Abstract: The Federal Reserve System has come to occupy center stage in the formulation and implementation of national and global economic policy. And yet, from the perspective of legal theory, the mechanisms through which the Fed acts autonomously from other governmental actors are barely analyzed. This article undertakes that analysis and describes the way that law does—and, more interestingly, does not—shape the Fed’s vaunted independence. The article demonstrates that nothing within the Federal Reserve Act’s independence-enhancing mechanisms are as they seem, including the Fed’s self-funding, participation of private banks’ representatives on the Fed’s monetary policy-making committee, the long tenures of the members of the Fed’s Board of Governors, and other examples frequently invoked by scholars and policy-makers to explain—incorrectly or incompletely—the Fed’s independence. In the process, the article challenges the prevailing accounts of independence in administrative law, economics and political science, all of which focus on statutory mechanisms of creating policy-making space between the central bank and other governmental institutions. Fed independence is not, then, simply a creature of statute, but an ecosystem of formal and informal institutional arrangements, within and beyond the control of the actors and organizations most interested in controlling Fed policy.
1.The Independence and Operations of the Federal Reserve System: Structure, Theory, and Law 5
2.What the Fed Is, What the Fed Does 6
2.What the Fed Is, What the Fed Does 6
3.Insufficient Approaches to Evaluating Fed Independence: Law, Economics, and Political Science 9
3.Insufficient Approaches to Evaluating Fed Independence: Law, Economics, and Political Science 9
4.Agency Independence 11
5.Central Bank Independence 13
6.Independence in Political Science 15
7.The Institutions of Federal Reserve Independence 16
7.The Institutions of Federal Reserve Independence 16
8.Congress and the Fed: The Curious Case of the Fed’s Budgetary Independence 19
A.Structure of Fed Budgetary Independence 20
9.Statutory Basis for Fed Budgetary Independence 22
9.Statutory Basis for Fed Budgetary Independence 22
10.The Compromise of 1913 and the Quasi-Autonomy of the Federal Reserve Banks, 1914-1935 22
10.The Compromise of 1913 and the Quasi-Autonomy of the Federal Reserve Banks, 1914-1935 22
11.Open Market Operations Under the Gold Standard and Real Bills Doctrine 27
11.Open Market Operations Under the Gold Standard and Real Bills Doctrine 27
12.Scholarly Engagement with Fed Budgetary Independence 28
12.Scholarly Engagement with Fed Budgetary Independence 28
13.Implications of the Fed’s Budgetary Autonomy 30
13.Implications of the Fed’s Budgetary Autonomy 30
14.The President and the Fed 33
A.Chair-President Independence 33
1.The Law: Removability of the Chair 33
15.Informal institutions of Chair-President Independence 36
16.The Chair-centric Federal Reserve 38
17.Appearances of Independence 39
18.Governor-President Independence 41
18.Governor-President Independence 41
1.The Myth of the Fourteen-Year Term 41
19.Independence from the Treasury 43
19.Independence from the Treasury 43
1.Legal Independence 44
20.In Practice: The Fed-Treasury Accord Independence 44
21.Implications: Presidential Control of the Federal Reserve 45
21.Implications: Presidential Control of the Federal Reserve 45
22.Private Banks and the Federal Reserve 46
A.The Legal Institutions of Fed-Bank Independence 47
23.The Non-Legal Institutions of Fed-Bank Independence 49
23.The Non-Legal Institutions of Fed-Bank Independence 49
1. New York Fed and the FOMC 49
2.Regulatory Capture 50
24.Political Power of the Reserve Banks 51
On December 23, 2013, the Federal Reserve System celebrated its centennial. Over the course of that century, the Fed1 has become one of the most important governmental agencies in the history of the American republic, a transformation one scholar has labeled “the most remarkable bureaucratic metamorphosis in American history.”2 Its policies influence nearly every aspect of public and private life. Given this importance and influence, “[n]o one can afford to ignore the Fed.”3
At the core of that “remarkable bureaucratic metamorphosis” is a much-invoked but as often misunderstood set of institutional arrangements that constitute the Fed’s unique independence. In the standard popular and academic account, law is at the center of that independence: indeed, it is the statute itself, under this view, that defines that independence. Economists and political scientists interested in central bank independence—having written enough on the phenomenon to give it an acronym (CBI)4—take as given that law is central bank independence.5 And legal academics, in the exceptional event that they have taken note of the Fed,6 have analyzed its independence within the context of administrative law7 and agency independence generally.8 Again, unsurprisingly, law is at the center of that analysis, too.
The idea that Fed independence is determined by law is wrong.9 That focus—by scholars and judges—provides a verisimilitude of Fed independence without capturing its complexity, and, in that way, both obscures the reality of the Fed’s policy processes and impedes serious scholarly and policy discussion about the ends and means of central bank design.
This article—part of a broader project10—argues that the focus on law assumes too much, and analyzes too little. Drawing on the language, structure, and history of the Fed eral Reserve Act of 1913 (especially as amended in 1935), other legislative materials, memoirs and biographies of Fed Chairs and other insiders, and other archival resources, as well as a secondary literatures from law, history, economics, and political science, the article provides a more comprehensive account of the legal and non-legal institutions of Fed independence. By institutions, I mean, following Douglass North, the “humanly devised constraints that structure political, economic and social interaction. They consist of both informal constraints (sanctions, taboos, customs, traditions, and codes of conduct), and formal rules (constitutions, laws, property rights).”11 The consequence of placing the legal and non-legal, formal and informal institutions of Fed independence side by side is the demonstration that the Fed is not as independent from politics as many have presumed, but nor is it beholden to closed-door pressures the way that detractors have insisted. Instead, Fed independence has evolved as a unique ecosystem that is both more and less than the standard academic and popular accounts assume.
The article’s main contribution is substantive: by examining every feature of the Federal Reserve Act that arguably creates independence—or its reciprocal, accountability—the article demonstrates the law’s subtlety and, sometimes, its irrelevance. For example, the Fed’s (in)famous budgetary independence is a different creature than the statutory framework originally created in 1913 and still on the books today. Instead, the changing context of monetary policy over a century has created a very different—and much more substantial—system of budgetary independence. On the other hand, the length of tenure for the members of the Board of Governors—often invoked as an example of Fed independence—is, by informal practice, essentially a dead letter. Although the statute makes clear the opposite policy goal, every President since 1935 has essentially named his Board of Governors. The uncritical tally of statutory provisions that usually accompanies the CBI and agency-independence literatures misses these complexities; an analysis of the institutions of Fed independence, broadly construed, does not.
The article’s value for legal theory is also important. The argument is not just about the chasm between law on the books and law on the ground, but is an argument in favor of analyzing the history and evolution of the interaction between them: the institutional development of Fed independence relies on statutory authorization, individual interpretation, and the subtle but steady drip of change exerted by individual personalities, outside forces, and the role of chance. In this way, the more comprehensive account of the institutions of Fed independence challenges the legal theory of agency independence by arguing against the durable, formal structure of that independence that prevails in courts.
This more comprehensive account of Fed independence comes at an important time in the history of American central banking. The financial and economic policies the Fed has undertaken during the financial crisis, in its aftermath, and now as it prepares to unwind those policies have generated criticism in Congress,12 the academy,13 the press,14 and in courts.15 Whatever one’s view of these policies, understanding the role of law in influencing—or not—the space within which the Fed operates is essential for anyone, whether in defense of the legal status quo, or in favor of legislative change.
This article proceeds as follows. Part I provides the context for the debate by outlining the structure of the Fed, describing the practice of monetary policy, and explaining why Fed independence is such an important—and controversial—topic. Part I.B. then explains the basic academic approaches to central bank and agency independence already in place, including their overreliance on (1) the relationship between the Fed and the President (and, to a lesser extent, Congress), and especially (2) the unchallenged assumptions about the centrality of law in creating that independence. Part I.B. then explains how Fed independence fits within, and in some ways pushes against, the framework of institutional change developed by North and others working within new institutional economics.
The rest of the article evaluates the way that formal and informal institutions separate, or don’t, the Fed from three outside groups: Congress, the President, and private banks, mediated by the Reserve Banks. Part II outlines the relationship between the Fed and Congress, with specific focus on the Fed’s budgetary autonomy and the circuitous history of the practice of open market operations and the consequences for Fed budgetary independence. This Part explains how this budgetary autonomy came to be, why scholars have missed or mischaracterized this independence, and why it matters.
