Chairman Barr, Ranking Member Moore, and distinguished members of the Committee on Financial Services Monetary Policy and Trade Subcommittee, my name is George Selgin, and I am the Director of the Cato Institute's Center for Monetary and Financial Alternatives. I am also an adjunct professor of economics at George Mason University, and Professor Emeritus of Economics at the University of Georgia.
I'm grateful to you for allowing me to take part in your consideration of various proposals for reforming the Federal Reserve System. With your permission, I wish to limit my testimony to one only of several proposals being discussed at today's hearing, namely, the proposal to make the FOMC, rather than the Federal Reserve Board of Governors, officially (and not just informally) responsible for setting the interest rate paid on banks' excess reserve balances.
Monetary Policy Authority before Interest on Reserves
Between the middle of the Great Depression and the recent financial crisis, responsibility for determining the stance of monetary policy has rested mainly, if not exclusively, with the Federal Open Market Committee, a twelve-member committee consisting of the seven-member Federal Reserve Board of Governors plus five of the twelve Federal Reserve Bank presidents, always including the president of the New York Fed, with the remaining four bank presidents serving on a rotating basis.
Title II of the Banking Act of 1935 (U.S. Code § 263) amended the Federal Reserve Act by creating the FOMC and vesting it, rather than either individual Federal Reserve regional banks or the Federal Reserve Board of Governors, with the authority to "consider, adopt, and transmit to the several Federal Reserve banks, regulations relating to the open-market transactions of such banks." The amendment also stipulated that the Fed's open market operations "shall be governed with a view to accommodating commerce and business and with regard to their bearing upon the general credit situation of the country."
The new arrangement essentially ended individual Federal Reserve banks' power to independently influence the stance of monetary policy. Whereas until 1935 each Fed bank was in charge of open-market operations within in its own district, in the new set-up, instead of pursuing their own, independent policies, "the district banks participated in the creation of a coordinated, national monetary policy."1 This outcome reflected a compromise between those who would have preferred, and those who feared, the complete centralization of control over monetary policy in Washington.
Strictly speaking, the 1935 legislation did not give the FOMC exclusive control over monetary policy. While it gave that committee complete authority over open market operations, it placed control over two other instruments of monetary policy—changes in banks' minimum reserve requirements and Fed banks' discount rates—with the newly established Board of Governors of the Federal Reserve System.2 However, these other monetary policy instruments have since fallen into desuetude. Regarding reserve requirement changes, by 1954 the Board had concluded that
Frequent changes in requirements even by very small percentage amounts would be disturbing to member banks and to the credit market. For these reasons this method of influencing bank reserve positions and the flow of credit and money is usually employed only when large-scale changes in the country's available bank reserves are desired. For day-to-day operations in influencing the flow of credit and money, the Federal Reserve depends principally on the more flexible instruments of discount and open market operations.3
Discount rate adjustments, in turn, became unimportant in influencing the stance of monetary policy when the Fed switched from reserve targeting to targeting the federal funds rate during the 1980s. Since 2003, moreover, the discount rate has been set above the fed funds target (or, since November 2008, above the upper bound of the fed funds target range). When the discount rate is above the effective fed funds rate, banks ordinarily have no reason to borrow at the discount window.
Consequently, for all intents and purposes, for several decades prior to October 2008, when the Fed began paying interest on banks' reserve balances, the FOMC – and the regional bank presidents taking part in it – exercised exclusive control over monetary policy. Moreover, as we shall see, Fed officials themselves now take for granted the FOMC's ultimate responsibility for the conduct of monetary policy.
Interest on Reserves transfers Formal Control over Monetary Policy to the Board of Governors
The 1935 compromise by which regional Fed bank presidents, through their participation in the FOMC, shared the legal authority to determine the stance of monetary policy with the Board of Governors, came to an abrupt—if generally unnoticed—end in 2008 as a result of the passage of the Financial Services Regulatory Relief Act of 2006 and the Emergency Economic Stabilization Act of 2008.
Section 203 of the 2006 Act allowed the Fed to begin paying interest on banks' reserve balances beginning on October 1, 2011. The 2008 Act advanced that date by three years, allowing the Fed to begin making interest payments as early as October 1, 2008. The Fed was authorized by these Acts to pay interest on both banks' required and their excess reserve balances. Importantly, the 2006 law assigned responsibility for setting both rates, not to the FOMC, but to the Board of Governors, and this provision remained unaltered by the 2008 law.
