Central-bank Independence
Our constitutional order does not countenance a completely independent agency. Thus, although the Federal Reserve enjoys great independence, it is also constrained by institutional rules and political oversight. Yet over time, the Fed has greatly expanded the scope of its activities. It has stepped into fiscal and political territory usually reserved for more politically accountable agencies. And it has presided over embarrassing institutional failures. The extent and nature of Fed independence are newly in question.
Reform must come sooner or later. The Fed could remain an expansive, powerful, and political agency, but necessarily lose some independence and face more direct control by elected officials, like the Treasury. Or the Fed could return to a narrower scope of activities consistent with its current or even greater independence. I argue for the latter course. Only the combination of independence and limitations surmounts the central pre-commitment problems of monetary policy and financial regulation.
But raising the drawbridge, hoisting the independence flag, and defending the status quo is not going to be tenable for long.
LIMITS ON INDEPENDENCE
The Fed is not absolutely independent. It may not print money and hand that money out as it sees fit. Rather, the Fed is limited by the political branches, by its mandates, and by the tools it has available to carry out those mandates. The Fed was created by Congress, to serve ends defined by Congress.
Fed officials are appointed by the president, confirmed by the Senate, serve limited terms, and must report regularly to Congress. Lawmakers have also established rules governing regional federal-reserve banks, the appointment of their officials, and the potential removal of those officials.
The Fed has a limited authority, crystallized in its "dual mandate," which features "price stability" and "maximum employment." Implicit in these mandates is, "and nothing else." A central banker may believe that reshoring manufacturing or combatting climate change are vital issues, and that spending money, channeling lending, or forcing banks to act would address those issues. But the Fed must leave these issues to Congress, the president, or other federal agencies.
Finally, the Fed's tools are limited. The Fed may set short-term interest rates, borrow and lend at plausible market rates, and buy and sell securities, always receiving something of value in return for money given out. But it may not give people money, or confiscate money. It may only transact with banks and a few other large qualified financial institutions. It may not transact directly with the Treasury, an important limit on inflationary finance. The kinds of securities it may buy are sharply limited. It may not tell banks who to lend to and at what rates — at least not directly, or of its own accord without congressional direction.
These limits are striking when you think about them. The Fed is charged with controlling inflation. The most powerful way to create inflation is to print money and hand it out. The most powerful way to stop inflation is to confiscate money. The Fed is forbidden to use these tools. The Fed must instead nudge inflation via overnight interest rates. The Fed is charged with maximizing employment. Yet the Fed may not lower or raise labor taxes, subsidize hiring, empower or restrict unions, fix social-program disincentives, change minimum- or prevailing-wage laws, halt or expand immigration, or just print money and hire people.
Why is the Fed subject to these limits? The answer is that in our republic, decisions regarding taxing, spending, and regulating must be reserved for politically accountable parts of government. Fiscal policy — handing out money or confiscating it — may be a powerful engine of inflation and disinflation, but it also transfers wealth, which is an inherently political task. Telling banks to lend to A and not to B is also political. These kinds of decisions must be made by elected representatives of the people, as imperfect as those representatives may appear to aspiring technocrats.
Monetary policy creates winners and losers, of course. Inflation and low real interest rates help borrowers, home buyers, and the government while hurting savers and government bondholders. Recessions induced by monetary policy hurt unemployed workers and cyclically sensitive industries. But these are indirect, diffuse effects.
President Donald Trump's threats to fire members of the Federal Reserve Board and the Fed chair, in addition to his pushing for lower interest rates, have thrust central-bank independence into the headlines. But questions surrounding central-bank independence have been simmering for far longer, and go far beyond the terms of office and removal of governors. What should the Fed's mandates be? How should they be enforced? What limits should there be on the Fed's scope of action? What rules should guide its decision-making processes? How far into the financial system should the Fed's crisis support and consequent regulation extend? When, inevitably, central-bank activities and fiscal or other policies are in conflict, how is that tension resolved?