Part III discusses the Fed’s relationship with the President, with specific focus on (1) the relationship between the Chair and the President as a matter of law, personality, and popular imagination; (2) the law and practice of Governor appointments meant to limit the President’s control of the Board, but in practice a means of extending that control; and (3) the Fed’s relationship with Treasury, and what that history means for legal and non-legal institutions of independence.16
Part IV then looks at the Fed’s relationship with the private banks it regulates. This is an important starting point, developed further in other work17: that central bank independence has become a shorthand for independence from government is an ahistorical interpretation that focuses on only one half of the Compromise of 1913 that created the System as a balance between private and public interests. Part IV explains more about that Compromise, and then identifies the legal and non-legal institutions of independence between the Fed and private banks, mediated largely by the Reserve Banks. The role of the Reserve Banks in the System is crucial, even after their authority was diminished in 1935, and is frequently overlooked. Part IV explains how the private banks and Reserve Banks—through legal and non-legal means—influence Fed policy politically and substantively.
The article concludes by explaining why a more nuanced understanding of the institutions of Fed independence matters, irrespective of one’s views of the Fed’s policy decisions of the last decades, and explains, briefly, how the article fits within a broader research agenda.
Scholars and public commentators frequently assert that the Fed is or is not independent. One goal of this article is to expose as incoherent such a binary characterization of the Fed. To argue that the Fed is or is not independent is meaningless without further specification as to how it is independent, to what end, and from whom. This article takes up the “how” question, and argues that the prevailing answers to it should be dismissed.
1.The Independence and Operations of the Federal Reserve System: Structure, Theory, and Law
The broader project, of which this article is a first contribution, defines Fed independence as the exercise of control over the Fed’s policy processes—whether monetary policy, systemic risk regulation, bank supervision, bank regulation, etc.—from internal and external audiences by means of mechanisms, both legal and non-legal. This article’s focus is on the role of law within that policy-making space, and the argument is that the law as defined in the Federal Reserve Act doesn’t function as either anticipated by legislative drafters or as described by academic and popular commentators.
In order to understand how this definition challenges and improves upon the existing literature, one must understand the nature and structure of the Federal Reserve System itself. Part I.A. provides that overview, focusing on monetary policy. Part I.B. then briefly surveys the agency and central bank independence literatures, and the structure and process literature in political science, each of which, in its way, focuses on the relationship between the administrative state—including central banks—and the President and/or Congress. It also shows how much of those literatures focus on the law, at the exclusion of other institutions of independence. Part I.C then explains the more robust approach undertaken in this article.
2.What the Fed Is, What the Fed Does18
The original Federal Reserve Act created the Federal Reserve System, which consisted of the Federal Reserve Board, based in Washington, D.C.; Federal Reserve Banks in what would become twelve cities throughout the country; and member banks, which were private, commercial banks that subscribed to the stock of Reserve Banks and gained access to the System’s regulatory apparatus.19 In 1933, Congress created the Federal Open Market Committee (FOMC), to be the System’s monetary policy-making arm (thereby replacing the only sometimes coordinated monetary policy efforts of the Reserve Banks). Under the Banking Act of 1935, the System was fully reorganized—effectively refounded—and the Federal Reserve Board replaced by the Board of Governors of the Federal Reserve System. At the same time, the FOMC was refashioned to include all seven members of the newly created Board of Governors. Today, the remaining seats on the FOMC are filled by the President of the Federal Reserve Bank of New York and four of the eleven other Reserve Bank Presidents on an annually rotating basis.
The FOMC controls monetary policy using one of four basic mechanisms, one of which came into use only recently. These tools are the federal funds rate, discount rate, reserve requirements, and what is called “quantitative easing.” The control of the federal funds rate is recently the most important and most commonly used of the Fed’s tools. The federal funds rate refers to the rate at which banks lend money to each other, usually for short-term loans (overnight, or slightly longer). The effective federal funds rate is the average of these rates reported by the banks. The FOMC, in its eight annual meetings, establishes the target federal funds rate, or the rate it wishes to see in the markets for interbank, short-term loans.
To reach this target, the Fed buys or sells securities on the open market, through the trading desk at the Federal Reserve Bank of New York. When the FOMC decides to raise interest rates, it sells securities; when it decides to lower interest rates, it buys securities. To understand why, one must consider the role the Fed plays in establishing and maintaining the money supply.
When the Fed buys a Treasury security on the open market, it provides its counter-party with cash—an electronic modification to the counter-party’s balance sheet. This purchase removes the security from a bank’s balance sheet, and replaces it with greater reserves in the bank’s account at its local Federal Reserve Bank. In this way, the Fed has expanded the money supply by removing from the banking system a more illiquid asset—the government bond or, more recently, the mortgage-backed security—and replacing it with cash, the most liquid of assets. If the bank already had the requisite level of reserves required by the Fed (more on that momentarily), the cash that now sits on the member bank’s balance sheet is something extra. Banks are generally in the business of taking the “something extra” and injecting it into the economy in the form of bank loans. Under normal conditions, the bank will lend the majority of what it has received from the Fed in exchange for its more illiquid security, expanding the money supply in the economy as the bank borrower spends that money and multiplies the money’s reach through a daisy chain of spending, investing, and saving. Because most consumers of bank credit—usually businesses, but also individuals—don’t carry around much cash on their persons or under their mattresses, the money lent by the Fed’s initial counter-party to another person or institution will eventually end up in another bank, who will then lend to another individual or institution with the same consequence. The effect is a more or less predictable expansion of the money supply throughout the banking system.
When such a monetary expansion occurs, banks start to feel flush. Projects that otherwise would not get funded, get funded. People who would otherwise not get loans, get loans. The result, under the best conditions, is economic growth. But under other conditions, what looks like economic growth is, in fact, inflationary pressures that threaten to undermine the economy’s stability and consumer confidence in the level of prices and wages. When the Fed fears that loss of confidence and senses that inflation, not growth, drives expansion, it intervenes. In the oft-quoted metaphor cited by Fed Chairman William McChesney Martin, the Fed acts “in the position of the chaperone who has ordered the punch bowl removed just when the party was really warming up.”20
If the Fed becomes the chaperone, it restricts the money supply by doing the reverse of the purchase of the Treasury securities described above—it sells them to its counter-parties. When the Fed sells these securities on the open market, at the market price, it replaces cash on the bank’s balance sheet with a less liquid government security. In turn, the Fed’s counter-party bank will either call in loans due—or, more frequently, initiate fewer subsequent loans—thus either diminishing or slowing the expansion of the money supply.
So far this explanation has skated over the significance of interest rates. After all, the Fed has done nothing explicitly with interest rates–it has merely injected or retracted liquidity to expand or shrink the money supply. The connection between open market operations and interest rates is as basic as a supply and demand graph from introductory economics. Here, the supply and demand is for money, a commodity for which there is a market, just as there are markets for crude oil, pineapples, or squirrel traps. The price of money in these markets is the interest rate. When there’s less money, people will pay more for it—and the interest rates will rise. When there is more money, people will pay less for it, and the price of money drops. Thus, while the difference between the federal funds effective rate and the federal funds target rate is actually more complicated than this simple explanation suggests, the reality is that the Fed can and does affect interest rates through open market operations similar to the process described above: by affecting the quantity of money, it changes the price of money.
Two other monetary-policy levers are more easily explained, with this background in mind. The second is the lending that occurs through the figurative “discount window,” referred to today, more accurately, as the discount rate. The discount rate is the rate, set by the Reserve Banks subject to the approval of the Board of Governors, at which the Fed lends directly to the banks themselves.21 The original conception of the Fed was as the lender of last resort,22 and the discount rate was the mechanism by which the Fed might make these loans to an otherwise solvent bank in crisis. Historically, the discount rate was of far more importance to the maintenance of the banking system. Today, it has been almost completely replaced by open market operations, although in times of crisis—including the recent crisis—the discount window is much more actively used, not without controversy.23
The third lever is the reserves requirement with which all banks—whether members or not of the Federal Reserve System—must comply.24 The Board of Governors can increase or decrease that rate, and by so doing, increase or decrease the money supply. There is little dramatic change, though, that a prudent Fed can do with reserve requirements: ratcheting up reserve requirements too high is unnecessary, since a more finely-tuned increase of rates is always possible, and decreasing reserve requirements too much only exposes individual institutions to idiosyncratic chances of default.
The final lever is the newest—and, with the potential exception of the direct loans through the discount window to non-banks in times of crisis, the most controversial—addition to the Fed’s toolkit. Called, euphemistically, quantitative easing, it is essentially the answer to the perplexing question of what a central bank can do when interest rates are already at zero, and yet there is inadequate economic expansion and indeed risk of deflation, rather than inflation. In such cases, conventional open market operations to influence interest rates are useless—conventional interest rates cannot go below zero.25 But the continued purchase of securities on the open market can have the same effect of injecting liquidity into the economy when that floor is reached. Quantitative easing is thus the increased purchase of these assets in order to inject even more money into the system. The hope of quantitative easing is that the injection of this amount of money will do what lowering interest rates to zero could not do—namely, get the economy moving again.
