In a recent op-ed, Norbert Michel of the Heritage Foundation and George Selgin of the Cato Institute argue that the Federal Reserve should stop paying interest on excess reserves (IOER) because it reduces bank lending and will increase the odds of a recession. I disagree. As I will explain, for any given level of market interest rates, paying interest on reserves improves the ability of the Fed to conduct monetary policy, which reduces volatility and encourages growth.
Paying interest on reserves does not, as Michel and Selgin argue, “encourage banks to park money at the Fed instead of lending to businesses and household.” The Fed’s influence on bank lending to businesses and households comes through the level of market interest rates, and the central bank achieves a given interest rate level through a combination of setting the quantity of reserves and the level of IOER. While the Fed can, of course, achieve its interest rate objective without IOER — as it did prior to the crisis — such a policy works less well, especially in a crisis, with a corresponding cost to the economy.
Reserve balances are deposits held by banks at the central bank. Paying interest on those balances is a standard monetary policy tool used by 13 of the 14 largest central banks to help keep market interest rates near their respective targets.
What is IOER?
Let’s take a step back first to understand what IOER is.
Reserve balances are a liability of the Federal Reserve. Because currency, the other main Fed liability, is essentially fixed at any point in time, the Fed controls the supply of reserves nearly exactly by adjusting the amount of assets it holds.
The Fed conducts monetary policy by doing three things: setting a target for the federal funds rate, setting the supply of reserves to meet demand at that target, and establishing a lending and deposit rate to put a ceiling and floor on the federal funds rate. IOER puts an approximate floor on the federal funds rate because no bank should be willing to lend reserves for less than it could get simply by leaving the reserves on deposit at the Fed.
Reserve balances are currently very elevated because of the Fed’s large-scale asset purchase programs that it executed during the recession to push down longer-term interest rates and stimulate growth. As a result, the federal funds rate is at or even a bit below the floor established by IOER. For the past year and a half, the Fed has been tightening monetary policy by raising IOER. In December 2015 and December 2016, the Fed tightened policy by raising IOER by 25 basis points, twice. Money market rates have all essentially risen in lockstep.
Why can’t the Fed once again conduct monetary policy without IOER?
It is true that before the crisis, the Fed conducted monetary policy without IOER. But that was in a normal interest rate environment and at a time when the Fed’s balance sheet was much smaller. Under current circumstances, with reserve balances so large, the Fed could not tighten monetary policy without paying interest on reserves.
Overnight interest rates would stay near zero. While the Fed could sell its holdings of longer-term securities to push up longer-term interest rates and drain reserves, the Fed would not be able to control longer-term rates with much precision. Longer-term rates, including household mortgage rates, could jump up sharply, leading to an unwanted slowdown in economic growth.
Should the Fed stop paying interest on reserves once the Fed’s balance sheet normalizes?
The Fed has indicated that it will soon start allowing its balance sheet to shrink by partially ceasing to reinvest principal payments. This process is expected to start sometime this year. As noted in a recent blog post, I agree with Michel and Selgin that the economy would be best-served if the Fed returned to conducting monetary policy with a small balance sheet, in a manner similar to how it did before the crisis. Where we disagree is that I think the Fed should retain and continue to use IOER.
Why? First, for a given target federal funds rate, paying interest on excess reserve balances does not reduce bank lending. The Fed influences the economy by raising or lowering its target for the federal funds rate. When market interest rates fall, economic activity — especially business and household investment — picks up, and bank lending picks up to fund that increased activity. The Fed can achieve its federal funds rate target with a relatively high level of IOER and high level of reserves, or a relatively low level of IOER and low level of reserves. The principal difference will be whether banks satisfy their liquidity needs by holding reserves or by holding other assets. Lending, and the level of economic activity, will essentially be the same in either case.
Essentially, but not entirely, which brings us to the second reason the Fed should use IOER as a monetary policy tool.
By improving the Fed’s ability to control interest rates relative to the pre-crisis situation, when the Fed did not have the ability to pay interest on reserves, IOER reduces interest rate volatility in normal times and especially in crisis times. It is hard to argue that increased volatility and uncertainty are not bad for economic activity.