The Fed's immediate goal in securing the authority to start paying interest on banks' Fed balances in October 2008 was to prevent the crisis-related emergency lending it was engaging in at that time from driving the fed funds rate below the FOMC's then-chosen target of 2 percent. By paying interest not just on required but on excess reserves, the Fed could encourage banks to retain newly-created reserves that came their way, instead of lending them. Interest on excess reserves (henceforth IOER) was thus deployed early so that it might bolster the Fed's ordinary means of monetary control.
As the crisis continued, however, the IOER rate came to perform, not merely a supplementary role, but the lead role in the Fed's setting of monetary policy. Instead of relying on open-market operations to achieve a target federal funds rate, the Fed switched to a new "floor" operating system in which the IOER rate itself took the place of open-market operations as its chief instrument of monetary control. The basic idea was that, instead of loosening or tightening its policy stance by buying or selling securities in the open market (and thereby adding to or subtracting from the total supply of bank reserves) the Fed could loosen or tighten by influencing banks' demand for reserves. A higher IOER rate would, other things equal, increase banks' demand for reserves, tightening credit by discouraging bank lending, while a lower one, by reducing banks' appetite for reserves, would loosen credit, encouraging them to lend more.
By keeping its IOER rate above corresponding market interest rates, as it has done since November 2008, the Fed has prevented additions to the supply of bank reserves from resulting in any general increases in the supply of credit. Instead, increases in total bank reserves were matched by roughly equal changes in banks' excess reserve holdings. Although the Fed could still purchase or sell assets on the open market, and although it did, in fact, ultimately undertake three rounds of Large Scale Asset Purchases, its open-market operations ceased to play their traditional role as the Fed's main instrument of monetary policy.
Thus the Fed's switch to an IOER-based operating system had the effect of transferring control over the Fed's monetary policy stance from the FOMC, where it had resided for decades, to the Board of Governors, which had previously exercised that control solely through its participation, together with several Fed bank presidents, in the FOMC.
A Change Not Anticipated by Congress
The just-described transfer of authority for conducting monetary policy, from the FOMC to the Board of Governors, had not been anticipated, and was certainly not intended, by Congress when it passed the 2006 Financial Services Regulatory Relief Act.
Instead, when Congress originally granted the Fed authority to pay interest on banks' Fed balances, it did so in order, as the Federal Reserve Board itself stated in its 2006 Annual Report, to "reduce unnecessary burden [sic] on banking organizations and improve operation of the financial system."4 Interest payments on required reserves, the report said, would "remove a substantial portion of the incentive for depositories to engage in reserve-avoidance measures," allowing "the resulting improvements in efficiency [to] eventually be passed through to bank borrowers and depositors."
As for interest on banks' excess reserves, although the 2006 Act also granted the Fed the authority to pay such interest, the Fed at that time anticipated employing the IOER rate, not as its chief device for regulating the federal funds rate, and for thereby adjusting the Fed's monetary policy stance, but merely to serve as an above-zero minimum possible value for the effective fed funds rate, so as to limit that rate's potential volatility.Because the Fed's target fed funds rate would generally fall between that minimum value and the Fed's discount rate, open-market operations were to continue to serve as the Fed's primary monetary control instrument.
These originally-intended functions of interest payments on banks' reserve balances were reflected in the 2006 law's stipulation that interest on Fed balances be paid "at a rate or rates not to exceed the general level of short-term interest rates" – which stipulation was not altered by the Emergency Economic Stabilization Act. Had Congress intended to have the Fed employ the interest rate on banks' reserve balances as an instrument of monetary control, and certainly had it intended to have that rate serve as the Fed's chief instrument of control, rather than as a mere means for offsetting the reserve requirement tax, it would certainly not have placed such a limit on the rates the Fed was authorized to pay.5
The decision to make the Board of Governors, rather than the FOMC, responsible for setting interest rates on banks' Fed balances, which was also carried over from the 2006 to the 2008 Act, likewise reflected the originally-intended purpose of interest on reserves. Because such interest payments weren't intended to serve as a primary means of monetary control, vesting control over them with the Board rather than the FOMC was not seen as contradicting the spirit of either the 1935 Banking Act or subsequent developments that had left the FOMC exclusively in charge of determining the stance of monetary policy.
