Overview

Recent policy discussions in Congress and the Administration have raised the possibility of eliminating the Federal Reserve’s authority to pay interest on reserves (IOR) as a means of bolstering the fiscal outlook.

Advocates for this proposal suggest that eliminating the Fed’s IOR authority could reduce interest expenses for the federal government by more than $1 trillion over a 10-year projection window. But without this tool, the Fed would need to transition to an entirely different means of implementing monetary policy—a so-called “scarce reserves” regime.  And in that regime, Fed net income and remittances to the U.S. Treasury would almost certainly be lower in the long run than under the existing ample reserves regime.

That conclusion follows from the fact that, in the long run, the average rate of return on the Fed’s assets exceeds the average interest expense on its interest-bearing liabilities.  In an ample reserves regime, the Fed will hold more assets on its balance sheet and thus generate more income from that interest rate spread.  Even over a shorter period such as the next ten years, transitioning promptly to a scarce reserves regime would likely lower the Fed’s income and remittances.

When the federal budget is on an unsustainable path, finding a seemingly easy way to reduce the Fed’s interest expenses by $1 trillion sounds appealing.  However, proposals to eliminate the Fed’s IOR authority don’t take into account changes on the asset side of the Fed’s balance sheet that would need to occur in order for the central bank to continue to keep short-term interest rates at levels consistent with achieving the FOMC’s maximum employment and price stability goals.^[2]

Below we provide some details on the longer-run fiscal consequences of eliminating the Fed’s IOR authority.  In addition, we present rough estimates of the shorter-run fiscal effects of removing IOR authority following a methodology similar to that used by the Congressional Budget Office (CBO).

These estimates suggest that eliminating IOR authority could reduce Federal Reserve cumulative remittances to the Treasury over the 2026-2035 window by about $150 billion relative to a baseline scenario.

While this note focuses narrowly on the direct effects of eliminating IOR authority on Fed remittances, it’s important to emphasize that the Congress has directed the Fed to conduct monetary policy to promote its statutory objectives of maximum employment and stable prices.  Eliminating the Fed’s IOR authority could severely hamper the Fed’s implementation of monetary policy and have very serious macroeconomic and financial consequences that would, in turn, have adverse effects on the fiscal position of the U.S. government.

Background

For most of its history, the Fed lacked statutory authority to pay interest on reserves.  The inability to pay interest on reserves combined with the imposition of reserve requirements presented significant challenges for the Fed in implementing monetary policy.

Prior to 1980, only banks that were members of the Federal Reserve System were subject to reserve requirements.  The requirement to hold non-interest bearing reserves to meet reserve requirements effectively acted as a tax on bank transaction deposits.  Moreover, the reserve requirement tax was higher during periods when interest rates were higher.  Banks naturally sought to minimize the effects of the reserve requirement tax.  Particularly during the high-interest-rate period from the late 1960s through the 1970s, many banks opted out of Federal Reserve membership.

In 1980, the Depository Institution Deregulation and Monetary Control Act (DIDMCA) extended reserve requirements to all depository institutions so this option for avoiding the reserve requirement tax was no longer feasible.^[3]  However, banks became increasingly aggressive and creative in restructuring their liabilities to reduce their reserve requirements.  As shown in Chart 1, for example, the average level of reserve balances trended lower throughout the 1980s and 1990s and had dropped to only about $10 billion in 2006.^[4]

For all these reasons, economists and the Fed had long sought the authority to pay interest on reserves.^[5]  In 2006, the Congress finally passed legislation authorizing the Fed to pay interest on reserves beginning in 2011.

With the onset of the Global Financial Crisis (GFC) in 2008 and the challenges the Fed faced in managing reserves in the banking system, the Congress accelerated the authority to pay interest on reserves to the fall of 2008.

Following the failure of Lehman in the fall of 2008, the Fed announced the first round of large-scale purchases of longer-term securities to put downward pressure on longer-term interest rates and ease financial conditions.

