Beyond Institutional Agnosticism

What comes after the policy state?

Today, the LPE Blog ran a piece by Michaeljit Sandhu responding to my own work on the Blog. Sandhu rejects my call for legislative primary in favor of institutional agnosticism. I found the blog post sharp but unsatisfying.

Every significant reconstruction moment in American politics, save one, adopted a specific program of what I have called “constitutional politics.” Reconstruction Republicans, the Progressives, and the New Dealers all tried to remake the state on more favorable terms.

They were almost too successful. After the New Deal, left-of-center Americans looked on their work in awe like they were in a Byron poem. They saw a rational state capable of delivering policy that could itself change the world. Constitutional politics was replaced by policy and institutional agnosticism. The goal was no longer a fundamental reordering of the American state.

That’s where the exception comes in—the Rights Revolution. Various entities within this epoch were agnostics. The civil rights movement played every slot machine: presidential, administrative, legislative, and judicial. In the process, the rights revolutionaries created a juristocracy that would, in due time, undo all of their work. In the infamous Powell Memo, a future juristocrat advised a nascent neoliberal coalition that they ought to seize on the fruits of the rights revolutionaries’ agnosticism by capturing the Court. Today, we live with the consequences.

Modes of governing authority never sit still. If you play at their games, you generate warrants on authority that you will not be able to control. Saying you want to be institutionally agnostic reminds me of a kid saying he wants to grow up to be a lawyer, a doctor, and a rocket scientist. These are different games. How they interact is contingent in the extreme. And in constitutional politics, not every game abounds to democracy’s benefit in the long run.

I think for Sandhu’s post to work he had to have tackled two interrelated points: (1) some modes of governing authority will in the long run tend to entrench hierarchy and (2) agnosticism reifies the power of these modes. Eg, it is suicide to put your chips down on juristocracy if it abounds towards conservative outcomes in the long run.

The Rights Revolution gives us the bounty of institutional agnosticism. You end up strengthening institutions that are set against the left and reconstruction. In the process you destabilize whatever you managed to accomplish.

Sandhu is surely right that Congress is no democratic Eden. It has repeatedly entrenched hierarchy, frustrated reform, and disappointed progressive aspirations. But this observation does not eliminate the need for an institutional theory. It simply sharpen it.

The problem is that institutional agnosticism is not itself an alternative. Every political movement eventually confronts questions about where governing authority should reside, how political victories should become durable, and which institutions ought to exercise final responsibility for democratic decisionmaking.

Refusing to answer those questions does not transcend institutional politics. It merely leaves intact the constitutional settlement we already have.

(Here it is important to remember that an LPE history of the last sixty years or so tells us that our current settlement was shaped to a tremendous degree by forces set against effective democracy. Why concede to the game they have constructed for us? Why do we think we are immune from the constitutional politics that the right wields so effectively?)

For nearly half a century, American government has operated through what Karen Orn and Stephen Skowronek described as the policy state: an order in which policymaking increasingly migrated away from legislatures and toward presidents, administrators, courts, and diffuse networks of governance. Executive discretion expanded as Congress was sidelined by an accommodation between presidentialists and juristocrats. Administrative agencies accumulated greater responsibility for national policymaking while becoming increasingly vulnerable to presidential control. Courts assumed a more prominent role in supervising administration and resolving political conflict. The result was not the disappearance of politics but its relocation.

Today, that settlement appears exhausted.

The policy state promised flexibility, expertise, and effective governance despite legislative fragmentation. Instead, it has produced escalating presidential unilateralism, constitutional hardball, judicial management of national policy, and administrative instability that leaves fundamental questions of governance swinging dramatically with each electoral cycle. Presidents increasingly govern through emergency declarations, aggressive statutory interpretation, and administrative improvisation. Courts increasingly determine the boundaries of public policy.

These are not isolated pathologies. They are signs of developmental exhaustion. The old order is not working. And no wish list of policy objectives can fix that.

Institutional agnosticism therefore risks becoming profoundly conservative. By declining to articulate an alternative architecture, it defaults to preserving one that has already ceased to function well.

No plan is not an alternative to a plan. The existing institutional order already embodies a theory of governance. The only question is whether we intend to replace it.

This is where legislative primacy enters the conversation. Properly understood, it is not a romantic celebration of Congress as presently constituted, nor a denial of Congress’s reactionary history. It is a reconstruction project aimed at replacing a failing constitutional settlement. Its claim is comparative rather than absolute. Every institution has disappointed democratic aspirations. The question is which institution is best positioned to convert democratic mobilization into durable public authority.

Congress remains uniquely situated to perform that function—not because legislators are wiser than presidents or judges, but because legislation institutionalizes coalition-building, public bargaining, and democratic accountability in ways executive governance cannot. Durable political settlements are rarely produced through unilateral administration. They emerge from legislative reconstruction.

Indeed, history points in precisely this direction. Reconstruction, the New Deal, the Great Society, and the Civil Rights Revolution all required strong executive leadership. But none became durable because presidents acted alone. Their achievements endured because they were translated into legislative settlements that reshaped the structure of American government itself. Executive action may catalyze constitutional transformation. It rarely completes it.

(The Trump Administration is a cautionary tale about what happens when a president with big ambitions meets a political movement with no sense of legislative politics.)

Sandhu asks why Congress deserves pride of place. The better question is what institutional project should succeed the policy state.

If the answer is simply that reformers should use whichever institution is presently available, then institutional agnosticism offers little guidance at precisely the moments when constitutional politics becomes most consequential.

It cannot tell us whether courts should entrench reform, whether presidents should bypass legislatures, or whether administrative governance should substitute for democratic lawmaking. It resolves these questions pragmatically only by leaving existing institutional distributions of power largely intact.

The central issue, then, is not whether Congress has failed. It plainly has. The central issue is whether the institutional order built around presidential administration and judicial supervision has succeeded where Congress faltered. After decades of executive and judicial aggrandizement, administrative instability, and declining democratic legitimacy, that answer increasingly appears to be no.

The debate over legislative primacy is therefore not a nostalgic dispute about Congress. It is a debate about what comes after the policy state. And if the existing constitutional order has reached the limits of its developmental capacity, refusing to propose a successor is not institutional humility. It is acquiescence.