Nominal interest rates are very low, and in a world of excess global saving, low inflation, and high demand for safe assets like government debt, there’s a good chance that they will be low for a long time.
That fact poses a potential problem for the Federal Reserve and other central banks: When the next recession arrives, there may be limited room for the interest-rate cuts that have traditionally been central banks’ primary tool for sustaining employment and keeping inflation near target.
That concerning possibility has led to calls for a new monetary policy framework, including by Fed insiders like John Williams, president of the San Francisco Fed.
In particular, Williams has joined Olivier Blanchard and other prominent economists in proposing that the Fed consider raising its target for inflation, currently 2 percent.[1] If the Fed targeted a higher average level of inflation, the reasoning goes, nominal interest rates would also tend to be higher, leaving more room for rate cuts when needed.
Few fans of negative rates
Interestingly, some advocates of a higher inflation target have been dismissive of the use of negative short-term interest rates, an alternative means of increasing “space” for monetary easing. For example, in a recent interview in which he advocated reconsideration of the Fed’s inflation target, Williams said: “Negative rates are still at the bottom of the stack in terms of net effectiveness.”
Williams’s colleague on the Federal Open Market Committee, Eric Rosengren, also has suggested that the Fed may need to set higher inflation targets in the future while asserting that negative rates should be viewed as a last resort. My sense is that Williams’s and Rosengren’s negative view of negative rates is broadly shared on the FOMC.
Outside the United States, Mark Carney, governor of the Bank of England, has expressed openness to targeting nominal GDP (which essentially involves targeting a higher inflation rate when GDP growth is low), but has also made clear that he is “not a fan” of negative interest rates.
As I explain below, negative rates and higher inflation targets can be viewed as alternative methods for pushing the real interest rate further below zero. In that context, I am puzzled by the apparently strong preference for a higher inflation target over negative rates, at least based on what we know now.
Yes, negative interest rates raise a variety of practical problems, as well as political and communications issues, but so does a higher inflation target. In this post, I argue that it’s premature for policymakers to emphasize the option of raising the inflation target over the use of negative rates. Pending further study about the costs and benefits of both approaches, we should remain agnostic about whether either or both should be part of the Fed’s policy framework.
Comparing a strategy based on a higher inflation target with the use of negative rates is natural because, as just mentioned, they work through the same channel. Economic theory suggests that aggregate demand (consumption and investment) responds to the real rate of interest, which is the nominal (market) interest rate minus the public’s expected rate of inflation.
As I noted in my earlier post on negative rates, the Fed has routinely set the real federal funds rate at negative levels (i.e., with the nominal funds rate below inflation) to fight recessions. However, with the inflation target at its current level of 2 percent, and assuming that the Fed does not set its policy rate lower than zero, the Fed cannot reduce the real policy rate below -2 percent, i.e. a zero nominal rate less 2 percent expected inflation.
History, including the experience of the past few years, suggests that—in the absence of a robust fiscal response—that may not be enough to deal with a bad recession. To reduce the real policy rate further, the Fed would either have to lower the nominal interest rate into negative territory, raise expected inflation (by raising the inflation target), or both.
Since negative nominal rates and a higher inflation target both serve to reduce the lower (negative) bound on the real interest rate achievable by monetary policy, they are to some extent substitutes.
The argument for negative rates
Which approach is preferable? Without trying to be exhaustive, I’ll briefly compare them on four counts: ease of implementation, costs and side effects, distributional effects, and political risks. I find that negative rates are not clearly inferior to a higher inflation target and may even be preferable on some dimensions.
Ease of implementation. Negative interest rates are easy to implement. In practice, central banks in Europe and Japan have imposed negative short-term rates by deciding to charge (rather than pay) interest on bank reserves, an action that is clear, concrete, and essentially instantaneous.
Experience suggests that the effects of imposing negative rates on reserves also spread fairly quickly to other interest rates and asset prices. Like other central banks, the Fed pays interest on bank reserves and presumably could use a similar approach—essentially charging banks to keep reserves at the Fed—to enforce a negative policy rate.
In contrast, while the Fed could announce at any time that it is raising its inflation target, the announcement would not increase the Fed’s ability to lower the real interest rate unless the public’s inflation expectations changed accordingly.[2] But, as the Japanese experience has shown, inflation expectations may adjust slowly or incompletely to announced changes in target, especially if actual inflation has been very low for some time.
The public might also have reasonable doubts about the Fed’s ability to reach the higher target or about the willingness of the Congress or future Fed policymakers to support a higher inflation goal, both of which would reduce the credibility of the new target and thus its ability to influence expectations.
Which approach provides, potentially, more policy “space” for the Fed? Some advocates of a higher inflation target, such as Blanchard, have proposed increasing the target to as much as 4%, which would allow a real policy rate as low as -4 percent, if the nominal rate is zero.
The extent to which rates can be pushed negative, in contrast, is constrained by the fact that households and businesses can always choose to hold cash, which pays a zero nominal interest rate, rather than securities. To date we have not seen policy rates below -0.75% (Switzerland), equivalent to a -2.75% real policy rate if expected inflation is 2%. That comparison favors a higher inflation target, obviously.