As shown by the blue line in Chart 2, the size of the Fed’s securities portfolio expanded greatly in 2009. And the counterpart of the expansion of the Fed’s securities portfolio was a comparable increase in reserves outstanding in the banking system (red line).  Over subsequent years, the quantity of reserves in the banking system increased even more, reflecting multiple rounds of large-scale asset purchases to address the economic effects of the Global Financial Crisis and the pandemic at a time when the policy rate had effectively reached zero.

The enormous expansion in the Fed’s balance sheet necessitated a new approach to monetary policy implementation.  Prior to the GFC, the Fed implemented monetary policy in a so-called “scarce reserves” regime.  In that regime, the Fed actively managed the supply of reserves to equal the estimated demand for reserves by banks at the intended level of the federal funds rate.  In the new approach, dubbed the ample reserves regime, the Fed controls the level of short-term market rates primarily by setting two administered rates—the interest rate on reserve balances and the offered rate on overnight reverse repurchase agreements.  The section below discusses this approach in more detail and explains how the new approach to policy implementation affects the Federal Reserve System’s income.

Ample Reserves, Policy Implementation and Federal Reserve Income

Since the Global Financial Crisis, and officially since 2019, the Fed has implemented monetary policy in an “ample reserves” regime.^[6]  In this approach, the Fed relies on administered interest rates—the interest on reserve balance (IORB) rate and the offered rate on overnight reverse repurchase agreements (ON RRP) rate—to keep the federal funds rate at levels consistent with the Federal Open Market Committee’s policy objectives.^[7]  (The FOMC is the primary Fed body responsible for conducting monetary policy; the FOMC meets to set short-term interest rates about every six weeks.)

Under the ample reserves regime, the Fed seeks to supply a level of reserves so that the system operates on the “flat portion” of the aggregate reserve demand curve.

Figure 1 shows one way of visualizing this key characteristic of an ample reserves regime.  The solid line depicts a hypothetical reserve demand curve that is steep at low levels of reserves and flattens out close to the level of IORB at high levels of reserves.  If the Fed supplies a sizable quantity of reserves, short-term interest rates move close to IORB.  Moreover, shocks to the supply of reserves (depicted by the shaded area) have little if any effect on the level of short-term interest rates.

The ample reserves regime has many attractive features.  First and foremost, it delivers excellent interest rate control for the Fed even in environments with very large quantities of reserves in the banking system.  In addition to its efficacy for interest rate control, the ample reserves regime is also a very simple system to operate for both banks and the Fed.  In contrast, the prior scarce reserves regime with reserve requirements was very complicated operationally for both banks and the Fed.  And, as noted above, the scarce reserve regime effectively imposed a tax on banks that, in turn, led banks to develop increasingly sophisticated ways of avoiding the tax.^[8]

An implication of the ample reserves regime is that the Fed’s balance sheet is considerably larger than under a scarce reserves regime.  As shown in Chart 2 above, the Fed has a much larger quantity of reserves outstanding at present than was the case under a scarce reserves regime.  The counterpart of the larger quantity of reserves is a larger portfolio of securities on the asset side of the balance sheet.  Of course, the changes in the size and composition of the Fed’s balance sheet have important implications for interest income and interest expense on the Fed’s income statement as well.

Under an ample reserves regime, the Fed’s interest expenses are higher than under a scarce reserves regime.  However, the Fed’s interest income is also higher stemming from its larger securities holdings.  Ordinarily, the return on the Fed’s assets is somewhat higher than the average rate on reserves and other interest-bearing liabilities.  As a result, the large size of the balance sheet under an ample reserves regime tends to generate higher net income for the Fed, on average over time, as well.  Of course, there may be times when the Fed’s interest expenses exceed its interest income.