An Untenable Situation
When, at the Fed's urging, Congress passed the 2008 Emergency Economic Stabilization Act, allowing the Fed to immediately begin making interest payments on banks' reserve balances, it cannot possibly have anticipated that the Fed would end up treating those interest payments, not only as an additional instrument of monetary control, but as its chief instrument of monetary control.6 Consequently, it was only inadvertently that Congress ended up transferring responsibility for monetary policy from the FOMC to the Federal Reserve Board, thereby denying to the regional Fed banks the influence they had long exercised, at least to some extent, in shaping the course of monetary policy.
It's true that the difference between control of monetary policy by the FOMC and control of that policy by the Board of Governors is not as great as it may seem. As Ben Bernanke has pointed out,
The seven members of the Board (when all seats are filled) each have a permanent vote on the FOMC, whereas only five Reserve Bank presidents vote at each meeting. Moreover, in practice, greater sway over policy is held by the chair and those close to her, as well as by those Committee participants (even those without a vote at a particular meeting) who are most persuasive in the internal debates.7
Furthermore, Bernanke writes, although authority to set the IOER rate formally rests with the Board, "Fed policymakers know that the expectation of the Congress and the public is that monetary policy will be made by the FOMC, not the Board—an expectation reinforced by decades of Fed practice." Consequently there is no "risk that the Board will try to block implementation of an FOMC decision" (ibid.).
However, with all due respect to Mr. Bernanke, if both the public and Congress expect monetary policy to be made by the FOMC rather than by the Federal Reserve Board, then Congress has a duty to see to it that that expectation is fulfilled, not simply by counting on the Federal Reserve Board to fulfill it, out of a supposed deference to past experience, but by legally placing the power to make monetary policy where everyone agrees that it belongs. What's more, anyone familiar with the Fed's history during the last decade will have reason to question the assumption that "decades of Fed practice" supply a reliable guarantee of what the Fed may or may not do in the future.
The proposal now before your committee, to amend the Federal Reserve Act so as to make the Federal Open Market Committee rather than the federal Reserve Board officially responsible for regulating the interest rate paid on banks' excess reserve balances, would correct the present, anomalous state of affairs, legally ensuring that monetary policy decisions rest with the FOMC, and not the Board.
Better Still, Restore Interest on Reserves to its Originally Intended Purpose
I have argued so far that, if changes in the IOER rate are to continue serving primarily as a means for monetary control, then responsibility for setting the IOER rate should rest with the FOMC, which has traditionally been responsible for monetary policy, rather than with the Federal Reserve Board of Governors, the members of which have, traditionally, taken part in determining the Fed's monetary policy stance only by virtue of being permanent members of the FOMC, on which five regional Federal Reserve bank presidents also serve.
There are, however, grave problems with the Fed's new, IOER-based operating framework. As I have previously testified at length before this same subcommittee concerning those problems, I will not repeat that testimony or any part of it here.8 I will only observe that those grave problems supply a powerful argument for compelling the Fed to return to relying on open-market operations, rather than changes in the interest rate paid on banks' excess reserve balances, as its preferred instrument of monetary control. That could either mean having the Fed revert to its pre-2008 operating system, or allowing it to implement the slightly modified version of that system that the 2006 Financial Services Regulatory Relief Act of 2006 was supposed to provide for, in which interest may be paid on banks' reserve balances, both required and excess, but only at rates low enough to discourage banks from amassing excess reserves. The same outcome might also be achieved by revising the current law to allow the Fed to pay interest on banks' required reserve balances only, but not on their excess reserves.
Should Congress choose to confine the Fed's interest payments as recommended here, those interest payments would no longer be capable of supplanting open-market operations as an instrument of monetary policy. Instead they would serve only to compensate banks for their reserve holdings, and perhaps to place an above-zero lower limit on the effective federal funds rate, without ordinarily encouraging banks to hold any excess reserves, and without becoming the chief means for regulating that rate—as was the original intent of the 2006 Act. In that case it would be perfectly appropriate for Congress to leave the Federal Reserve Board in charge of setting the rates paid on banks' reserve balances, though only assuming that the Board is no longer allowed to make a mockery of the stipulation that those rates not "exceed the general level of short term interest rates."
