When people hear the phrase “a new Fed–Treasury Accord,” they tend to react reflexively. They imagine subordination. They imagine monetisation. They imagine the end of central bank independence.
History suggests the opposite.
The 1951 Accord did not weaken the Federal Reserve. It restored its independence. During and after World War II, the Fed was effectively committed to maintaining low long-term interest rates by purchasing Treasury debt as necessary. It monetised to preserve a rate peg. Inflation followed. The Accord re-established the Fed’s authority to allow long-term rates to rise — and, more importantly, to refuse automatic deficit financing.
To invoke the Accord historically is to invoke independence. Not coordination. Not fiscal dominance. Independence.
The present debate is therefore not about whether the Fed should serve the Treasury. It is about clarifying institutional boundaries that have blurred over the past fifteen years.
The real issue is structural.
The central technical question is this: who controls the maturity structure of U.S. government debt?
The Treasury chooses duration. That choice determines the taxpayer’s exposure to future interest rate movements. Issuing long-term debt provides protection: payments are fixed; fiscal sensitivity to rate increases is reduced.
Quantitative easing altered that structure. By purchasing long-term bonds and issuing short-term reserves, the Fed effectively shortened the consolidated government’s duration exposure. One arm of the state lengthened maturity. The other shortened it. The result was greater fiscal sensitivity to rising rates.
That is not merely monetary policy. It is fiscal intervention by other means.
If institutional clarity is to be restored, the principle must be straightforward: the Treasury controls duration. If the Fed wishes to supply reserves, it should operate at the short end. It can lend against collateral. It can conduct explicit duration swaps. But it should not unilaterally reshape the government’s interest-rate risk profile.
The same logic applies to credit allocation. Mortgage-backed securities purchases. Corporate bond backstops. Sectoral interventions. These are distributive choices. If the government wishes to subsidise housing or particular industries, that responsibility lies with elected officials and the Treasury. An independent central bank should not be allocating credit.
Monetary stability and financial stability are sufficient mandates. Expanding beyond them weakens independence rather than strengthening it.
The most difficult boundary, of course, appears in crises. In 2020, the Fed monetised extraordinary volumes of new Treasury issuance while suppressing long-term rates. Whether one attributes subsequent inflation primarily to that episode or not, the institutional precedent is clear: deficits were financed at scale through central bank balance sheet expansion.
A new Accord, properly conceived, would address that question in advance. Coordination may be necessary in crisis. Automatic monetisation should not be.
Having said that, one must ask: what is Warsh actually trying to do?
I believe Warsh’s framework is frequently misunderstood. He is not arguing for austerity. He is not a mechanical monetarist. Nor is he advocating fiscal dominance.
I believe his core conviction is that inflation is fundamentally institutional. It arises when fiscal credibility erodes and when the line between fiscal and monetary authority disappears. It becomes persistent when deficits are implicitly financed rather than transparently funded.
His critique of the post-2008 Fed is not primarily technical. It is institutional. Quantitative easing did more than stimulate demand. It altered the structure of public liabilities. It reduced market discipline. It shifted duration risk. It fostered the perception that fiscal expansion would always be backstopped by the central bank.
I believe Warsh sees this as the true long-term danger — not high rates, but blurred boundaries.
But if not monetisation, and not austerity, then what?
The United States faces structurally high deficits. The politically feasible options are limited. Harsh austerity is improbable. Sustained inflation would erode credibility and capital markets. Financial repression carries its own distortions.
That leaves one durable path: sustained nominal growth.
And here lies what I believe is the central strategic bet.
Artificial intelligence has opened a historic window. It creates the possibility of a broad-based positive supply shock. If productivity accelerates materially, potential real growth rises. If real growth rises, nominal growth can increase without destabilising inflation expectations. And if nominal growth exceeds the effective cost of debt, the debt-to-GDP ratio stabilises through expansion of the denominator rather than erosion of the numerator.
This is not fiscal dominance. It is not monetisation. It is a bet on a supply-driven resolution of a fiscal problem.
I believe Warsh understands that in a high-debt regime, monetary policy is no longer the dominant variable. The dominant variable is the credibility of the fiscal path relative to nominal growth. If growth materialises, fiscal sustainability improves organically. If it does not, markets will impose discipline through higher long-term rates.
Ultimately, the bond market will arbitrate the experiment.
The intellectual architecture, however, is coherent: restore institutional boundaries; reduce monetary activism; return fiscal responsibility to the Treasury; and allow structural growth — potentially powered by AI — to carry the burden that monetary engineering cannot.
This is not a tactical adjustment. It is a regime bet.
And like all regime bets, it will be judged not by speeches, but by the long end of the curve.