3.Insufficient Approaches to Evaluating Fed Independence: Law, Economics, and Political Science
The conventional justification for Fed independence is that the process just described—changing the quantity of money and thereby influencing interest rates throughout the economy—is necessarily a controversial exercise.26 Under the classic formulation, creditors in society prefer to see higher interest rates and lower inflation; debtors prefer to see lower interest rates and higher inflation. It is entirely because of the authority to adjust these interest rates—which necessarily influence how much it costs the government to service its debt, Jane Doe to pay for a mortgage or student loans, or the relative attractiveness of investments in the stock market—that makes the decisions and institutional design of the Fed so controversial. Society must be able to assume that those monetary levers are pulled for reasons other than a politician’s desire to inflate away the public debt, to cater to some electoral interest, or to pursue pure venality.
The debate regarding why politicians would ever cede even partial control over that money-regulating process—in other words, why politicians would ever create independent central banks—is ongoing.27 So too is the debate about to what end, the regulation of money, whether price stability, economic growth/employment regulation, systemic risk regulation, or some combination of the three.28 The article will leave to the side that debate, and instead focus on the mechanics of independence. In other words, the questions are from whom is the Fed be independent, and how is that independence accomplished and maintained?
There are three literatures of research that provide insight into these last questions: agency independence in law, central bank independence in economics and political science, and structure and process theory in political science. All three are useful starting points. But all three are either focused on different questions—the constitutional contours of appointment and removability for law, the empirical consequences of legal separation between central banks and the government for CBI, the consequences for specific features of institutional design on agency performance. Or, more provocatively, they miss the nuance of the institutions of Fed independence by focusing exclusively on law, at the expense of non-legal, informal institutions of independence.
Courts and legal scholars have long analyzed the nature of agency independence. But this is something of a misnomer: as Gersen noted, agency independence is a “legal term of art in public law, referring to agencies headed by officials that the President may not remove without cause. Such agencies are, by definition, independent agencies; all other agencies are not.”29 Thus, “agency independence” is not concerned with agency independence in the generic sense of that term—whether the agency can pursue its own agenda without outside interference—but only whether the President can summarily fire the agency’s head.
Other scholars have documented the removability focus in administrative law’s historical development,30 but the doctrinal gist is easily summarized. Congress may not require the President to seek Senate advice and consent prior to removal, as the “reasonable construction of the Constitution” would forbid that kind of blending of legislative and executive functions without express authorization.31 But Congress may condition Presidential removal of an agency head to a more limited range of causes, depending on the nature of the office in question. For offices that are created to “perform . . . specified duties as a legislative or as a judicial aid”—that is, independent commissions like the Federal Trade Commission—the Court deemed removability conditions on agency heads constitutionally permissible.32 So too for lower-level executive appointees like the independent counsel,33 but not if the agency head and the lower-level appointee are both deemed to be protected by for-cause removability protection.34
As a quick-and-dirty overview, the doctrinal summary isn’t very satisfying. The point is only that some kinds of restrictions are permissible, some are not, and the meaning of agency independence for judicial purposes is narrowly circumscribed within that President-and-removability framework.35
On their own terms, these doctrinal conclusions are controversial to scholars of presidential authority on each side of the cases just summarized.36 But as a means of evaluating agency independence writ large, the removability focus is even more susceptible to criticism: that is, leave aside the questions of a unitary executive, and focus instead on the meaning of agency independence, and one sees immediately why the removability focus captures only a small part—and, this article argues, a not very important part—of the agency independence picture.
A growing chorus of scholars has challenged the removability paradigm as too narrow. For example, scholars contend that the paradigm: focuses on the wrong mechanisms of independence,37 ignores the ways in which executive agencies (i.e., those whose heads are removable at will and, separately, are subject to Presidential regulatory review through the Office of Information and Regulatory Affairs) use presidential review to increase “self-insulation,”38 creates meaningless distinctions between executive and independent agencies,39 is focused on the wrong problems40 and the wrong parties,41 reflects a misunderstanding of how the administrative state actually functions,42 elides ways in which the President controls independent agencies beyond removability,43 and gives to courts review of decisions that are fundamentally incompatible with judicial review.44 Vermeule summarizes the point well. Identifying a “mismatch” between “the doctrinal law as embodied in judicial decisions and the revealed behavior of political actors,” he notes that “the legal test that courts deem central to agency independence is neither necessary nor sufficient for operative independence in the world outside the courtroom. The legal test . . . does not capture the observable facts of agency independence in the administrative state.”45
The Federal Reserve’s independence illustrates some of the scholars’ frustrations with removability as a paradigm for comprehensive evaluation of agency independence, as some scholars have noted.46 Indeed, the relationship between the President and the FOMC especially touches on appointment and removability in ways that scholars have not fully analyzed.47 But otherwise, the removability paradigm points only to one small part of the phenomenon of Fed independence, and is essentially useless (as will be argued more fully below)48 at clarifying the extent of Fed independence.
5.Central Bank Independence
Although legal scholars have mostly either ignored the Fed or analyzed it in conjunction with other agencies of very different stripes, economists and political scientists49 have long focused on the inputs and outputs of central banks and central banking.50 Interestingly, although the CBI and agency independence literatures rarely overlap, their conceptions of independence are strikingly similar. While the term “independence” in the CBI context has meant “different things to different people,”51 the focus is, as with agency independence, almost exclusively on the legal mechanisms that separate the central bank from interference by the political branches, especially the executive.52 That literature is summarized in the footnotes below.
There is an important conceptual focus to CBI that focuses more on the why of independence—to what end—than on the mechanics of Fed independence. Under Stanley Fischer’s now famous articulation, CBI is divided between “goal independence” and “instrument independence.”53 Goal independence refers to the freedom to select the ends of monetary policy; instrument independence is the freedom to select the means of pursuing statutorily specified goals.
Fischer’s formulation has been, for the most part, the last word on those mechanical questions.54 And even there, the focus is mostly on the statute.55 It is the law, passed by Congress and recorded in the U.S. Code, that establishes the “goals” of central banking; and the law, passed and recorded in the same way, that provides the freedom to select the “instruments” of central banking. And it is the law that empirical economists cite when they attempt to determine the extent of independence and the correlation between that independence and economic indicators such as GDP growth, inflation, and unemployment.
6.Independence in Political Science
Political scientists looking at agency design have explored more contours of the administrative state, even as they have, for the most part, focused on law as the primary mechanism in that design. An earlier generation of scholars of the bureaucracy viewed the question of congressional delegation to agencies under the view that delegation was abdication, the creation of a “headless fourth branch” that controlled governmental decision-making without accountability.56 But in the 1980s and 1990s, first McNollGast57 and other scholars58 introduced a renewed confidence that Congressional design can shape the goals and behavior of agencies. This work came to be known as the “structure and process” approach to agency control. As Gersen summarizes, “[a]lthough the structure and process thesis now has many variants, its simplest form asserts that legislatures can control agency discretion (policy outcomes) by carefully delineating the process by which agency policy is formulated.”59
While administrative law looks to removability restrictions to define agency independence, the structure-and-process theorists focused on the variety of mechanisms available to Congress through which that control might be asserted. That focus is therefore inherently on institutional design, whether via ex ante legislative decisions of Congress60 in its role as institutional designer, or more broadly on, well, the structures and processes associated with a given agency, whether legislative or executive.
The structure and process thesis is an important extension of the narrower focus on removability in administrative law. And in many ways, the new work in administrative law scholarship—as opposed to administrative law in courts—that challenges the judicial conception of agency independence, is expressly or by implication a subset of the structure and process thesis. But the focus in this context, as in administrative law generally and CBI, remains on law: for example, the nature of Presidential review of agency work product;61 threats and practices of auditing;62 limits on jurisdiction;63 and expansiveness of the wording of authorizing statute.64 Largely missing from the analysis is the role of informal institutions and their interaction with the legal mechanisms that are the traditional focus of these scholars.
7.The Institutions of Federal Reserve Independence
The agency-independence and central-bank-independence literatures both rely, then, on the flawed assumption that law is the defining feature of this independence. As Parts II-IV demonstrate, this assumption is inaccurate. To be clear, the flawed assumption does not eliminate the value of this scholarship. In many cases, the aim in these literatures is to assess the President’s and Congress’s constitutional prerogatives on the institutional design of the administrative state, or to establish a theory for testing the efficacy and implications of a central banks’ separation from specific governmental organizations, and the use of law to accomplish some end desired by Congress, interest groups, or the President. In that sense, the literatures can’t be criticized for failing to take a more comprehensive view of Fed independence because that view wasn’t a part of, or required for, that research.
