Federal Reserve Policy in a World of Low Interest Rates.

• Author: Sims, Eric R.

The Federal Reserve (Fed) is tasked with maintaining price stability and achieving maximum employment. In practice, over the last decades, the Fed has sought to achieve its objectives primarily through the manipulation of a short-term interbank interest rate, the federal funds rate (FFR).

At die height of the Great Recession of 2007-2009, the Fed pushed its benchmark policy rate to zero. With its principal tool unavailable, the Fed resorted to a sequence of unconventional policy actions in an attempt to provide further stimulus to the economy. These actions included large-scale asset purchases (more commonly referred to as quantitative easing, or QE) and forward guidance. These programs were viewed by most as solutions to the temporary problem of the zero lower bound (ZLB) on the short-term policy rate. Market participants never expected the ZLB to last more than a couple of years (Bauer and Rudebusch 2016; Wu and Xia 2016), but in actuality the FFR was at zero for seven years. And though the Fed began raising the FFR at the end of 2015, it has since cut it three times, and at present, the FFR sits less than 200 basis points above zero. Markets are expecting further rate cuts in the near future.

A substantial body of research finds that the so-called natural rate of interest, or sometimes "r-star," is on a continuing secular downward trend. Figure 1 plots the estimate of the natural rate from Laubach and Williams (2003) through the second quarter of 2019. The dashed lined is a best-fitting trend line, and shaded gray regions are recessions as dated by the National Bureau of Economic Research (NBER). While the Laubach-Williams estimate of r-star declined substantially in the wake of the Great Recession, this decline is part of a longer-run downward trend. In standard models, optimal policy entails adjusting the policy rate to track movements in the natural rate. With the natural rate hovering so close to zero, there is little room for conventional policy rate cuts should the need arise.

All signs therefore point toward an extended period in which interest rates are significantly lower than their average levels from the 1980s to 2000s. This means that the problem of the ZLB and the inability to push the FFR down in response to deteriorating economic conditions is likely to arise again. As a consequence, the Fed must move away from its conventional operating framework--for example, by significantly raising its inflation target, experimenting with negative rates, or more regularly using unconventional tools like QE as a substitute for conventional rate cuts at the ZLB. Which of these options should the Fed and other central banks choose?

The Problem of the ZLB and Policy Proposals to Avoid It

The macroeconomic models popular prior to the Great Recession--chiefly, New Keynesian dynamic stochastic general equilibrium (DSGE) models--were developed in the context of the Fed's precrisis operating framework. (1) These models feature one short-term interest rate (the policy rate) and abstract from the myriad debt instruments that are ubiquitous in modern economies. Decision rules for consumption and investment are derived from microeconomic decision problems. Nominal rigidities in the form of price and/or wage stickiness are introduced, giving rise to a Phillips curve-type relation between inflation and resource utilization. Monetary policy is characterized via some sort of rule, such as the famed Taylor (1993) rule, for the short-term policy interest rate.

The ZLB poses a substantial constraint for stabilization policy in these models. After all, there is only one policy instrument, and at the ZLB, this instrument is unavailable. Kiley and Roberts (2017) survey the costs of the ZLB in standard New Keynesian DSGE models and conclude that they are sizable. At the ZLB, the economy is much more susceptible to adverse demand shocks, and supply shocks can have nonintuitive effects on output and other aggregates (Cochrane 2017; Wieland 2019). Furthermore, at the ZLB, the economy can get stuck in a self-fulfilling trap of deflation and negative output gaps (Benhabib, Schmit-Grohe, and Uribe 2001).

Based on the prevailing view that the ZLB imposes substantial costs, many economists have pushed for policy changes meant to reduce both the likelihood and length of ZLB episodes. One popular proposal is to raise the Fed's long-term inflation target. The logic behind such proposals is the celebrated Fisher relationship, which says that the nominal interest rate equals the real rate plus the rate of expected inflation. For a given real rate, higher expected inflation raises the nominal rate one-for-one. An inflation target of say, 4 percent instead of 2 percent, would therefore give the Fed and other central banks an average of two more percentage points of room for rate cuts before hitting the ZLB. (2)

A number of other economists have argued for wider implementation of negative interest rates as a policy tool (e.g., Kimball 2017; Rogoff 2017; and Agarwal and Kimball 2019). Conventional wisdom holds that the existence of currency paying a zero nominal return places a floor of zero on interest rates on other assets. Contrary to this wisdom, a number of central banks--with the Fed being a notable exception--have successfully implemented negative policy rates without much trouble, and at present, a large amount of sovereign debt is trading at mildly negative yields. Nevertheless, rates have not gone substantially negative anywhere in the world, and the existence of a zero-yielding substitute like cash, as well as other features of financial markets and institutions, likely puts a cap on just how far into negative territory rates can fall. For this reason, some economists have called for the (near) abolition of paper currency (e.g., Rogoff 2016).

The elimination of barriers on how negative nominal interest rates can fall might entail significant changes in central bank operating procedures, but if successfully implemented, it would render the problem of low rates moot--the Fed and other central banks could adjust policy by moving rates up or down as needed without regard for a binding floor, zero or otherwise. Increasing the inflation target by a few percentage points would give the Fed significantly more room to cut rates in the face of deteriorating economic conditions without having to worry about pushing rates into negative territory. Enabling deeply negative rates or increasing the inflation target would both represent a significant departure from Fed practice and would certainly entail some potentially large costs. But if the ZLB poses a substantial threat, perhaps those costs are worth incurring.

The Fed's Unconventional Policy Actions

The large costs of a binding ZLB presuppose that central banks cannot do anything...