COMPLAINTS
Despite limits on its independence, the Fed has dramatically expanded the scope of its activities in the recent past, and arguably waded into political waters. The Fed bought trillions of dollars in government bonds, including most of the bond issues that financed massive deficits in 2020 and 2021. By buying long-term bonds, the Fed shortened the maturity structure of federal debt. When interest rates rose, higher interest costs on the debt cost taxpayers hundreds of billions of dollars. In quantitative-easing operations, the Fed also bought trillions of dollars of mortgage-backed securities, with the explicit goal of funneling credit to the housing market and, thereby, not to other borrowers. In 2008 and again in 2020, the Fed supported asset prices by making direct asset purchases and lending to intermediaries to finance purchases.
The Fed has advocated fiscal stimulus but seldom fiscal restraint. It has dabbled in "inclusive employment" initiatives. It addressed highly speculative climate risks to the financial system while other regulatory problems festered. It has cooperated with the de-banking of unfavored industries and has forced banks to fund political groups. It has increased the reserve supply by trillions of dollars and innovated by paying interest on those reserves.
Whether you approve or disapprove of this expansion (I approve of the last one), these interventions far beyond the Fed's traditional limitations all demand examination and consent from the American people via our elected representatives.
Complaints about the Fed's crisis interventions and related financial-stability regulation go back further in time. The Fed was founded in 1913, after a series financial panics, to be the "lender of last resort." It promptly failed in 1933. Since then, it has cycled through periods of expanding intervention matched by expanding risk-taking.
The Fed and the Federal Deposit Insurance Corporation (FDIC) guarantee bank depositors and other creditors, in the hopes of preventing bank runs. These institutions try to regulate bank assets to offset the disincentives of a deposit guarantee. But when regulation and supervision fail, a new run breaks out, and the Fed and the FDIC guarantee more liabilities. The 2008 crash, bailouts, and Dodd-Frank Act were just one turn of the cycle. The large interventions of 2020 and the Silicon Valley Bank (SVB) affair prove that the underlying fragility remains.
Today, people count on the Fed and the FDIC to guarantee all bank deposits and money-market funds, quash price decreases in Treasury- and corporate-bond markets, and provide liquidity in pretty much every market. Nobody expects a big bank to fail. A perpetual put option, with private gain in good times and the ability to sell quickly to the Fed in bad times, creates dangerous incentives for financial-market participants to take on too much leverage and risk, which then demands ever greater Fed intervention.
The SVB case is revealing. SVB took large uninsured deposits from institutional investors and bought long-term Treasury securities. If interest rates rose, SVB would swiftly be insolvent, as the market value of its assets would be less than its deposits. SVB's large institutional depositors would then quickly run to get their money out. Despite an army of regulators and supervisors deploying hundreds of thousands of pages of regulations, and despite the utter simplicity of the problem — no complex toxic derivatives here — the Fed did not detect and remedy interest-rate risk combined with large institutional deposits. SVB suffered a run, failed, and the Fed and FDIC extended deposit insurance ex post to institutional depositors with accounts in the hundreds of millions of dollars. Several other banks failed at the same time, and many larger banks came close.
More deeply, under pressure from banks, the Fed has not taken the transparently simple step of adequately raising equity requirements and reducing the financial system's reliance on run-prone and illiquidity-prone short-term financing, the one sure cure for financial crises.
Inflation that approached 10% in 2022 and financial meltdowns seem like self-evident institutional failures. Perhaps these were events that even a perfectly run Fed could not have forestalled. Perhaps another tweak of the strategy and a few hundred more rules would preclude a recurrence. But the Fed has never had a transparent public accounting of what went wrong during these events, or how it might avoid a repetition. In fact, it seems proud of its increasing interventions — and ready to go again.