But if the aim of these judges and scholars is to evaluate the ways that a “headless fourth branch” can exist outside the traditional structure of government, as many critics in the judiciary and the academy have expressed,65 or to quantify and evaluate claims that the Federal Reserve is or is not independent of other governmental actors or organizations, then the law-as-independence approach is incoherent, for two reasons.
First, the assumption of law’s centrality pays too much attention to the relationship between the Fed and the President, or the Fed and the rest of the government (in the CBI context). These dynamics speak only to a small—albeit important—part of the Fed’s independence. In reality, the Fed faces a number of “audiences,” to use Carpenter’s term. In his words:
Government agencies live among numerous audiences, and these audiences overlap and blend into one another. Audiences include the political and judicial authorities who endow organizations with power; interest groups and civic associations; organizations of professional and scientific expertise; media syndicates in print and broadcast, and the mass publics who digest the information produced by these syndicates; the companies, corporations, and citizens who are governed by agencies; the clienteles who rely upon agencies for benefits and for order. In political systems like the U.S.—with formal separation of powers among legislative, executive, and judicial branches; with federalist structures that multiply and refract government capacity; and with pluralist political structures that often scatter the forces of business, labor, religion, race, and ethnicity—these audiences stand ever more diffuse.66 So it is with the Federal Reserve: the audiences that surround, provoke, influence, and ultimately determine the shape of its policy processes are several, identifiable, and diffuse. One can say nothing about the Fed’s independence without specification of one or more such audience.
Of course, scholars writing in one vein of the public choice tradition have long focused attention on the ways in which the bureaucracy interacts with, for example, “the key committees and members of Congress and the private interest” in order to shape administrative policies.67 And there are, to be sure, examples of scholars who have looked closely at private influence on the Federal Reserve itself, although not directly within the framework of Fed independence.68 Barkow, in the administrative law context, is a notable exception. She has argued that agency independence primarily focuses on independence from private interests—that is, as a means to insulate the bureaucracy from agency capture—and not from governmental actions.69
This article, in this respect, builds on Barkow. If the theoretical inquiry for Fed independence is concerned with the Fed’s ability to pursue its policy course without dictation from outside parties, that analysis must combine references to audiences beyond the government. Future work engages more comprehensively with external and internal audiences that fight for control of the Fed’s policy space.70 For now, this article takes up the first aspect of that challenge by placing the role of non-governmental parties—specifically, the private banks the Fed regulates—squarely within the context of Fed independence.
The second problem with the legal approach to evaluating Fed independence is more important for present purposes. And that is that the near exclusive focus on law fails to engage with the law’s statutory and historical context and evolution. This is especially true in the administrative law context, where the removability of the Chair is the singular focus.71 But even a broader focus on a variety of legal mechanisms is inadequate. Extra-legal sources also circumscribe agency activities in a variety of ways. Here, Vermeule’s argument that conventions are distinct from law, but also shape the way that institutional independence is practiced and evolves, is important.72 To evaluate Fed independence, scholars and policymakers must be sensitive to these kinds of non-legal mechanisms.73 As demonstrated throughout the rest of the article, these other, non-legal institutions can be just as important, sometimes much more important, than the legal mechanisms identified by statute.
In that sense, the better approach is not the law of Fed independence, but its institutions.74 Here, North’s iconic formulation, quoted in the introduction, bears repeating: institutions are “humanly devised constraints that structure political, economic and social interaction. They consist of both informal constraints (sanctions, taboos, customs, traditions, and codes of conduct), and formal rules (constitutions, laws, property rights).”75 The key distinction is between the formal and informal, the legal and non-legal, but not just on those differences. It is also essential to note how the formal and informal institutions play off one another. In this sense, the article’s exploration of institutional change pushes the Northian view by deemphasizing the role of human engineering in the creation of these institutions, and focuses more on the role of slow, intermittent, sometimes unexpected change, against which individuals—be they Presidents, members of Congress, Fed Chairs, or others—interested in shaping institutional change must act and react, sometimes expressly, sometimes not. When taken together, it is the institutions of Federal Reserve independence that demonstrate how that independence is created, evolves, and influences policy formation and implementation.
8.Congress and the Fed: The Curious Case of the Fed’s Budgetary Independence
While the administrative law focus on agency independence is on the relationship between the President and the agencies, scholars have long analyzed Congress’s relationship to the Fed.76 But one of the primary means of Congressional control over agencies—the power of the purse—represents a unique and often misconstrued aspect of Congress’s relationship with the Fed. Part II discusses this largely uncharted statutory and historical framework: the Fed’s ability to fund itself from the proceeds of open market operations that it controls without interference from the political branches. This budgetary independence is not uncharted because it is unacknowledged. To the contrary, the Fed includes on its own website a frank admission that the System’s “income comes primarily from the interest on government securities that it has acquired through open market operations.”77 Instead, the story is an interesting one because of the interplay between the current practice and the statutory language authorizing this practice, separated as they are by a century of dramatic change in the conduct of monetary policy. As a result, although this feature of the Federal Reserve Act has been widely cited as a defining characteristic of Fed independence, the relationship between the Act and the current practice has caused scholars to mislabel or incorrectly analyze the Fed’s budgetary independence every time it has been addressed. Part II provides the first analysis, based in law and history, and illustrates how legal and non-legal institutions interact to create greater distance between Congress and the Fed.
To be sure, there are other important ways that Congress interacts with the Fed that shape the System’s independence. The statutory discretion afforded the Fed regarding the Fed’s dual mandate of price stability and maximum employment,78 Congressional hearings,79 and proposed legislation, not enacted,80 represent additional features that will be analyzed in future work.81 Part IV focuses extensively on the Fed’s budgetary independence given its prominence in securing the Fed’s independence from Congress and the mischaracterizations of that independence that have prevailed in legal scholarship to date.
Structure of Fed Budgetary Independence
The Federal Reserve is the only truly autonomous budgetary entity in the entire federal government, including the Congress and the President.82 To understand this dynamic, one must first understand how the rest of the federal government is funded, and compare it to the unique budgetary independence of the Federal Reserve.
There are three dominant forms of funding in government.83 First, the vast majority of governmental institutions—from the Congress to the Courts to the White House, and most agencies, institutions, programs, and commissions in between—is funded through Congress’s annual appropriations process.
Second, the majority of actual government expenditures do not occur through this appropriation process, but instead are part of the government’s mandatory commitments.84 These include entitlement programs such as Social Security, Medicare, Medicaid, and other forms of direct assistance; some kinds of disaster relief; and interest on the national debt. And third, some governmental agencies, including the Comptroller of the Currency, the Federal Deposit Insurance Corporation, and the now-defunct Office of Thrift Supervision, are funded through the fees assessed against their own regulated entities.
And then there is the Fed. The Fed funds itself with a portion of the proceeds from its open-market operations. In the Fed’s own words, from a recent budget report, “[t]he major sources of income were interest earnings from the portfolio of U.S. government securities and federal agency mortgage-backed securities in the System Open Market Account. Earnings in excess of expenses, dividends, and surplus are transferred to the U.S. Treasury—in 2009, a total of $47.4 billion.”85 The Fed also receives income for “priced services” provided to private banks, which include the cost of transporting and printing new currency, check clearing, and other services related to currency distribution and the general payment system.
Some scholars have supposed that the Fed, like some other banking regulators, funds itself through assessments on private banks.86 While true that the Fed collects money from member banks for “charged services,”87 such assessments, cover just 25% of its expenses.88 The rest, as stated, comes from the proceeds from its open market operations.89
That the Fed funds itself largely from the proceeds of its substantial assets, taken together with the nature of the Fed’s ability to create money in pursuit of its monetary policy objectives, means that the Fed’s funding is without parallel in the federal government. The Fed conducts monetary policy by, among other options, creating money with which it can buy government—and more recently, non-government90—securities.91 These interest-bearing assets generate money that the agency can subsequently use to fund itself.92 The Fed thus has the ability to create from nothing the money it eventually uses to pay its employees, funds its conferences, and renovate its buildings.
9.Statutory Basis for Fed Budgetary Independence
The Fed’s budgetary independence is thus without equal in the federal government. But here is the striking reality about this independence: It is not expressly authorized by Congress. The Federal Reserve’s funding mechanism is located in Section 10(3) of the Federal Reserve Act. That section grants the Board of Governors the
power to levy semiannually upon the Federal reserve banks in proportion to their capital stock and surplus, an assessment sufficient to pay its estimated expenses and the salaries of its members and employees for the half year succeeding the levying of such assessment.93 Unquestionably, this statutory authorization exempts the Fed from the Congressional appropriations.94 But, on its face, it merely allows the Fed to make the Reserve Banks pay for its expenses, “in proportion to their capital stock and surplus.”95 It does not allow the Fed to create, and fund itself with, its own Federal Reserve notes.
