Administration, there is an equal and opposite Trump Administration policy. Just in the environmental field, you have the Clean Power Plan, the Waters of the United States rule, and automobile-emission standards, all of which are now undergoing what is politely called “reconsideration.” Again, if you’re a modern-day Ted Kennedy, you can’t count any more on making durable gains during Democratic administrations. What you can count on is major policy losses during Republican Administrations.

This new state of affairs must be incredibly frustrating. The deep strategy that worked for decades has run out of steam. Serious reversals come regularly from the courts and from the administration. What’s to be done?

The obvious answer is a return to legislating. There’s a persuasive alternative history that if the Obama Administration hadn’t pivoted so quickly to regulation, we would have a law—that is, an actual statute, right there in the U.S. Code—addressing greenhouse gas emissions. And the Trump Administration wouldn’t be able to anything about it except at the margins. The same can be said of any number of progressive regulatory initiatives that have floundered, in one way or another, over the past decade.

But, so far, the mindset hasn’t changed, and the battle lines over the administrative state haven’t budged. There’s still this idea that the old strategy can still work, even though it very clearly isn’t. Why is that? Part of the answer is inertia. And there is much truth to Kochan’s point that Congress still faces skewed incentives against actually legislating.

But that could easily change. Imagine, if you would, a second Trump Administration, one with more effective agency leaders, rather than the firebrands who were there at the start, and with more waves of judicial nominees. Could that be enough for all the right-thinking people, the ones whose hearts are in the right place, to start demanding more of Congress or, at least, that the agencies be reined in? It might well be, and we’re already seeing some stirrings in that direction, as the Members of the House think through what they can do to restrain the Administration.

So the incentives may already be changing. And, so far, the so-called anti-administrativists have held firm to their views, despite having some allies in Administration. And, so far, the Trump Administration has also held firm in its rhetoric against administrative overreach. So what we may see in the future is less a realignment on these issues than a new strange-bedfellows coalition that agrees that the status quo isn’t working for anyone. Give it another few years, and someone like Justice Kagan may proclaim that we’re all anti-administrativists now.

In other words, Kochan’s conclusion about Congress’s “unilateral disengagement” is entirely accurate, but not inevitable or eternal.

Andrew M. Grossman is a partner in the Washington, D.C. office of Baker & Hostetler LLP and co-leader of the firm’s appellate practice. He is also an adjunct scholar of the Cato Institute. This post is based on remarks delivered at a Federalist Society panel discussion on Donald Kochan’s Strategic Institutional Positioning: How We Have Come to Generate Environmental Law Without Congress.