Too much or too little independence does not jump out as the central cause of this common litany of complaints. The Fed does not seem to have followed policies of recent decades unwillingly under political pressure.
The Congress and past presidential administrations do not seem terribly unhappy with the Fed's actions, either. If anything, during oversight hearings, individual members of Congress argue for more credit to their constituencies, or for the Fed to lean more into their own political projects. The Congress has largely not complained loudly that the Fed acted beyond its mandate. Mandates, like constitutions, are written in broad language, reinterpreted over time, and must be enforced to retain salience. When the Fed reinterprets or expands its mandate or takes on a new tool, and Congress does not object, eventually that silence becomes acquiescence.
In the end, the Fed is not fundamentally different from the rest of the alphabet soup of semi-independent federal agencies. Once created, they expand steadily; increase their staff, budget, and authority; grow cozy with the industries they regulate; start to take on more political roles; and accumulate institutional grandeur and power. When trouble looms, they circle the wagons, defend the institution above all else, and never admit mistakes that might weaken the institution's prestige.
This tendency is not the fault of individuals. Every Fed official or employee I have known is a well-intentioned, smart, honest, hardworking, and dedicated public servant. The Fed has historically been one of the most competent and well-run public agencies. This is not a story of greed, incompetence, malice, or corruption. Throwing "bad people" out and putting "good people" in will make no difference.
There was no design to the Fed's expansion, either. Events provoked expedients, each of which became permanent. The expansion occurred because neither the Fed nor the government wished to reform and return to a more limited scope after crises had passed.
Faced with a more powerful institution stepping into classically political arenas, one could advocate that we accept the institution's accumulation of additional power and try to make it more politically accountable, like other regulatory agencies or even the Treasury. Advocates of such "democratic accountability," however, must face the fact that this sort of Fed would be just as accountable to the next administration, which might have drastically different plans for it.
The alternative is for Congress, working in conjunction with the Fed, to restore and enhance the limits on the Fed's mandates and powers, and to better hold the Fed accountable to those guardrails. Some of the politically questionable new powers — including bailouts, some crisis-management tools, term-structure management, or ongoing subsidies for particular industries — can be transferred to the Treasury or other agencies. Then the Fed could act with restored independence within narrower limits.
INDEPENDENCE AND LIMITATIONS
What is distinct about monetary policy that benefits from independence? What are the questions to which independence is the answer, and uniquely for the Fed?
The most common response is that independence insulates monetary policy from short-term political pressure, specifically pressure to goose the economy in ways that cause inflation. But that rationale doesn't distinguish monetary policy from other economic policies. Politicians are always tempted to hand out stimulus checks, regulatory favors, tax cuts and exemptions, loan forgiveness, and other goodies. Yet we do not entrust these policies to independent agencies.
Moreover, "political pressure" is also "democratic accountability." If independent central bankers run off with their own policy preferences, become dysfunctional, or are captured by narrow political groups or industry, we want our elected representatives to have the power to correct the central bank. The ballot box is our restorative force, not rule by an insulated technocratic elite.
Economists usually argue that central-bank independence addresses a "time inconsistency" problem and the concomitant need to pre-commit against future temptations. In the conventional analysis, the government faces a "Phillips curve" tradeoff between inflation and employment. The more inflation people expect in the future, the more inflation there is today for any given level of employment. The government thus wishes for people to expect low future inflation. Then, the government can choose a little extra inflation today to get more employment today. But people know the government will want to make the same inflationary choice in the future, so they expect future inflation no matter what the government promises today. Thus, the government ends up with high inflation today and in the future without increasing employment.
To avoid this scenario, the government wants to pre-commit to less inflation than it will desire in the future, when the time comes. Like Odysseus, it wants to tie its hands to the mast.
Reflecting this analysis, today's consensus view of monetary policy emphasizes the importance of "anchored" expectations of future inflation. Anchored expectations allow temporary inflation to offset other shocks without shifting the inflation-employment tradeoff and therefore turning into long-lasting stagflation.