To understand how this relatively modest statutory authorization metamorphosed into the Fed’s present and complete budgetary independence, one must understand more about the evolution of the Federal Reserve System over the past century. Three features are of particular importance: (1) the quasi-autonomy of the Reserve Banks, terminated in 1935 when monetary policy came under the exclusive purview of the newly forged Board of Governors; (2) the Fed’s history with what was called the “real bills doctrine”; and (3) the Fed’s history with the gold standard.
10.The Compromise of 1913 and the Quasi-Autonomy of the Federal Reserve Banks, 1914-1935
The conventional story of the Fed’s creation describes an acute financial crisis in 1907, resolved by a bailout orchestrated by JP Morgan. As the story goes, the Panic of 1907 made bankers and politicians wary of continued reliance on the private bailout model. The Federal Reserve System was the political response to those concerns.96
This story is technically but deceptively true. The primary reason is that it links, almost ineluctably, the Panic of 1907 and the Act of 1913. For understanding how the Fed’s budgetary independence came to be, this uncritical link is a mistake. The Panic of 1907 occurred in, well, 1907; the Federal Reserve Act of 1913 in 1913. The six years in between were extraordinarily important for the fate of the Federal Reserve System, including as they did two Congressional elections in which Democrats first seized control of the House (in 1910)97 and then the Senate (in 1912).98 Most important, the presidential election of 1912—a four-way race between incumbent Republican President William Howard Taft, erstwhile Republican former President Theodore Roosevelt, Socialist Eugene Debs, and Democratic New Jersey Governor Woodrow Wilson—was one of the most significant elections in American history. In the words of one historian, the 1912 election “verged on political philosophy.”99 That political philosophical moment intervened between the Panic and the Act in ways that were essential to the ultimate shape the System took.
On a most basic level, the elections mattered because of partisan control. The first proposals following the Panic of 1907 were entirely Republican; the final bill was almost exclusively Democratic.100 Senator Nelson Aldrich was the Republican leading the monetary reform efforts. In 1908, Congress passed the Aldrich-Vreeland Act, which created the National Monetary Commission with Aldrich at the head.101 The Commission imagined a structure very different from the system the Federal Reserve Act eventually created. That structure, the National Reserve Association,102 was to be a mix of public and private appointments, but dramatically weighted toward the private. For example, the board of the NRA was to have forty-six directors, forty-two of whom—including the Governor and his two deputies—were to be appointed directly and indirectly by the banks, not by the government.103
The election of 1912 intervened, and capped a change of the partisan guard in the House, Senate, and White House, and the Democrats made the cause of monetary reform their own. The key consequence of this political transformation was what might be called the Compromise of 1913. Under that Compromise, the final result was the mostly supervisory, leanly staffed Federal Reserve Board, based in Washington, DC, and the quasi-autonomous twelve “Reserve Banks,” considered by several active participants in the Act’s drafting to be essentially private institutions.104
The tension between the two poles—public and private, accountable to the political process and independent from it—is essential to understanding the nature of Fed independence, then and now. Paul Warburg, the German-American banker whose ideas in the early 1900s set the stage for much of the debate preceding the enactment of the Federal Reserve Act, described it this way: “The view was generally held that centralization of banking would inevitably result in one of two alternatives: either complete governmental control, which meant politics in banking, or control by ‘Wall Street,’ which meant banking in politics.”105 One of the central debates preceding the passage of the Act centered on how to navigate those two poles, or the “whirlpool of socialism and the jagged rocks of monopoly.”106
The consequence of that Democratic navigation was the existence of the government-controlled Federal Reserve Board on the one hand, and the private Reserve Banks on the other.107 The System would not, in theory at least, be dominated by one faction or the other.
So it was that one of the consequences of this Compromise was that the Reserve Banks—not the Federal Reserve Board—was tasked with the conduct of what we now call monetary policy. The idea was that the System was a federalist one, with decentralized authority located in the Reserve Banks. Carter Glass, a zealous guardian of the Compromise,108 emphasized the federal in the Federal Reserve System in these terms:
In the United States, with its immense area, numerous natural divisions, still more numerous competing divisions, and abundant outlets to foreign countries, there is no argument, either of banking theory or of expediency, which dictates the creation of a single central banking institution, no matter how skillfully managed, how carefully controlled, or how patriotically conducted.109 E.W. Kemmerer, an early observer of the creation of the Fed, called the arrangement of “twelve central banks with comparatively few branches instead of one central bank with many branches” the “most striking fact” about the System.110 Glass shared the view of the Reserve System as a series of central banks; indeed, he did not view the Federal Reserve Board as in charge of the central banking aspects of the system at all.111
So it was that the Reserve Banks—those entities in charge, by statute, of generating the funds from which the Federal Reserve Board would, through semiannual assessment, fund itself, became financially autonomous organizations. Their financial autonomy meant that they raised money through the business of banking for banks: that is, “discounting,” or lending money at interest, to the member banks within the System. In the words of the first Secretary of the Federal Reserve Board,
The banks, in short, have all those banking powers that are not expressly mentioned in the Federal Reserve Act or directly implied as having been invested in the Federal Reserve Board. . . . There is nothing, either in the Federal Reserve Act or in the regulations of the Federal Reserve Board, to indicate that the reserve banks are to be operated in groups or through communication with one another, resulting in the establishment of a single policy as to detail. Neither is there any to prevent officers of the Federal Reserve Banks from communicating with one another, getting such information as can be exchanged by that means, or adopting their own policies as the circumstances and business needs of each district or of all appear to require.112 In other words, the Reserve Banks, not the Federal Reserve Board, controlled the purse strings under the original Compromise.
The original assessment provision of the Federal Reserve Act—which is identical to the provision in place one hundred years later—thus functioned via a Federal Reserve Board without an open market operations policy and without member banks to provide a source of income.
The assessment function, then, was from the Reserve Banks’ profits (including from the Banks’ open market policies) to the Board, not the Board to the Reserve Banks. And even though the Federal Reserve Board participated to a limited extent in shaping the tenor of monetary policy, the reality is that the Reserve Banks could and did pursue their own monetary policy.113
The era of autonomy for the Reserve Banks ended with the passage of the Bank Act of 1935, which placed the authority for open market operations of the individual banks into the Washington-based Board.114 The assessment provision, however, was unchanged.115
During the entirety of that early period, though, the Federal Reserve Board could assess the Reserve Banks for its expenses, including by using income generated through open market operations. But the Federal Reserve Board could not dictate the outlines of those operations. There existed therefore a separation between the assessment authority and the operations authority. The Fed in this early stage could not create the money with which it funded itself. This change in what constituted the identity and function of the Reserve Banks illustrates how the statutory funding mechanism has not kept pace with U.S. central banking practice.
11.Open Market Operations Under the Gold Standard and Real Bills Doctrine
Even if monetary policy throughout the Fed’s history had always been left to its discretion, the statutory authorization to levy assessments on the Reserve Banks would still be different from the authority the Board uses today. When the Act was passed, the United States was on the “gold standard,” a concept that in fact refers to a set of standards that, depending on the particulars, limit the central bank’s discretion in pursuing any given monetary policy.116 Under that regime, neither the Board of Governors nor the original Reserve Banks had the unlimited power to create the money the Board would assess from the Reserve Banks, and from which it would pay its own expenses.
So too with the real bills doctrine. While an important principle at the time, the real bills doctrine was not universally accepted, even during the Federal Reserve Board era. As Friedman and Schwartz indicate, “the real bills criterion . . . provided no effective limit to the amount of money.”117 This is because of the inherent difficulty in determining what counts as a “real bill” that a Reserve Bank can permissibly discount.118 As mentioned, not even every Reserve Bank practiced the principle, which has allowed for some fascinating comparisons between, for example, the Atlanta and St. Louis Reserve Banks, which each had oversight over different parts of the state of Georgia, but practiced different kinds of discounting techniques.