The return on the Fed’s assets tends to adjust slowly to changes in the level of interest rates.  As a result, in periods when short-term interest rates have moved up quickly such as the policy tightening cycle beginning in March of 2022, the Fed’s interest expense can move above interest income for a time.  In those circumstances, the Fed records a “deferred asset” on its balance sheet representing the cumulative excess of expenses over income.  This deferred asset declines over time when Fed income turns positive.  Conversely, during periods when short-term interest rates move down quickly, the Fed’s expenses fall relative to its income and net income may surge.^[9]

The Implications of Loss of IOR Authority for Policy Implementation

The Fed’s IOR authority is the primary tool that allows for effective interest rate control in an ample reserves regime.  At present, the banking system holds about $3.4 trillion in reserve balances.^[10]  If the Fed could not pay interest on those reserves, the immediate impact would be that banks holding those reserves would seek to invest in alternative higher yielding assets.  Their efforts to do so would put downward pressure on short-term interest rates until short-term rates dropped to the zero rate of return on reserves.

Figure 2 illustrates this in terms of the reserve demand curve.  The elimination of IOR authority would mean that banks would no longer be able to earn interest on reserves.  The reserve demand curve from figure 1 then would shift down (dashed curve) so that it flattens out along the x-axis at a level of short-term rates close to zero.  In order to maintain short-term interest rates at the level prevailing prior to the elimination of IOR, the Fed would need to reduce the aggregate quantity of reserves and likely by very substantial amounts, denoted by the dashed vertical line.^[11]

As noted above, in 2006 prior to the GFC, banks held only about $10 billion in reserve balances.  Even accounting for the grossing up in the size of bank balance sheets since then, a roughly equivalent amount today would be on the order of only $30 billion—more than 100 times smaller than the current level of reserves.

Long-Run Fiscal Implications of a Return to Scarce Reserves

The fiscal implications of a change in the Fed’s approach to policy implementation can be viewed through the lens of the consolidated balance sheet of the Treasury and the Fed or, equivalently, in terms of the gross interest costs of federal debt less Fed remittances.  Figure 3 provides a simple way of illustrating the connection between Fed remittances and the government’s net fiscal costs.  In this diagram, the total gross Treasury debt outstanding is measured along the x-axis.  The level of interest rates is measured along the y-axis.  Assuming a fixed long-run rate on Treasury debt, the gross interest cost of federal debt is shown by the sum of the blue and green shaded areas.  The Fed holds a sizable quantity of Treasury debt that is effectively financed by currency and reserves.  With a zero rate of interest on currency and the rate on reserves set somewhat below the average rate on Treasury debt, the Fed’s net income is shown by the blue shaded regions.  This net income is returned to the Treasury.  As a result, the net interest cost of federal debt outstanding is shown by the green shaded region.

Figure 4 shows how this picture would change after a hypothetical transition to scarce reserves.  With a much smaller quantity of reserves outstanding, the Fed would hold a much smaller securities portfolio.  Total Fed remittances then would be shown by the blue shaded area.  Treasury securities held by the non-Fed public would increase and the net interest cost to the Treasury is shown by the green shaded area.  The net effect of the transition to scarce reserve depends on the magnitude of two opposing effects.  On the one hand, the increase in relatively high cost Treasury debt and corresponding reduction in relatively low cost reserves increases net interest costs for the Treasury.  This effect is shown by the darker green shaded region.  On the other hand, the reduction in the rate paid on reserves to zero reduces net interest costs for the Treasury to the extent that banks hold non-interest bearing reserves in the scarce reserves regime.  This effect is illustrated by the darker blue shaded region.  The upshot is that the transition from ample to scarce reserves increases the net cost of Treasury debt if the dark green area exceeds the dark blue area.

We can get a sense of these magnitudes using information the CBO publishes on its long-run forecasts for the average cost of Treasury debt and the federal funds rate.  In its most recent long-range budget forecasts, the long-run cost of Treasury debt and the long-run federal funds rate were projected at 3.6 percent and 3.2 percent respectively.  Using these estimates, the area of the dark green box in Figure 4 is given by  where  is the level of ample reserves and  is the level of scarce reserves.  Similarly, the dark blue area is given by   The net cost of Treasury debt will be the same in the ample and scarce reserves regime when  .  Based on these interest rate assumptions, when the ample level of reserves is more than nine times the level of scarce reserves, the net cost of Treasury debt will be lower than in the scarce reserves regime.  Conversely, when the scarce reserves level is more than one ninth of the ample reserves level, the net cost of Treasury debt will be lower than in the ample reserves regime.  Using a long-run level of 9% of reserves to nominal GDP as an approximation of an ample level of reserves, the scarce level of reserves would need to be above 1% of nominal GDP before there was any benefit to the federal government relative to the ample reserves regime.  For comparison, the average level of reserves as a share of nominal GDP in 2006 was about 0.07% —far below the level that would be necessary to generate a net benefit for the federal government relative to an ample reserves regime.