But aside from general expressions of intent, few at the Fed are willing to say aloud where "anchoring" comes from. The Fed may be committed to its inflation target, eventually, but just what is the Fed willing to do about it? Anchors are cheap talk in a calm sea. If 1979-style inflation breaks out again, will the Fed really repeat the painful recessions of the 1980s, if that's what it takes to bring inflation back under control?
Independence alone is a weak pre-commitment mechanism. Independent central bankers whose preferences on inflation versus unemployment align with the government's preferences will want to inflate every bit as much as the government does. Independence only provides an effective pre-commitment to low inflation if the government appoints central bankers whose natural preferences are more hawkish than the government's — central bankers who have, from the government's point of view, an irrational or ideological commitment against inflation. Such choices are not evident in appointments or Senate confirmations, to put it mildly. Rules, mandates, and limits that prohibit the Fed from inflating are likely more effective pre-commitments than independence alone.
Moreover, the same pre-commitment issues pervade other government policies. The government is tempted to expropriate wealth, impose rent and price controls, or default on debts. Each has a short-run benefit and a long-run cost: If the government takes these actions, people don't invest, build apartments or businesses, develop new products, or invest in government debt. Limitations on government action including constitutional and property rights are the core structures of this pre-commitment, not independence of discretionary decision makers.
The tension between monetary and fiscal policies poses a harder issue than inflation versus employment. This tension is historically more significant, and likely to be the core tension going forward. Governments are tempted to print money to finance deficits or repay debts. If governments too often inflate away debts, they can't borrow in the first place. An independent central bank that will not or cannot follow inflationary policies — buying government debt with newly created money, forcing banks to hold government debt, artificially lowering interest rates — pre-commits the government to enact the politically difficult tax, spending, and microeconomic policies rather than inflate its way out of trouble. In the end, if we never wish to see inflation, a truly independent central bank must force the government into explicit default and debt restructuring rather than implicit default via inflation, as countries without the reserve currency must do.
Resisting fiscal inflation is much harder than resisting a president's desire for a Phillips-curve election sweetener or offsetting Keynesian-multiplier fiscal stimulus by monetary restraint. If a central bank raises interest rates to fight inflation, that raises interest costs on the debt. With 100% debt-to-GDP ratio, each percent higher interest rate translates to 1% GDP greater deficit. With interest costs already around $1 trillion a year, this pressure is with us today. President Trump cited U.S. government borrowing costs as a reason he desires lower interest rates. A much harder tension lies ahead: Imagine the government wishes to borrow money during a war, a severe recession, or another pandemic, to finance spending that voters desire. Or suppose the bond vigilantes finally arrive and refuse to roll over our debt. Will the Fed really allow rates to skyrocket, force sharp fiscal austerity, or force explicit default or restructuring, all in the name of containing hypothetical inflation? Moreover, will it do so knowing the financial system's continued reliance on periodic bailouts and risk-free Treasury debt?
It's difficult to imagine a central bank that is strong and independent enough to resist monetizing debt under these circumstances. The Fed participated willingly in huge government debt purchases and low interest rates in 2020-2021, and of course during World War II.
Would we even want a Fed so independent that it could, of its own volition, produce unemployment over 10% as in 1982 — the highest since the Great Depression — or financial turmoil in an anti-inflationary quest, over the objections of an elected president and Congress? Should we want a Fed that can decide all on its own to veto the sort of emergency fiscal actions the president and Congress took during the pandemic, or to force tax increases and spending cuts on the government? Shouldn't such painful national ventures require the approval of the people's representatives?
Laws and mandates barring the Fed from allowing inflation that clearly indicate the government's political support for that choice represent more promising means of pre-committing against fiscal inflation than pure independence.