The point is not that the real bills doctrine actually provided no limit on which bills could theoretically be discounted. It is that (1) some people within the System perceived such limitations, and acted accordingly, and (2) that Congress thought that it was not granting to an institution unfettered access to money creation with which it could then, in turn, fund itself. Woodrow Wilson felt the same way:
Let bankers explain the technical features of the new system. Suffice it here to say that it provides a currency which expands as it is needed and contracts when it is not needed: a currency which comes into existence in response to the call of every man who can show a going business and a concrete basis for extending credit to him, however obscure or prominent he may be, however big or little his business transactions.119
Thus, Congress’s authorization to the Fed to levy assessments against the Reserve Banks under a gold-standard and real bills regime, when the Reserve Banks enjoyed autonomy to determine their own monetary policy, is radically different from the same authorization without those features. As one historian described it, the “automaticity” of the gold standard and the real bills doctrine “was expected to reduce the need for specific guidance by the government.”120
Neither the gold standard nor the real bills doctrine survives today. The gold standard has a more circuitous history, and survived in fits and starts until the U.S. formally withdrew its support for the international gold standard in 1971.121 The limitation of the gold standard on central banking practice is that the money supply must be managed with an eye toward long-term balance of international payments. When one country’s gold supply gets so low that market participants can doubt the convertibility of currency to gold, central-banking theory under the gold standard requires interest rate increases to attract more gold into the economy, even if that economy is in recession.
Debating the relative merits of the gold standard, real bills doctrine, or decentralized central banking are not the point here. The purpose is only that all three principles limited the ways in which the Federal Reserve Board could raise its revenue. The modern Board of Governors, on the other hand, does not face these limits. The consequence is that the Fed can create its own budget using a statutory authorization from a different era, subject to none of the restraints that existed at that time.122
12.Scholarly Engagement with Fed Budgetary Independence
Scholars have long noted that the Fed is not subject to the appropriations process, and that its non-appropriations status is a source of its independence. What is more interesting is that every legal scholar to have engaged this question has mischaracterized it. Some scholars mistakenly claim that the Board is funded by assessments on member banks.123 Others correctly note that the Fed is funded by assessments on the Federal Reserve Banks, rather than the member banks, but do not note the role played by proceeds from open market operations.124 Others correctly note that the Board uses the proceeds from open market operations, but then cite the provision that authorizes assessments on the Reserve Banks.125 One prominent legal scholar and historian has cryptically cited the Reserve Board assessments provision of the Federal Reserve Act for the conclusion that it creates a “straightforward accountability system,”126 although the author does not explain what that system is nor how it promotes accountability. A more recent article argues that “independent agencies such as the Federal Reserve . . . still ‘cannot afford to flout the views of the President,’ who continues to exercise substantial control as a consequence of his effective power of the purse,” without reference to the Fed’s unique budgetary independence.127In another article, the authors expressly mention the Fed as having a “significant interest in securing the goodwill of the President to enlist the chief executive’s aid in budget battles with Congress,” despite the Fed’s unique budgetary independence.128 The explanation is that “[e]ven agencies with an independent source of funding will have a recurring need for new authority and new sources of funding that outstrip existing demands.”129 And while the Fed may well find itself in a situation where its conventional means of securing funding will be inadequate, that eventuality, if it occurs, seems a flimsy basis for anticipatory reliance on the President for “aid in budget battles with Congress.” The reality is that the Fed’s budgetary independence is extraordinary, based only in part on statute, and illustrative of how the institutions of Fed independence—legal and non-legal—interact side by side to create the space within which the Fed operates. And that is a space that scholars have repeatedly mischaracterized.
13.Implications of the Fed’s Budgetary Autonomy
One point should be emphasized, as statements about how the Fed interacts with the money supply tend to provoke spirited arguments, to put it mildly: there is nothing secretive or nefarious about the Fed’s use of open market operations to fund itself. The Fed includes its accounting of its open market operations in its annual reports, and has done so—with varying degrees of transparency—for its entire one-hundred-year history. Moreover, the Fed has, in a century under intense scrutiny from market participants and existential critics alike, had no major financial scandal.130 This is an impressive feat for any agency, let alone one that generates as much controversy as the Fed. Indeed, Ben Bernanke, even when he flies to far off conferences in remote towns in South Korea or in the far-off Arctic, still flies commercial.131
This is not to say that the Fed’s funding decisions shouldn’t be scrutinized. There are important empirical questions about whether any other agency has matched the Fed’s budget growth, for example. A proper empirical inquiry would assess whether budget growth of the entire System matches or deviates from the growth of other agencies. Attention to the variance would also be useful. To take an example topical in 2013, the “sequester” that required mostly indiscriminate reductions in agency budgets did not apply to the Federal Reserve.132 And unlike non-appropriated agencies funded through market assessments, the Fed is not subject even to the ebbs and flows of their own assessments. How these realities affect the Fed’s budgetary decisions—from salaries to perquisites to hiring decisions—are important topics of scholarly inquiry.133
Rather than an exposé, the point of this analysis is to explain the way that the Federal Reserve’s funding structure has moved beyond its statutory mooring. The legal mechanism provided by statute in 1913 removed the Fed from the annual legislative appropriations process. But the legislative change away from autonomy for the Reserve Banks, the non-statutory rejection of the real bills doctrine, and the executive decision to abandon the gold standard have moved away from that originally limited funding apparatus. Whereas the statutory mechanism anticipates checks on the Fed’s ability to create the money with which it funds itself, the current practice has no such limitation. Scholars have all but ignored this statutory quirk, and even those who make passing reference do not engage in legal or historical analysis of its features.134
There are more than academic reasons for focusing on the Fed’s budgetary independence. Between 2007 and 2013, the Fed’s balance sheet increased from $870 billion to over $3.3 trillion.135 Several studies, including one from the Board of Governors, have highlighted the coming difficulties that the Fed is likely to face when the era of accommodative monetary policy subsides and the Fed must unwind the massive balance sheets it has accumulated during the last five years.136 There is a not implausible risk that as the Fed does so, it will face net losses such that it cannot cover its own expenses.
If that occurs, the Fed will face four options: (1) slash its operating budgets such that its expenses match the more modest assessments already available from member banks through the services rendered by Reserve Banks; (2) seek to expand the base of those assessments, consistent with the original policy intention behind the System’s status as a non-appropriated entity; (3) seek capital infusions from Congress; or (4) manipulate its open-market operations sufficient to cover its operating expenses. The first option is the most politically palatable and financially plausible, but could come at real cost to both the ability of the Fed to pursue monetary policy and bank supervision, and the ability of the Consumer Financial Protection Bureau to execute its mission (since the latter institution’s budget is linked, by law, to the Fed’s budget). The second option is the probably even more politically palatable—it doesn’t include cuts to the institutions—but is more financially risky: the banks may not want the services offered in sufficient quantity to support the increased demands that such a budget would place on them. The third option is politically uncertain—the prospect of the Fed Chair going to the Congress, hat in hand, to seek additional funding could result in extraordinary legislative measures that could redo the Fed according to the prevailing monetary zeitgeist. And the fourth, of course, is scandalous. How the Fed would proceed in this scenario is open to speculation.
The Fed has dismissed the idea that balance sheet considerations will influence its monetary policy decisions.137 And at least one of the papers’ models depends on a scenario where Congressional gridlock and public debt will push interest rates higher, earlier than the Fed anticipates.138 In May 2013, however, the Congressional Budget Office released a report139 that drastically reduced deficit projections such that at least some commentators view the fiscal outlook as significantly less dire than was previously projected.140 The focus of these debates has been, as with most aspects of academic and popular analysis of the Fed, on monetary policy and the ways that conduct of that policy will be affected by the change in the balance sheet.
The point is simply that the Fed’s actions, extraordinary and controversial though they have been since the beginning of the 2007 crisis, will continue to push the limits of our understanding of what central banks are designed to do. As the Fed continues to innovate, appropriate questions of Fed independence, democratic accountability, and institutional design will be part of the conversation. A more sophisticated understanding of how Fed independence operates, in theory and practice, will help guide those conversations.
14.The President and the Fed
Legal and non-legal institutions also play important roles in understanding the relationship between the Fed and the President. That relationship is multi-tiered, depending on what is meant by “President” and what is meant by “the Fed.” First, Part III looks at the main public interface between the Administration and the Federal Reserve System, the President and the Fed Chair respectively. Part III.B. then addresses the role of the non-Chair Governors of the Board, and again flags a misunderstanding about the relationship between legal and non-legal institutions: while the Governors have extraordinary fourteen-year terms meant to protect independence and stagger appointments across a Presidential Administration, the practice of Board service is different. Governors almost never serve their full terms, and because the President fills appointments when they become vacant, Presidents have routinely had the ability to fill a majority of the Board and often the Board in its entirety.
Part III.C. then briefly turns to the curious history between the Fed—meaning, most generally, the FOMC—and the Treasury, including with a preliminary reinterpretation of one of the most important episodes in the Fed’s history, the Fed-Treasury Accord of 1951 that liberated the Fed from its role in managing the process and interest rates for the public debt.
