Central banks have come a long way, but not far enough
PROGRESS in the practice of monetary policy occurs one disaster at a time. From the depression of the 1930s economists learned that money matters, and that a contraction in the money supply can produce a painful downturn. From the inflation of the 1970s economists learned that inflation is a monetary phenomenon which can be controlled through the proper application of monetary policy. One has the sense that altough the world’s present disaster is coming—slowly, fitfully—to an end, central bankers are still quite a long way away from understanding how they failed and how they might do better in the future. The most recent moves from the Bank of England and the Federal Reserve suggest that this particular revolution remains unfinished.
Monetary lessons are never learned quickly—unfortunately, since an earlier understanding of the nature of the disasters might allow policymakers to change course more quickly and prevent a lot of human suffering. Governments began experimenting with reinflation in 1932-33, touching off a proper recovery after four long years of depression, but America blundered back into recession in 1937 after tightening monetary policy once again. Only when monetary policy was entirely subordinated to the war effort was the depression well and truly beaten. And only in the decades after the war would economists begin articulating the connection between monetary contraction, deflation, and depression; Milton Friedman and Anna Schwartz’s “Monetary History of the United States” was not published until 1963.
Inflation began its long march upwards in the mid-1960s and raged in double digits, across the rich world, for much of the 1970s. Not until the early 1980s did governments begin wringing inflation out of the system through monetary tightening. In that case, the intellectual heavy lifting preceded the policy experimentation. Monetarists and rational expectations theorists had ready explanations for galloping inflation, but central bankers took convincing that hammering economies with tighter policy would make a difference.
The world’s most recent economic disaster began in the late 1990s, when the Japanese economy fell into a liquidity trap and embarked on its lost decade. Economists quickly set to work diagnosing the Japanese economy and building policy recommendations. But arguments over just what Japan’s crisis meant were unresolved when a Japan-like disaster descended on much of the rest of the rich world. And central banks have only gingerly and belatedly begun trying out ideas meant to bring the crisis to a quick and definitive end.
The rich world’s disaster hinges on the problem of zero, in two ways. First, as the interest rates under the control of the central bank approach zero, policymakers are at a loss as to how to continue loosening policy. And second, when inflation approaches zero the relationship between economic weakness and disinflation breaks down. When inflation is high a surge in unemployment will quickly slow price increases, but when inflation is low wage and price rigidities blunt the effect of unemployment, leaving inflation targeting central banks uncertain how to proceed.
These puzzles provoked an ongoing debate. In one corner stands a group with a long and proud history: those ready to give up on monetary policy. Some have been in the corner since rates first touched zero in 2008 while others have joined over time as unconventional policies failed to bring back full employment. But like those who said monetary policy couldn’t fix the depression or rein in inflation, this bunch reckons that zero means it’s time for other strategies, or to simply accept present suffering.
In another corner there are the mechanics, who reckon central banks have their framework right and only need to find new tools. If inflation falls dangerously below target while the main policy rate is near zero, then one should simply target longer rates and bring them down to zero. If even more oomph is needed, then the central bank can turn to rates on private (rather than government) loans and securities, and drive them down to zero. This group won the day in Japan, where the central bank debuted “quantitative easing” in the early 2000s. It held the early edge elsewhere in the rich world in the aftermath of the Great Recession; both the Fed and the Bank of England reached for QE after their benchmark rates dropped to near zero.
Then there are those who emphasise expectations. It is not a controversial statement in central banking to note that the stance of monetary policy is determined by expectations of how a central bank will react to changing economic circumstances. But that of course implies that by telling markets about what sorts of things will trigger rate increases in the future, a central bank can tighten or loosen monetary policy in the present, all without adjusting any inerest rates. So if markets are behaving—that is, saving, borrowing, and investing—as if the central bank is likely to start tightening when x, y, and z thresholds are crossed, and the central bank then says to markets that it is actually prepared to allow expansion beyond those thresholds before raising rates, then markets should change their behaviour immediately.
This bunch has been ascendent over the past two years. Beginning in 2011 Federal Reserve policy statements included an approximate calendar date on which interest rates would begin to rise; in August of 2011 rates were said to be likely to stay low until at least mid-2013. For the next year the Fed adjusted its calendar guidance in response to economic conditions. Then in December of last year it switched things up, changing its guidance to revolve around economic variables, namely, the level of unemployment and inflation. Rates would stay low, the Fed said, until the unemployment rate had fallen to at least 6.5%, provided that inflation expectations were no more than half a percentage point above the Fed’s 2% target. Fed officials have since indicated that ongoing QE will also be pegged to particular changes in the economy, timed to end as the unemployment rate falls to 7%. This week the Bank of England embraced a similar approach. The Bank’s main policy rate will not rise and QE will continue until the unemployment rate has fallen to at least 7%, provided that inflation two years ahead is not expected to be more than half a percentage point above 2% and the financial system is not judged to be tilting toward instability.