On the other hand, it is not clear that an inflation target as high as 4% would be politically tenable and hence credible in the U.S. or other advanced economies, whereas arguably feasible institutional changes, some as simple as eliminating or restricting the issuance of large-denomination currency, could expand the scope for negative rates. The question of which approach creates more policy “space” is thus still somewhat open. Of course, nothing rules out using some combination of the two strategies.
Costs and side effects. Negative rates and a higher inflation target both have costs and side effects. As I discussed in my earlier post, negative rates can create problems for money market funds, banks, and other financial institutions, costs that would have to be managed if rates remained negative for very long. These concerns are legitimate, since effective transmission of monetary policy requires a properly functioning banking and financial system.
For what it’s worth, the effects of negative rates on banks’ net interest margins in Europe appears to have been moderate thus far. There are also means by which central banks can limit the effects of negative rates on bank profits—by charging a negative rate only on a portion of bank reserves, for example, as the Bank of Japan has done.
Higher inflation has costs of its own, of course, including making economic planning more difficult and impeding the functioning of markets. Some recent research suggests that these costs are smaller than we thought, particularly at comparatively modest inflation rates. More work is needed on this issue.
Higher inflation may also bring with it financial stability risks, including distortions it creates in tax and accounting systems and the fact that an unexpected increase in inflation would impose capital losses on holders of long-term bonds, including banks, insurance companies, and pension funds.
In comparing the costs and side effects of the two tools, a difference worth keeping in mind is that negative rates would be in place only in periods when they were needed (i.e., when the zero lower bound on interest rates would otherwise be binding), while higher inflation (assuming it could be achieved) would be a permanent condition, affecting the economy in good times as well as bad.
Changing the inflation target also carries the risk of being perceived as opportunistic, which could result in inflation expectations becoming unstable. Less-anchored inflation expectations would make inflation harder to control and give the Fed less scope to use monetary policy to offset fluctuations in employment.
Distributional effects. Either policy would give the Fed more scope to fight recessions and keep inflation near target, potentially providing broad benefit. On the margin, though, the two approaches would differ in their distributional implications, with the net effects difficult to assess.
The most direct costs of higher inflation are borne by holders of cash, and, again, with a higher inflation target those costs would be experienced at all times, not just during recessions. More generally, less wealthy people may find it more difficult to protect themselves from inflation. In contrast, negative rates would probably most affect more financially sophisticated and market-sensitive firms and households. In particular, banks would probably not pass on negative rates to small depositors, with whom they want to maintain profitable long-run relationships, but instead would more likely impose negative rates on “hot money” investors who place less value on longer-term relationships.[3]
The transition to a higher level of inflation would hurt holders of bonds and other non-indexed assets while providing a windfall for debtors, including mortgage borrowers. In the medium term, nominal returns to saving (including the investments of pension funds, life insurance companies, etc.) would be higher with a higher inflation target, but the real (net of inflation) returns received by savers would be similar under either regime.
Political risks. Both negative rates and a higher inflation target would be politically unpopular, possibly leading to reduced support for the policies of the central bank and for its independence. In particular, as already noted, the credibility of a higher inflation target could be reduced if political support for it were seen to be tenuous. Political viability is thus an important concern in judging these policy options.
In the political sphere, the fact that negative rates would be temporary and deployed only during severely adverse economic conditions would be an advantage. Like quantitative easing, which was also unpopular in many quarters, a period of negative rates would probably be tolerated by politicians if properly motivated and explained.
We have some evidence on this point: Negative rates are disliked by many in Europe and Japan but central banks have been willing and able to use them without facing high political costs, at least so far.
In contrast, a higher inflation target would be a permanent, or at least very long-lasting change, not restricted to an emergency; and it would raise questions about the flexibility of the Fed’s legal mandate to achieve price stability. It thus might need explicit approval or at least some sort of review from Congress.
A possibility, recently proposed by a comprehensive study on monetary policy options, would be to set up a commission to assess potential changes in the Fed’s policy regime and to report to Congress and the public.
Although commissions can serve important public purposes, proponents of a higher inflation target should be careful what they ask for. In the United States, as in Europe, there is a substantial element of public opinion (well represented in legislatures and even in the central banks themselves) that holds that central banks should concern themselves only with inflation, and that efforts to use monetary policy to stabilize employment are illegitimate or impractical.
These views have manifested as opposition to the Fed’s accommodative policies in recent years, and even in legislative efforts to eliminate the employment part of the Fed’s dual mandate. Holders of this perspective would be unimpressed by the cost-benefit analyses of the Keynesian proponents of a higher inflation target.
To the contrary, they would strongly oppose choosing higher inflation in order to give the Fed more room to respond to employment fluctuations, and indeed might seek a lower target. In their efforts they would be aided by the public’s money illusion (the tendency to confuse general inflation in both wages and prices with changes in real wages). Whatever the abstract merits of a higher inflation target, if it is not politically achievable then it is of no benefit.
Conclusion. It would be extremely helpful if central banks could count on other policymakers, particularly fiscal policymakers, to take on some of the burden of stabilizing the economy during the next recession. Since that can’t be assured, and since the current low-interest-rate environment may persist, there are good reasons for the Fed and other central bankers to consider changes in their policy frameworks. The option of raising the inflation target should be part of that discussion. But, as I have argued in this post, it is premature to rule out alternative or potentially complementary approaches, including the possibility of using negative interest rates.
© 2016 The Brookings Institution.