The Law: Removability of the Chair
The Chair of the Board of Governors serves two statutorily-defined roles: she is the Chair of the Board, nominated by the President and confirmed by the Senate to a four-year term. She is also one of seven members of the Board, nominated by the President and confirmed by the Senate to a fourteen-year term. The Federal Reserve Act is clear that, in her capacity as Governor, she is only removable for cause. The statute is silent, however, with respect to her removability as Chair.
Given the hallmark of independence that removability has become in administrative law, and given the prominence that the Fed Chair receives within the System and indeed within government and the public imagination generally, it is perhaps remarkable that the Federal Reserve Act is silent on the question of Chair removability. What, then, does the law do to provide independence for the Chair?
This is an open question, and one unlikely ever to find judicial resolution. One scholar has argued that, given this silence regarding Fed Chair removability, a “convention” of for-cause removability must be inferred in light of the presumptions in favor of Fed independence generally.141 Such an inference is not unprecedented: the entire architecture of the removability opinion in Free Enterprise Fund v. PCAOB, rests on just such an inference with respect to the Commissioners of the SEC.
But I am skeptical that a court would, or even must, reach this conclusion, and would argue instead in favor of the more literal lack of removability for the Governor qua Chair. The literal reading of the statute is warranted for two reasons. First, even on its own terms, it is not entirely accurate to claim the statute puts no condition on the Chair’s removal: again, the Chair, in her role as a member of the Board of Governors, is removable only “for cause.”142 The Chair is, of course, far more important than the other Governors in the public estimation, and there is evidence from the legislative history of the 1935 Act bearing on the argument that the Chair should be removable only for cause. But the nature of the Chair’s participation on the Board is more complicated than, say, an SEC Commissioner without the statutory removability protection. The Fed Chair’s joint service as Governor and Chair is different from other agencies, such as the FCC, FEC, and SEC, and it matters for understanding how the removability restriction for the Governor affects the Chair’s independence. A removed Chair remains a Governor; whether the President likes it or not, that very existence can cause problems for the management of the System. The statutory restriction on the Governor’s removability is thus a partial protection against arbitrary removability of the Chair.
The history of the Fed Chairmanship bears out this theory of Chair protection, even without the presumption of non-removability. While it’s accurate to say that “no President has ever formally discharged the Fed Chair,”143 the history of Chair removal is more complicated than that. That is, the President has never written a letter to a Fed Chair terminating his employment of the kind that prompted litigation in Humphrey’s Executor v. United States, where President Roosevelt made that demand on a Commissioner of the Federal Trade Commission.144 But three Chairs in the modern era have functionally been removed from their positions as Chairs before the end of their fourteen-year terms as Governors. Two of those were removed before the conclusion of their four-year terms as Chairs.
The first is Marriner Eccles, the Father of the modern Fed, appointed to the Chair by Roosevelt in 1934 immediately prior to the Fed’s 1935 reorganization. But when he was up for reappointment as Chair under President Truman, the President refused, contrary to Eccles’s wishes.145 Eccles chalked up the denial of reappointment to his taking too hard a stand on a banking enforcement in California,146 but Truman’s Secretary of the Treasury thought Eccles too independent of the President on monetary policy.147 But unlike subsequent Chairs Burns148 and Volcker—neither of whom was reappointed as Chair when the terms of Governor would have permitted reappointment—Eccles refused to leave the Board of Governors, and continued to make his influence felt on some of the highest-profile decisions in the Fed’s history.149 Thus, removal from the Chair—formally or informally—couldn’t eliminate the voice of the Governor who once occupied it.
Thomas McCabe, Eccles’s successor, also provides an important counter example. Because of his role in the Fed-Treasury Accord that Truman had opposed, he felt pushed out—before his four-year term had ended—and resigned, paving the way for William McChesney Martin, then an official in Truman’s Treasury Department.150
The third Chair to be “removed” was William Miller, President Carter’s first Fed Chair. He was removed after one year for what was widely viewed as incompetence at managing the Great Inflation of the 1970s.151 In what may well be unique in the annals of executive appointment, Miller’s removal was not to the ignominy of the non-government sector, but to his place as Secretary of the Treasury. To be sure, it’s difficult to call the appointment as the President’s spokesman for the Administration’s economic policies a “removal,” but the episode has led several to reach this very conclusion.152
These examples bear on the question of legal independence of the Chair by demonstrating that, even without that legal independence, the President has, in these limited cases, exercised some control over the person occupying the central chair.
But the argument that the Chair is protected from removability by convention fails for a more basic reason: the lack of evidence that there is a widespread presumption in favor of the existence of a bar on removability. One can argue that the statutory silence on Chair removability is “contrary to widespread belief,”153 but there is no direct evidence for this conclusion. The point may be instead that there is a widespread belief that the Board is an independent agency, or even the most independent of agencies. This conclusion certainly appears among academic economists. To cite a trivial example, economists do not acronymize SEC or FCC independence in the way they do CBI. But, as this article has argued at length, the independence of the Federal Reserve is about much more than the removability or not of the Fed Chair, and while economists certainly focus on the legal mechanisms of that independence, no model of CBI has ever focused exclusively on removability. Thus, the characterization of the non-removability of the Fed Chair as implied indulges in the same indefensible shortcut—that non-removability is the sine qua non of an independent agency—that this article and many others have challenged.154
The open question of the Chair’s removability is an interesting, but relatively minor question in assessing the Chair’s independence from the President. The political costs associated with such a challenge will only arise at a time when a President deems the Chair’s actions sufficiently noxious to warrant removal. If that situation arises, the Chair may recognize it and step aside, as did McCabe. He may acquiesce to the removal as Chair but stay as an Administration antagonist on the Board, as did Eccles.155 The President may also provide the cover of appointment to another office, as may have been the case with Miller. But as a matter of structure, history, and logic, the presumption that there necessarily exists a convention of non-removability is inaccurate.
The Chair’s vulnerability to at-will removal, then, only exposes the debility of Chair removability as a metonym for agency independence. The perhaps startling reality is that, as a formal matter, there is no legal separation between the President and the Chair: the Chair is, or at least should be, fireable at the President’s will.
15.Informal institutions of Chair-President Independence
The legal conclusion that the Fed Chair is removable at the President’s will is emphatically distinct from the informal institutional independence that a Fed Chair can wield. That informal independence may render that legal technicality irrelevant. Even assuming the President has the authority to remove a Governor from the Chair, a President may be politically unable to exercise that authority midstream, or, more often, even at the end of the Chair’s four-year term. In other words, a President tired of the policies of a Fed Chair who wishes her removal has two options. First, the President may, as did President Roosevelt to Federal Trade Commissioner Humphrey, fire her. Or, given the unique relationship between the Governor’s fourteen-year term as Governor and four-year term as Chair, the President may also refuse renomination even when the Chair is eligible for another term.
But then again, even if the President has the legal right in the first case, he may not have that political luxury in the second. The general expectation is that a sitting Chair is a candidate, if not the leading candidate, for reappointment to the Chairmanship, even if her initial appointment was by the sitting President’s political opponent. Because of the nature of past resignations and the interaction between the Chair’s four-year term and the fourteen-year term of the Governor who occupies the Chair—and, indeed, the statute’s permission to serve the balance of a previous Governor’s unserved term—the current situation is that the President will nominate a Chair roughly half-way through the President’s term, usually when the Chair is eligible for another four-year term as Chair.
The consequence of these staggered terms of a Presidential Administration and the Fed Chairmanship would render the Chair more independent of the President, since the Chair’s renomination is in the hands of a potential successor. There is much to be said for that view of the independent Chair. But there are several informal institutions—aside from the statutory silence on the Chair’s removability—that make the Chair’s renewable four-year term less independent of the President. First, a Chair interested in retaining the position may either seek to curry favor with the sitting President, or establish a relationship sufficient to render the non-renewal of that Chair politically costly.156 The most obvious mechanism to accomplish the first is to pursue an accommodative monetary policy to eliminate recessionary concerns from the election. There is indeed evidence of this phenomenon.157 And in other cases, a Chair post-election has changed his views throughout his tenure to accommodate those of the Administration. William McChesney Martin, Fed Chair from 1951 through 1970 during the administrations of five presidents is the best example here.158
But there are also examples of the Chair with an independent power base. President Reagan was disinclined to renominate Paul Volcker as Chair in 1983, but felt that the inflationary signal of non-appointment made reappointment a political necessity.159 President Clinton’s reappointment of Alan Greenspan—despite the latter’s credentials as a leading Ayn Randian libertarian160—was also influenced by Greenspan’s then-extraordinary reputation that might have made his non-renewal politically costly to Clinton.161 And sometimes the cost of failing to renominate a predecessor’s Fed Chair is financial, not political: President Obama reportedly renominated Chair Bernanke at some political cost out of fear that the financial markets would respond adversely at the suggestion of monetary and regulatory policies other than those pursued during Bernanke’s management of the financial crisis during his four-year term as Chair.162
It is impossible, then, to evaluate in the abstract whether the staggered terms of the President and Chair, and the Chair’s usual ability to be renominated, means that the Chair is more independent of the President, or less. This inability is what makes the historical analysis of Fed independence so important: the legal provisions, even read technically and carefully, are indeterminate. The practices of specific Presidents and specific Chairs is more telling. And it is telling indeed that of the eight Chairs of the Board of Governors since the position was created in 1935, five were appointed by a successive Administration. And four of the five were reappointed by a successor President of a different party—Martin appointed by President Truman, reappointed by Presidents Eisenhower (twice), Kennedy, and Johnson; Volcker, appointed by President Carter and reappointed by President Reagan; Greenspan, appointed by President Reagan, reappointed by Presidents George H.W. Bush, Clinton (twice), and George W. Bush; and Bernanke, appointed by President George W. Bush, reappointed by President Obama.