The expectations crowd seems to have gained an edge among policymakers for a few different reasons. First, central bank assessments of the risk-return profile of QE seem to have been worsening. Central bankers never seemed particularly comfortable splashing out on government bonds, but appear to have been less than impressed with the results of their purchases and sick of hearing complaints about unpleasant QE side effects (real and imagined), from financial-system distortions to looming hyperinflation. Some may also have worried that in the absence of clear communication about future policy QE’s effects were being blunted. Markets might react to expansionary QE by simply moving forward the date on which they expected policy tightening, thereby cancelling out the good done by the asset purchases. Better expectations management could help secure some of the QE gains. And lastly, the thresholds approached has given hawkish central bankers an exit strategy to cling to: a sense that all the largesse will end, and perhaps in the not-too-distant future.
But the Fed has been tinkering with expectations management for a while now without generating the rapid recovery one would hope for from appropriate monetary policy. While the Bank of England approved of the Fed’s innovations enough to adopt them for its own, Governor Mark Carney seemed to go out of his way to warn not to expect too much from the new guidance.
And yet the sun has yet to shine on those lingering in the last corner, who suggest that maybe there is something slightly more fundamental amiss with the current monetary policy approach: namely, that it still has a 2% inflation target at its heart.
That a 2% inflation goal might be a major culprit in the rich world’s current economic mess is a hard thing to stomach for economists and policymakers who came of age in the 1970s. Yet there are good reasons for fingering the 2% rate as a real problem. Troubles began before the crisis, as the near-total absence of inflation pressure helped usher interest rates steadily downward, leaving much less breathing space between the Fed’s policy rate and the dread zero. The Fed’s main policy rate was just 4.25% when the recession began, giving the central bank little room to cut rates to battle the downturn. Economists have had to raise their estimates of just how often an economy will end up stuck with interest rates near zero when central banks target low rates of inflation.
A second difficulty emerged soon after: low and stable inflation translates into very rigid wages and prices, even in the face of big declines in demand and soaring unemployment. When the variability of inflation drops (as it did, dramatically, from the early 1980s on) workers and firms get in the habit of adjusting prices less frequently, they drop automatic wage indexing, and they generally increase the rigidity of the economy-wide price level. This rigidity is especially problematic at near-zero inflation rates given resistance to accepting absolute cuts in wages and prices. As a result even the epic economic collapse of late 2008 and early 2009 translated into relatively moderate and delayed disinflation. And that meant big trouble for economies in which central banks used deflation as their depression canary-in-a-coal-mine. Prior to the crisis many central bankers would have insisted that there could be no demand collapse if inflation were kept above 1% and that keeping inflation above 1% would be sufficient to prevent a demand collapse. But they were wrong.
And though the link between sufficient demand and stable inflation would seem to have been called into question, central bankers have yet to update their mental models accordingly. They have evolved a bit. As the shift toward dual thresholds, including unemployment as well as inflation, indicates, central bankers are recognising that there is information out there about demand than is contained in the inflation rate. And yet their evolution is incomplete, because they insist on targeting a higher level of demand subject to continued fidelity to the inflation target. They have given themselves a bit of wiggle room by saying they will tolerate expected inflation a shade above 2%. But what they have manifestly not done is declare that they want more demand, whether or not that happens to involve inflation sustained above 2%.
Within the pre-crisis policy paradigm that view makes perfect sense. Rising inflation, in that worldview, is a sign that the economy is approaching its structural limits: is growing as fast as it potentially can. Goosing demand over and above that level is worse than useless; any growth that results is unsustainable and will fuel accelerating inflation.
But the crisis ought to have led central bankers to reconsider this view. For one thing, as noted above, the relationship between inflation and unemployment is not as clear as one would think at very low inflation rates. Study after study, including many produced by Fed economists, indicates that most of the gap between the current unemployment rate and the “normal” level is temporary—there really is quite a lot of labour-market slack—and yet inflation is not much below target. If inflation is less informative about slack than it used to be, then a central bank interested in getting rid of slack should probably discount the inflation signal to some extent.
Or to put things somewhat differently, an inflation rate of 3% or 4% is not in and of itself dangerous. It could be worrisome if it seemed to signal an economy being pushed beyond its capacity. But if there is good reason to believe that the economy is not close to capacity, then there is no particular reason to fear 3% inflation or 4% inflation. Inflation accelerating from 3% through 4% and beyond yes, but of all the problems rich economies have had over the past half decade turning a rising inflation rate to a falling one has not been among them.
It is worth acknowledging that there can be costs to a high but stable rate of inflation relative to a low but stable rate. If some prices adjust less quickly or easily than others, then a higher rate of inflation may entail greater distortions in relative prices that entail some efficiency costs. Such things need to be kept in perspective, however. The American economy, for instance, has been operating with an output gap close to $1 trillion for half a decade now. Even if that estimate is wildly overstated, and the actual gap is closer to $200 billion or 1.3% of GDP, that cost probably swamps any damage from relative-price distortions.