The historic gold standard was essentially a pre-commitment against fiscal inflation. The gold standard was a rule, not a grant of independence. The government promised redemption at a fixed rate; bring in a dollar and receive, say, 1/20 oz of gold. Under the gold standard, the Fed was not authorized to independently choose whatever exchange rate for gold it wanted each day. The United States didn't even have a central bank during much of the gold-standard period. Joining a currency union, adopting a currency board, dollarizing, or borrowing in foreign currency are similar rule-based devices that governments use to pre-commit against fiscal inflation. (However attractive many of its features are, the gold standard will not work for the United States today.)
Now, pre-commitments against fiscal inflation should not necessarily be absolute. We would not want to have lost World War II on the altar of no inflation. In economic theory, the rare "state-contingent default" forces bondholders as well as taxpayers and spending beneficiaries to pay in part for a war or crisis. Inflation, suspension of convertibility, and financial repression have been part of war and crisis finance for centuries. Yet every day is not a crisis (though to politicians and their speech writers, it often seems so). Thus, a good monetary-fiscal regime pre-commits that state-contingent default will be rare, like "just this once" wealth taxes, so that bondholders will lend to the government with only a moderate risk/inflation premium.
The delicate task then, is to construct a pre-commitment regime that largely avoids fiscal inflation but allows an exception for genuine national emergencies. Tie hands to the mast, for sure, but not so tightly that one goes down if the ship starts to sink.
That consideration might seem to argue for independent discretion; letting the Fed independently decide when to break the rules. But the choice between inflation, sharp tax increases and spending cuts, or letting a crisis burn, is also intensely political. The elected branches need to own such a decision. Thus, it seems better that the administration — or better yet, Congress — declares an emergency, temporarily suspends legal restraints on inflationary finance, and thereby takes responsibility for the choice of inflation over default or austerity. In the 2008 crisis, the Congress suspended usual limits by establishing the Troubled Asset Relief Program. Whatever one thinks of the act, it shows that Congress can and does suspend rules in an emergency.
FINANCIAL REGULATION AND CRISIS MANAGEMENT
Crisis intervention poses a classic and clear pre-commitment problem. A crisis is a systemic run on many banks at the same time. Once a run has started, creditor bailouts are essentially the only way to stop it. Creditors only stop running to get their money from banks if they know that the government will stand behind their deposits. Yet if the government always quickly bails out creditors, props up prices, and provides liquidity, then people, banks, and financial institutions have little incentive to avoid too much short-term borrowing, contain risk exposures, or keep some cash or balance-sheet capacity around to buy on price dips.
The government would like to pre-commit not to intervene, so that people take adequate care on their own. But once a crisis ensues, it will intervene to the extent of its power. People know that, so they do not take adequate care, and when the inevitable crisis hits, the government is forced to intervene.
Once again, independence alone appears a secondary issue for this pre-commitment problem. People will only believe the Fed (and Treasury) won't intervene if those agencies can't intervene and don't need to intervene. The Fed will not voluntarily abstain from intervening in a large crisis just to contain some abstract moral hazard next time around, no matter how independent it is. And it's not obvious that once a large crisis has started, failing to intervene is wise — 1933 was pretty bad.
Moreover, crisis management is necessarily political, involving bailouts, capital "injections," asset purchases designed to raise asset prices, and lending at below-market rates. Some people will get very rich; others will lose. Letting a crisis run its course is equally political.
So crisis fighting is not plausibly technocratic and apolitical, as the decisions of a fully independent agency need to be. Indeed, the Fed habitually works closely with the Treasury in such cases. The tradition that the Fed lends to a special-purpose vehicle in which the Treasury takes credit risk is a useful arrangement. Formal restrictions, which Congress can suspend, are likely to be much more effective pre-commitments.
But do not think only of extreme systemic crises such as 1929-1933 and 2008. The trouble is that to the Fed, every hiccup seems like a crisis. "Systemic risk" once meant systemic runs on short-term debt; today, it appears to mean that someone, somewhere, might take a mark-to-market loss on a portfolio, might find they have to roll over short-term financing at unattractive rates, or might not fully pay their creditors.