Of course, the structure of the financial system has changed since 2006 and it’s possible that new liquidity regulations and the general increase in demand for safety and liquidity would drive a higher demand for non-interest bearing reserves as a share of GDP now than in the past.  That said, the demand for non-interest bearing reserves in the longer-run would need to be 15 times larger as a share of nominal GDP than in 2006 just for scarce reserves to “break even” with an ample reserves regime in terms of the net interest cost of federal debt.

It’s worth noting that even in the case in which the level of scarce reserves is relatively large and exceeds the breakeven threshold above, it does not follow that moving from ample to scarce reserves is a sensible way to reduce the federal government’s interest expense.  Indeed, a policymaker interested solely in minimizing the net cost of federal debt might choose to greatly expand the Federal Reserve’s balance sheet, thereby reducing high-cost Treasury debt in the hands of the public and effectively financing a larger share of debt outstanding with relatively low-cost reserves.

Near-Term Fiscal Implications of a Return to Scarce Reserves

Based on the types of back-of-the-envelope calculations above, it seems likely that an ample reserves regime would result in a lower net cost of federal debt over the longer-run relative to a scarce reserves regime.  But what about over a shorter-run ten-year projection period of the type that often informs deliberations in Congress?

At this horizon, evaluating the effect of a transition from ample to scarce reserves is more complicated than for the very long-run horizon considered above.  In particular, the rate of return on the Fed’s assets at present is lower than the long-run average rate on Treasury debt and, in fact, lower than the level of interest on reserves.  That should provide some benefit in moving from ample to scarce reserves because relatively high-cost reserves would be reduced and “replaced” with an increase in relatively low-cost Treasury debt held by the non-Fed public.

However, a reduction in the Fed’s securities holdings would result in realized losses on sales of securities and corresponding reductions in Fed remittances.  The discussion below works through some of these details using projections from the Congressional Budget Office as a baseline.

The Congressional Budget Office regularly prepares projections of Fed remittances to the U.S. Treasury.  In their latest projections released in January, 2025, total Fed remittances over the 2026-2035 period were estimated at $775 billion.  Table 1 below reports rough projections of Fed remittances based on a methodology similar to that described by the CBO.^[12]  Based on these calculations, total Fed remittances over the next 10 years would amount to $740 billion—close to the CBO’s remittance projections.  In these projections, total Fed interest income over this period cumulates to $2.3 trillion while cumulative interest expenses total $1.3 trillion.  The difference between the total net income over this period of about $1 trillion and the roughly $740 billion in projected cumulative remittances arises because a portion of Fed net income over the period 2026-2035 is assumed to pay down the deferred asset accumulated by the end of this year.^[13]

Table 2 shows similar projections for the case in which Congress eliminates the Fed’s IOR authority and the Fed reverts to a monetary policy implementation framework based on scarce reserves.  We assume that banks would wish to hold only about $100 billion in non-interest bearing reserves—a level considerably higher than the comparable level of reserves based on the experience prior to the GFC.  In this case, the cumulative expenses of the Fed drops to $72 billion—more than $1 trillion lower than interest expense under the ample reserves regime.  However, interest income declines substantially as well to a level of $1.3 trillion.  Moreover, with assumed sales of Treasury securities of $2.5 trillion—a decline sufficient to reduce reserves to the assumed “scarce” level of reserves of $100 billion in 2026—the estimated realized loss on these sales would amount to about $350 billion.  This estimate for realized losses is based on the fair value of the Fed’s securities holdings reported in its quarterly financial statements.^[14]  Assuming the sale of securities in 2026, the realized loss would depress Fed net income in 2026 and substantially increase the level of the Fed’s deferred asset.  As shown in Table 2, taking into account the likely realized losses on sales of securities holdings, cumulative Fed remittances to the U.S. Treasury would be projected to decline to a little more than $590 billion, or about $150 billion lower than projected under the ample reserves regime.