To rule out that possibility, Congress should consider amending the 2006 Act so as to give a precise meaning to the phrase "the general level of short-term interest rates." Given the statute's intent, the interest rates to which that phrase refers are presumably market-determined rates on instruments similar in duration and risk to the reserve balances on which the Fed is authorized to pay interest. Because reserve balances are themselves risk-free assets of zero maturity, private overnight repurchase agreements collateralized by Treasury securities are the closest private-market equivalents.
Although private overnight repo rates vary, the Federal Reserve Bank of New York has recently conducted extensive research aimed at establishing overnight repo benchmark rates, using transaction level data. Based on this research, the New York Fed has developed a "Broad Treasury financing rate" that is very well suited to serve as an IOER benchmark rate, that is, as a reference "general" rate for the purpose of implementing the statute. 9 The 2006 statute could therefore be amended by having it define the "general level of short-term interest rates" as the average of the "Broad Treasury financing rate" over the 6-week period preceding any FOMC rate-setting announcement.10>
However, until or unless the Fed's use of interest payments on banks' reserve balances can be confined as described—as long, in other words, as adjustments to those payments continue to serve as an important determinant of the Fed's monetary policy stance—the power to make those adjustments should rest solely with the FOMC, where it clearly belongs.
1 See Gary Richardson, Alejandro Komai, and Michael Gou, "Banking Act of 1935," Federal Reserve Bank of Richmond Federal Reserve History website, https://www.federalreservehistory.org/essays/banking_act_of_1935.
2 Among its other provisions the 1935 Banking Act replaced the previous "Federal Reserve Board" with the present "Board of Governors of the Federal Reserve System." Although this change was at first largely cosmetic, after 1936 the Secretary of Treasury and Comptroller of the Currency no longer sat on the Board, as they had done previously (the first of them as Governor or Chairman). Confusingly, the Board of Governors continues to be routinely referred to as the "Federal Reserve Board."
3The Federal Reserve System: Purposes and Functions (Washington, DC: Board of Governors of the Federal Reserve System, 1954 (https://fraser.stlouisfed.org/files/docs/historical/federal%20reserve%20history/bog_publications/bog_frs_purposes_1954.pdf).
4 93rd Annual Report: Board of Governors of the Federal Reserve System, 2006 (https://www.federalreserve.gov/boarddocs/rptcongress/annual06/sec2/c5.htm).
5 Since October 2008 the Fed has evaded the prescribed rate limit by allowing the Fed's own primary credit (discount) rate to represent "the general level of short-term interest rates" (see Regulation D (Reserve Requirements of Depository Institutions, 12 CFR Part 204). In fact, as a matter of policy, the primary credit rate is set well above market short term rates, so as to make it impossible for banks to borrow from the Fed's discount window for the sake of relending the borrowed funds at a profit.
6 Thus the Federal Reserve's strategy for "normalizing" its policy stance has it doing so primarily by gradually raising the IOER rate, which defines the upper limit of the Fed's federal funds rate "target range," to just under 3 percent (https://www.federalreserve.gov/monetarypolicy/reqresbalances.htm).
7 Ben Bernanke, "The FOMC, the Board of Governors, and Fed interest rate policy." The Brookings Institution, Ben Bernanke's blog, Tuesday, June 9, 2015 (https://www.brookings.edu/blog/ben-bernanke/2015/06/09/the-fomc-the-board-of-governors-and-fed-interest-rate-policy/).
8 I refer to my July 20, 2017 testimony at the Committee on Financial Services Monetary Policy and Trade Subcommittee Hearing on "Monetary Policy v. Fiscal Policy: Risks to Price Stability and the Economy" (https://financialservices.house.gov/uploadedfiles/hhrg-115-ba19-wstate-gselgin-20170720.pdf).
10 The proposed amendment might read as follows: "Section 19(b)(12) of the Federal Reserve Act (12 U.S.C. 461(b)(12)) is amended by inserting after Subparagraph (C): "(D) General level of short-term interest rates defined.—For purposes of this paragraph, the term ‘general level of short-term interest rates' shall be defined as the average value over the preceding six-week interval of the Federal Reserve Bank of New York's benchmark Broad Treasury financing rate on overnight repurchase agreements."