We need an institutional structure in which the Fed and Treasury can pre-commit to less frequent intervention, both large and small. The reform must combine simpler and more effective financial regulation so that they don't need to intervene with formal pre-commitments not to intervene, and with the usual break-glass-in-case-of-emergency exception. Financial institutions that hold risky assets should get far more of their money by issuing equity and long-term debt, so that runs cannot break out. When it comes to achieving this goal, independence is a weakness. It is unlikely that an independent institution will institute such politically contentious changes on its own. Nor, arguably, should it be able to do so.
Going one step further, it is hard to argue that the Fed's regulation and supervision of individual banks and other financial institutions, outside of any objective "systemic" concern, benefit from much greater independence than agencies like the Securities and Exchange Commission, the Commodity Futures Trading Commission, the Office of the Comptroller of the Currency, the FDIC, and so forth. Yes, it would be nice if financial regulation were more "insulated from political interference," but it would also be nice if it were more "subject to democratic accountability." Pundits favor one or the other depending on which party is in power. However, we should not expect any immediate advantages if the Fed's financial regulation and supervision becomes only as independent as the other regulatory agencies, or even as non-independent as the Treasury Department. None of those are, after all, exemplars of efficient, transparent, apolitical, social-welfare maximizing, technocratic regulation founded on clearly understood market failures.
COSTS OF INDEPENDENCE
It is common in academic economics to applaud independent central banks. There is empirical evidence that independent central banks lead to lower inflation and greater economic growth. However, much of that evidence comes from the early years of each central bank's independence, not those banks' middle age, when institutional waistlines tend to spread. The creation of independent central banks usually came with a broader reform, which partially accounts for good subsequent economic outcomes.
And independence comes with costs. Central banks can be too independent. Agencies that are more independent are more prone to overreading their mandates, ignoring their limitations, and accumulating more power. They are also vulnerable to the creeping sclerosis that plagues all large organizations. The failures of the 1933 Fed, which led to a disastrous bank run, came in part from too much independence. Milton Friedman and Anna Schwartz's monumental history of monetary policy is in many ways a forceful critique of independence, showing how an independent Fed often made terrible decisions. They argued for less independence, in the form of monetary policy rules.
Independence is no panacea. Independent central banks can make mistakes. Other than the famous 1972 Nixon-Burns anecdote (which turns out to have been exaggerated), the 1970s Fed is not usually thought of as lacking independence, wishing less inflation, or pushed by politics to inflate. The Fed simply made policy errors. Inflation declined and the "great moderation" broke out under the independent Fed of the 1980s, but this was only possible because the Fed had the Reagan administration's support. It would not have been independent enough to do it alone. The complaints listed at the outset of this essay all emerged under the current independent Fed. It's not obvious that more or less independence would have helped.
Agencies that are more independent of pressure from the administration or Congress are likely to grow cozier with regulated industries. This temptation is especially strong with the Fed, as the banks are (famously) where the money is.
More independent institutions, anxious most of all to guard their independence, powers, budget, and institutional prestige, are also likely to be more timid in reform than institutions that must bend to the reality of electoral change. Witness how long it takes the Fed to change strategies, or its inability to contemplate mild reforms to clearly ill-functioning financial regulations. Indeed, all administrative agencies are notoriously difficult to contain or reform, even when they are headed by rotations of political appointees. Fundamental reforms to monetary policy or financial regulation will surely have to come from legislation or energetic executive action.
AN AGENDA
So, given the list of complaints, and given the much-changed environment, what does a good set of reforms look like? How do these reforms balance greater or lesser independence, greater or lesser accountability and oversight, and greater or lesser limitations and mandates? I phrase the following as answers, but one should really take them as questions requiring debate and consensus.