Neither is the relationship between slack and inflation the only thing to consider. With interest rates near zero it may be very difficult to raise demand without lifting inflation expectations above the 2% rate. Normal monetary policy operates on the principle that there is a “market clearing” real interest rate (or set of real market rates), that balances desired saving and desired borrowing and prevents resources—or willing workers—from sitting around idle. The central bank’s normal goal, then, is to adjust its policy rate to nudge market rates toward that market-clearing level. But what if the market-clearing real policy rate is -3% and the actual policy rate is stuck at 0.5% while inflation is just 1.5%? In that case the central bank’s policy rate is way too high, and will remain way too high until either the market-clearing rate moves back toward positive territory or the central bank makes up its mind to lift inflation expectations. With an expected inflation rate of 3.5% the real policy rate drops to the appropriate level and the economy should leap in response.
But that may seem an unlikely story. Or one could simply wonder about the underlying health of an economy in which investors can’t find any attractive positive-return investments and must be cajoled into splashing out by the threat of inflation-driven loss of principal. The simpler way to conceptualise the current situation may simply be to conclude that inflation isn’t as meaningful a concept as we thought. If we’re interested in demand, and the reason we’re interested in demand is that too much of it causes accelerating price increases and too little of it causes unnecessary joblessness, and we’re pretty certain that there is too little demand despite the fact that the our traditional metric for assessing demand is signalling that all is well…well, perhaps that traditional metric isn’t useful anymore, at least not for assessing demand.
One would then need an alternative metric, which of course takes us to the nominal output brigade. Nominal GDP, or total spending or income in the economy in dollar terms, has long been one of the gauges on the central banker’s dashboard. It is hard to think of a better measure of economy-wide demand than the total amount of money spent each year in dollar terms. Supporters of an NGDP target (either a rate of growth or a trend level) point to several advantages. One, which should be clear already, is that there is little risk of monetary policy fumbling situations like the present in which demand swings are not entirely reflected in changes in inflation. Another is that by anchoring market expectations around a demand path rather than an inflation path the business cycle should be substantially attenuated. And another is that a focus on NGDP makes for better management of supply shocks, since it implies looser policy, relative to an inflation target, when the economy is already being hit by a negative supply shock (rather than a tighter policy which would amplify the blow of the supply shock), and tighter policy when the economy is enjoying a positive productivity shock, which could conceivably prevent economic exuberance from taking a turn toward the irrational. The upshot of a switch to NGDP targeting at the present moment would be much more expansionary policy; the central bank would focus on targeting the obvious shortfall in demand and ignore the misfiring inflation signal.
The advantages of the NGDP approach have not been lost on today’s central bankers. Some will admit that flexible inflation targeting, particularly a strategy built around thresholds, is either similar to or could be adapted to mimic an NGDP target. Obviously, the inclusion in Fed and Bank of England policy of an unemployment threshold is an admission that inflation targeting, flexible or otherwise, isn’t getting the job done. Yet both central banks have declined to embrace the most important and powerful aspect of the NGDP target: the freedom to tolerate or even court higher inflation without giving up the nominal anchor—the NGDP target—that you need to keep market expectations from flying off into the atmosphere, bringing back the 1970s.
It would be presumptuous to say that the inevitable, glorious conclusion of the ongoing revolution is an NGDP target, though that may turn out to be true. A general rise in inflation targets, to 3% or 4%, and a shift toward more flexibility around the target could be an option. More radical ideas, like a shift to “e-money”—fully electronic currency that would allow the government to use negative interest rates—could eventually win support. Or fiscal policy could evolve to become more responsive to business cycles; control over tax rates on wages, for instance, could be handed to central bankers to better allow them to deploy a “helicopter drop” of cash, or wage-tax rates could be linked to the unemployment rate or other economic variables.
But despite the considerable distance monetary policy-making has come over the past half decade, it seems no closer to recognising the brutal truth: that in the absence of higher inflation or negative nominal interest rates rich economies are going to continue to suffer unnecessary and enormously costly shortfalls in demand, output, and employment.
It isn’t surprising, really, that this realisation has yet to dawn on the world of central banking. Even if central bankers themselves weren’t chilled by the very thought of rising inflation public antipathy to higher inflation or negative rates (especially among older workers and net creditors, among the most influential of groups in today’s rich, democratic societies) is intense. It is an uncomfortable truth that the rich world is to some extent getting the recovery it wants.
But the role of the central banker is absolutely not to give the public what it wants. They’re charged with the duty of serving up unemployment when inflation is too high. By rights they should be as ready and enthusiastic about serving up inflation when, as now, it stands a good chance of eliminating cyclical unemployment. Instead our current crop of central bankers has readily limited itself to providing only the lift in hiring that can be managed without delivering an inflation that starts with a 3. I know of no cost-benefit analysis that justifies the choice, but there it is all the same. If history is any guide, it may fall to the next generation of monetary policymakers to see the error and fix it. Unfortunately for us.