The point here is that the Fed Chair comes with it political capital that makes continued service—given the possibility of such guaranteed by the lengthy tenure possible by the combination of the Governorship and Chairmanship terms—a more plausible, sometimes compelled, political reality than other Presidential appointments. In the informal institution of Chair independence, the President is not simply free to appoint or fail to appoint whom he would.
16.The Chair-centric Federal Reserve
Part of this informal restriction on the President’s appointment power comes from the equivalence, in the popular mind, of the Fed Chair with the entire System generally.163 The foregoing discussion of the Chair’s relationship to the President tells us about the Chair’s independence, but it shouldn’t say much about Fed independence. After all, “[b]y statute . . . the chair decides almost nothing herself: The Federal Reserve System is supervised by a Board of seven presidentially appointed, Senate-confirmed governors, of whom the chair is but one.”164
Even so, and despite strong policy reasons for abandoning the Chair-centric Fed,165 in the public imagination, the Fed Chair is the entire ballgame. Consider a test, without looking at the footnotes: any reader who has made it this far almost certainly knows that the current Fed Chair is Ben Bernanke, soon to be succeeded by Vice Chair Janet Yellen. Most readers can also probably name at least Bernanke’s two immediate predecessors.166 But who are Yellen’s two predecessors as Vice Chairs?167 Who is the Vice Chair for Bank Supervision?168 The Vice Chair of the FOMC?169 And who, after Chair Bernanke and Vice Chair Yellen, are the remaining five governors?170 These individuals are virtually unknown among the general public. To put a finer point on it, in the words of one prominent journalist: The Chair of the Fed is an A-list political celebrity; at best, the Vice Chair is “B-list”; the other Governors are C-List, “anonymous politically and socially.”171
This equation of the Chair with the Fed has led one scholar to conclude that “the Fed’s history—and the growth of its power—is largely the product of the leadership of its Chairmen.”172 But the public face of the Federal Reserve is only one small part of that dynamic. Unlike, the Supreme Court for example, there is not a tradition of the Chair losing a majority of the vote on the Board of Governors or the FOMC. While some members of the latter Committee have issued dissenting statements from FOMC policy decisions,173 there are very few instances where the Chair is in dissent.174 Not only, then, is the Chair the popular equivalent of the System, he is also the policy equivalent.
Exactly how that unanimity or near unanimity arose in practice is unclear.175 But the informal institutions of Chair dominance—despite no basis in statute for that conclusion—renders the Board susceptible to the pressures of the Chair; in many senses, Board independence of the President and Chair independence of the President may be one in the same.
17.Appearances of Independence
Finally, the informal appearances of independence from the President are an important part of the role, and depend entirely on the personalities of the President, Chair, and—to a lesser extent—the Secretary of the Treasury. For example, keeping up the appearance of Fed independence, whatever the legal mechanisms, were obsessions of Chairs William McChesney Martin,176 Paul Volcker,177 and Alan Greenspan178 who seemed constantly preoccupied by the maintenance of this informal independence. And the tenure of Nixon/Ford Era Chair Arthur Burns is widely regarded as a failure in large part because of his proximity to the President. First reported in 1974,179 the recently published Burns diaries are filled with references to a close personal and emotional proximity between Burns and Nixon that raise modern eyebrows about that policy proprieties of that relationship. A few examples illustrate the point. Nixon told Burns about his appointment of prominent Democrat John Connolly as Secretary of Treasury before announcing it publicly, and then told Burns that Connolly—a politician, not an economist or businessman—would learn the ropes of his new position from Burns.180 Burns attended cabinet meetings;181 had his speeches vetted by Nixon’s staff;182 cleared his talking points with the President ahead of a meeting with other central bankers in Basel, Switzerland;183 advised Nixon on tax, wage, and other fiscal policy;184 made pledges to remain the President’s “true friend” on economic policies before the public;185 and more.186
Perhaps in part following the anti-example of Burns, Chairs have appropriately sought to maintain their distance from the informal pull of the office of the Presidency. But it remains an active dynamic, and no assessment of Fed-Chair independence is complete without analysis of the specific relationship and the specific personalities that inhere in each.
What, then, is the net effect for Fed independence based on the legal and non-legal institutions of Fed independence between the Fed Chair and the President? As the foregoing illustrates, the answer to that question is impossible to predict in the abstract. The nature of that relationship is entirely dependent on the personalities of the individuals who occupy the offices. While the law of Chair reappointment, with its interaction with appointments to the Governorship, favors across-Administrations appointments when the economic climate favors the policies of the incumbent Chair, it also suggests the incentive to cater to an incoming President’s wishes on economic policy. Thus, how that dynamic will play out in practice will depend on those individuals and their individual and historical contexts. The open legal question—and this article argues that the question should be resolved in favor of pure removability—is whether the President can legally remove at will the Chair qua Chair. But that question is, absent unforeseeable circumstances, essentially irrelevant: because of the political and financial (market) costs with such a disruption, not to mention the legal risk, a President is very unlikely to attempt a mid-term removal. However interesting as a matter of administrative law, the impact of that exact institution is far less important than the other legal and non-legal institutions that regulate the Chair’s relationship with the President.
While the Chair is perceived in substance and form as the power behind the System generally, the presence of the other Governors, and the formal and informal institutions that support their independence from the President, are worth highlighting. Here, the legal protection of a non-renewable term and the practice—by no means compelled, but widely followed—of significantly shorter tenure are at cross purposes. The result of that combination is also unexpected, given the emphasis in the Federal Reserve Act and in the scholarship interpreting the Fed’s independence: instead of limiting the President’s ability to choose his Board, the practice of frequent resignations has enhanced it. Since the second founding of the Federal Reserve in 1935, the President has effectively chosen his Board.
The Myth of the Fourteen-Year Term
As mentioned, by statute, each member of the Board of Governors is appointed for one non-renewable term of fourteen years.187 This is one of the longest terms of service in the federal government. Scholars have long discussed the Fed Governors’ lengthy tenure, usually uncritically for the propositions that, first, the fourteen-year term is “staggered”188 such that the President cannot immediately stack the Board in his favor; or, second, that the term represents a “term of office for each member . . . made long enough . . . to prevent day-to-day political pressures from influencing the formulation of monetary policy.”189 But a tradition of early resignation—a non-legal institution—makes this legal guarantee less important than it seems. Excluding the Chairs, the average term of the governors since the Board was constituted in 1935190 is just over six years, well within the mainstream of independent agencies.191 Including the Chairs, the figure is just under seven years. Indeed, it appears that only one non-Chair governor in the history of the Federal Reserve served a full 14 year term,192 although two others served portions of two terms totaling fourteen years or more.193
The non-renewable fourteen-year term is meant not only to insulate the Governors from the need to curry favor with the President—a principle undermined by the ability to serve unexpired terms—it is also meant to limit the President’s ability to overrun the board. The fourteen-year term was not arbitrarily decided: it corresponds to the seven members of the Board of Governors, just as the ten-year term corresponded to the five-member Federal Reserve Board prior to the 1935 reorganization. The idea is that each President should get but two appointments to the Board during a four-year administration. So much is suggested by the statute itself. Section 10 of the Federal Reserve Act instructed the President, in 1935, to “fix the term of the successor to such member [as served as a member of the Federal Reserve Board, rendered defunct by the 1935 Act] at not to exceed fourteen years, as designated by the President at the time of nomination, but in such manner as to provide for the expiration of the term of not more than one member in any two-year period.”194
Table 1 shows how, in practice, a convention of frequent resignations has made this legal mechanism of independence effectively inert.
Table 1: Presidential Appointments to the Board of Governors, 1935-2013195