Of course, the sale of $2.5 trillion of securities over a relatively short period of time would almost certainly put upward pressure on longer-term interest rates, further depressing the value of the Fed’s securities holdings and amplifying the realized loss on sales of securities relative to the figure presented here. ^[15]

In short, the analysis here indicates that a rapid transition from ample to scarce reserves with the elimination of IOR authority would result in a significant deterioration in the U.S. federal budget outlook with the net cost of federal debt increasing by about $150 billion relative to the baseline scenario.

Caveats and Complications

The estimates presented above embed many assumptions.  As a result, the confidence intervals around the specific numerical values cited above are quite large.  That said, the basic message is clear.  In evaluating the fiscal effects of a proposed change like eliminating IOR authority, it’s critically important to consider how such proposals would affect the Fed’s ability to implement monetary policy and the full implications of the proposed change for the Fed’s balance sheet and income.

Moreover, there are good reasons to believe the net effect of a transition from ample to scarce reserves on the federal budget would be far larger than the $150 billion loss estimate developed above for Fed remittances.  The $2.5 trillion in securities sales associated with this transition would likely put considerable upward pressure on longer-term interest rates.  The pass through to mortgage rates could be somewhat larger still if the securities sold included large volumes of agency MBS.  In addition, an increase in yields of that magnitude would significantly increase the interest cost on new debt issued by the U.S. Treasury.  More generally, a rapid sale of $2.5 trillion in longer-term Treasury securities could well create substantial strains in financial markets that would have broad economic and fiscal repercussions.

Conclusion and Coda

The estimates above incorporate a number of assumptions and are intended only to provide a sense of rough orders of magnitude.  Nonetheless, the calculations illustrate why focusing on the Fed’s full income statement rather than just its expense is critical in assessing the fiscal implications of eliminating the Fed’s IOR authority.  The bottom line conclusion in the analysis above is that transitioning to scarce reserves would almost certainly reduce Fed remittances over the longer-run.  Even in near-term projections over the 2026-2035 period, apart from some special cases, remittances under a scarce reserves regime also seem very likely to run lower than under an ample reserve regime, particularly in light of the potential for substantial realized losses on sales of securities that would be required to implement a transition to scarce reserves.

The estimates presented above are largely focused on very narrow direct accounting costs to the federal government.  Transitioning to scarce reserves in a short period of time would likely greatly impair the Fed’s ability to implement monetary policy effectively and also put significant upward pressure on longer-term rates.  These types of effects would likely have adverse repercussions for the economy, financial markets and financial institutions that would, in turn, amplify the total costs to the government in eliminating IOR authority.

As a final word, the proposal to eliminate the Federal Reserve’s IOR authority did not find its way into the final version of the just-passed fiscal legislation.  It bears repeating that the Congress has provided statutory goals for the Fed to pursue in conducting monetary policy.  The Fed is directed to conduct monetary policy to promote maximum employment and stable prices.  In assigning these objectives, the Congress recognized that achieving maximum employment and stable prices would also foster moderate long-term interest rates.  Importantly, the Congress did not direct the Fed to conduct monetary policy to minimize the net interest costs of federal debt.

Of course, during World War II, the Fed played a key role in supporting wartime finance during a national emergency.  After the war, however, requirements to support the Treasury’s debt financing increasingly became a constraint on the Fed’s ability to promote price stability.  Eventually, this tension was formally resolved with the Fed-Treasury Accord of 1951.