Given the dangers of plausibly too-independent central banks that we see elsewhere in the world, our Fed's overall level of independence versus accountability seems about right. Rather than strengthening the institution's independence or subjecting it to more direct political control, the Fed needs strengthened mandates, limitations, and accountability. More clearly living within its mandates will allow for a renewed independence. These are largely questions for Congress to decide in the end.
The Fed's mandates — currently price stability, employment, and financial stability — need to be spelled out more explicitly. They must also be limited, to ensure that the implied "and nothing else" means "and nothing else." Congress, not the Fed, should decide what "price stability" means — whether the Fed should target the inflation rate or the price level, and whether it need only announce a goal or be accountable for the outcome measured over a given period. Holding the Fed accountable to the achievement of its mandates, as well as ensuring it doesn't exceed its mandates, should be a regular part of congressional oversight. I favor a price-level target and dropping the employment mandate, but again, the questions matter more here than the answers.
The concept of a financial crisis requiring a Fed response must be confined to a systemic run on short-term debt. It should not mean that some voter might lose money, or that some business might suffer bankruptcy reorganization. Individuals and individual private institutions are not inherently "systemic." The Fed and Congress need to ensure that bailouts are not frequently needed and that the Fed is not able to pursue them absent specific authorization from the political branches.
The "and nothing else" element of the Fed's mandates requires an explicit understanding of where the Fed's responsibility stops. The Fed should eschew responsibility for long-term interest rates, credit spreads, stock prices, mortgage rates, and other prices, and avoid interfering in those markets. Trying to spot and correct asset-market "bubbles" or "manage the credit cycle" should be beyond its mandate. The European Central Bank (ECB)'s view that controlling long-term rates and sovereign spreads is vital to address market "dysfunction" or "fragmentation" that impede "monetary policy transmission" goes far beyond any solid understanding of that transmission.
Limits on the Fed's tools need review. The Fed should retain authority over short-term interest rates, achieved by paying interest on reserves, discount window or other lending, or repurchase agreements. Some have put forth good proposals for changing the basic operating procedure, but they are a subject for another day. The Fed's habit of paying banks higher interest than other depositors needs review.
It is also time to rein in the Fed's asset purchases outside of crises designated by Congress. Quantitative easing was an interesting experiment. But maintaining a permanent large-asset portfolio of long-term Treasury and mortgage securities in an attempt to lower those interest rates embroils the Fed in fiscal policy and credit allocation to a greater degree than it does any good. If the Fed needs a large balance sheet, it can fund that with short-term Treasury debt. The Fed can also, like the ECB, create reserves from thin air via collateralized lending to banks, counting the loan as the corresponding asset. If what in the end is taxpayer money is going to be used to buy mortgages, the Treasury, not the Fed, should do that.
The Fed and the Treasury need a new accord on who is in charge of the maturity structure and hence interest-rate risk of Treasury debt. Since this is a fiscal decision, it should be the Treasury. To the extent that the Fed takes duration or credit risk temporarily into its portfolio, it should swap it back to the Treasury promptly. The Fed's job would be a lot easier if the Treasury would issue fixed-value floating-rate electronically transferable debt to reduce the demand for the Fed to buy the Treasury's securities and transform them into that flavor of debt.
Other countries, including Switzerland and Japan, have even larger and riskier balance sheets. But they seem to have political arrangements that allow a central bank to tread much more deeply into fiscal policy. Conversely, a balance sheet that excludes Treasury securities, or holds only indexed securities, may be useful someday if fiscal-monetary tension heats up, but not yet.
Congress should not impose an interest-rate rule, such as the Taylor Rule, as a legislative restraint. Rules are a useful benchmark for the Fed's reporting on how it is achieving its mandates. Each time is different, and solid and unchanging knowledge of how monetary policy works is not sufficient to restrict how the Fed should achieve its mandate. For these reasons, we need a central bank in the first place, which exercises limited and independent discretion, rather than a mechanical rule like the gold standard. Congress should say where we're going, not how to get there.