Even after the Accord, there was a long period involving efforts of various Administrations and members of Congress—perhaps most famously including the decades-long campaign of Representative Wright Patman—to limit the Fed’s independence and enjoin the Federal Reserve to provide direct support for government financing.  As discussed above, the proposal to eliminate the Fed’s IOR authority is misguided even from the narrow perspective of minimizing the net interest costs of federal debt. But of far greater importance, such proposals would undermine the independence of the Fed and its ability to conduct monetary policy free of political influence and for the benefit of all Americans.  Unfortunately, given the pressures stemming from the unsustainable path of federal deficits and debt, we almost certainly have not seen the last of political efforts to enlist the Fed in various schemes to support federal finances.

Methodology

The calculations underlying Table 1 and 2 above are based on simple balance sheet projections and a set of assumptions regarding interest rates.  In the first step, we project the size and composition of the Federal Reserve’s balance sheet at the end of 2025 assuming that the FOMC continues to reduce the size of the balance sheet following its current procedure for the remainder of this year.  Reserves are assumed to reach the ample level by the end of 2025.  In subsequent years, the SOMA, currency, reserves, the Treasury General Account (TGA), and Other Liabilities are assumed to grow at the rate of nominal GDP projected by the CBO.  Agency MBS is assumed to continue to run off the balance sheet following the pattern projected in the Federal Reserve Bank of New York Annual Report (Annual Report on Open Market Operations – FEDERAL RESERVE BANK of NEW YORK).  Holdings of Treasury securities are determined as the difference between the projected size of the SOMA and projected agency MBS holdings.  The full balance sheet projections in the baseline are as shown below.

Note: $ Trillions

The income projections assume a path for interest rates similar to that projected by the CBO.  The rate for longer-term Treasury securities is projected at 3.40 percent over the entire period from 2026-2035.  The interest rate on reserves is projected at 3.40 percent in 2026 and 3.00 percent in subsequent years.  For simplicity, the interest on reserves rate is also applied to the catch all category “other liabilities.”  The return on legacy agency MBS holdings is set at 2.2 percent in every year—the effective return on agency MBS reported in the Federal Reserve’s quarterly financial statements.  Note that this return incorporates the gradual decline in the unamortized premium for such securities.  Similarly, the return on the Federal Reserve’s legacy holdings of Treasury securities is set at 2.2 percent, again matching the effective return on legacy Treasury securities holdings reported in the Fed’s quarterly financial statements.  The return on legacy Treasury securities also incorporates a gradual decline in the unamortized premium.  The Federal Reserve’s projected legacy holdings of Treasury securities are based on data for maturing issues of Treasury securities reported on the Federal Reserve Bank of New York’s website (System Open Market Account Holdings of Domestic Securities – FEDERAL RESERVE BANK of NEW YORK).  New Treasury securities purchased to maintain ample reserves in the outyears are assumed to yield 3.40 percent.  The effective return in the Federal Reserve’s holdings of securities is calculated as a weighted average of the yields on legacy Treasury and agency MBS holdings and the yield on holdings of new Treasury securities.  The Federal Reserve’s net interest income determines the bulk of its total net income.  However, there are some non-interest expenses reflecting salaries and other operating expenses.  Projections for these expenses are based on data reported in the quarterly financial statement for 2025 Q1 with assumed growth again matching projected nominal GDP growth over time.  Based on these assumptions, the projected income for the Federal Reserve is shown below.  Total remittances amount to about $740 billion over the 2026-2035 period, comparable to the CBO projection for remittances over that window in published data released earlier this year.  The net income projections include a projection for the deferred asset of $250 billion at the end of 2025.  Over the period 2026-2028, the deferred asset is paid down and remittances to the Treasury resume in 2029.

Note: $ Trillions

In the alternative scenario, the Federal Reserve is assumed to sell $2.5 trillion in Treasury securities to effect a return to a scarce reserves regime.  The balance sheet projections in this scenario are shown below.

As discussed, above, the sale of Treasury securities results in realized losses of $350 billion; as a result, the deferred asset jumps from $250 billion in 2025 to $520 billion in 2026.  The deferred asset is paid down over the period 2027-2030 and remittances to the Treasury resume in 2031.  Cumulative remittances to the Treasury total $590 billion or $150 billion lower than in the baseline scenario.

Note: $ Trillions

---