Financial stability, crisis management, and monetary policy are intertwined. To be credible, the lender of last resort and crisis manager must be able to print money. Only a potentially infinite supply can decisively communicate "whatever it takes." But too frequent last-resort lending with printed money creates inflation. If, when the Fed wishes to raise interest rates to quash inflation, all the undercapitalized, over-leveraged, and too-big-to-fail financial institutions need bailouts, the Fed will be unable to move. The SVB fiasco could return on steroids.
The temptation to implement monetary policy via financial regulation remains, especially as suspicions grow that short-term interest rates are not that effective. For example, credit-centric views of the Fed's effect on the economy tempt the Fed to manage credit directly. These are also attractive to those who applaud the Fed's expansion to general macroeconomic and financial management and welcome every new "tool" that raises the Fed's power. When monetary policy requires more or less stimulus, regulators can tell banks to lend more or lend less, or adjust capital requirements, as regulators once adjusted reserve requirements. In 1980, the Fed and the Carter administration imposed credit controls to restrain the economy — a move generally regarded as a failure. Granting the Fed credit-management authority is a bad idea. But again, this is a central issue in defining the Fed's mandates and limits.
The Dodd-Frank Act has failed. Congress should burn it and start over. At a minimum, Congress can demand that the Fed periodically review its subsidiary regulation under Dodd-Frank to determine whether regulations are functioning as designed. The Fed could do this on its own, but it has been very slow to implement even obvious reforms, and is unlikely to do anything fundamental.
Congress can demand greater accountability in general, and a revised mandate offers a framework for doing so. The Fed does not voluntarily undertake any cost-benefit analyses or risk assessments of its monetary policy and it ignores fiscal impacts of quantitative easing.
Given that monetary and fiscal policy will face off in the years ahead, the Fed needs stronger pre-commitments against fiscal inflation. Simple independence will not be enough, as we saw in 2020. Congress's desire not to see inflation, with the concomitant promise that Congress will sooner or later figure out how to repay debts rather than demand inflation, is a crucial part of the deal. A formal congressional commitment to repaying debt at the price-level target — by a "debt brake," for example, that ties the overall budget to the price level and the debt to GDP ratio — would be ideal. It would then be up to Congress to declare an emergency, and say "we are willing to risk inflation, but the fiscal situation is dire. We direct the Fed to keep interest costs on the debt low and monetize debt," as it did in the 1940s.
CONGRESS-LED REFORM
The bottom line is this: Yes, the Fed should remain independent. But the Fed cannot remain independent with its currently large and political footprint. Refreshed independence must be balanced by a limited mandate, limited tools, and stronger accountability. Only this package will address the Fed's expansion into fiscal and political territory while maintaining and enhancing the pre-commitments needed for successful monetary policy and financial regulation.
We must take seriously the problem of a vastly expanded Fed, one that has moved into fiscal and political territory. Crying "independence" and defending the status quo will not work forever. If the genie is not returned to the bottle, the Fed will necessarily drift into the orbit of the imperial presidency.
Reform will require thoughtful congressional action. There is a tendency in public discourse to view Congress as dysfunctional, bitterly divided, partisan, and unable to get anything done, and thus to advocate more energetic if questionably legal or constitutional administrative action, court intervention, or greater independence for agencies including the Fed.
It is also true that our government has difficulty separating rules — creating an institutional structure that should last another 100 years through trial and tribulation — from scoring short-term political advantage. On the other hand, Congress did create the Fed, with an impressive and well-considered institutional structure, and it has reformed and improved it several times. It can do so again.
Our job is to find long-lasting general principles so that when the time comes, the library of ideas is not empty. Though we may say "Congress is dysfunctional," we should not rush heedlessly to install an autocracy (stronger presidential control) or aristocracy (great power in the hands of an unresponsive elite) in place of the checks and balances of representative democracy.
