A Monetary Correction

Thursday, February 22, 2018
David Beckworth

ur economy is finishing its ninth year of expansion following the economic crisis of 2008 and 2009. While this expansion has been long-lasting, it has also been weak by post–World War II standards. Theories abound for this weakness. One common explanation, particularly on the right, is that the Federal Reserve’s attempts to stimulate the economy have had the perverse effect of keeping economic growth low.

In 2007 and 2008, the Federal Reserve reduced the federal-funds rate — the interest rate it targets — from 5.25 to 0.25 percent. It held that rate down for seven years. Since 2015, it has increased it in increments to 1.5 percent. The Fed has also made large-scale asset purchases to ease credit. Most of the Fed’s conservative and libertarian critics allow that some of these moves were justified as a response to a sharp recession, but maintain that the Fed has kept monetary policy too loose for too long. Its actions distorted markets so much as to amount to “financial repression.” With interest rates held low, investors moved to risky assets, including stocks, to reach for yield. Inequality worsened but the real economy did not prosper.

Many of these same critics spent the first years of the recovery warning that easy money would lead to a surge in inflation. With no such surge having occurred, the argument has shifted. The current critique is that the expansion has been artificial, and will be revealed as such as the Fed retreats from its extraordinarily accommodative policies. As interest rates rise to normal levels and the Fed’s asset holdings shrink, that is, we will see how dangerously dependent the economy has become on support from the Fed to achieve even modest growth.

There is an alternative perspective that better fits the facts. The truth is that poor Fed policy has contributed to the weakness of the expansion. But the Fed has erred by keeping money too tight, not too loose.

The best indicator of the stance of monetary policy is the rate of growth of spending throughout the economy. Accelerating growth in spending signals an expansionary policy and decelerating growth a contractionary one. In the 1970s, spending growth was high and rising, leading to high rates of inflation. During the Great Moderation that followed that period, spending grew at a relatively steady pace, averaging 5.3 percent a year. This steady pace promoted macroeconomic stability. Economic actors were able to make and coordinate their plans against a backdrop expectation that the volume of spending would keep growing at the accustomed rate.

The crash of 2008 and 2009, however, saw the steepest decline in spending since the Great Depression. In its aftermath the Fed did not pursue a policy of letting spending bounce back to maintain its average growth rate. The red line in Figure 1 shows what a historically normal spending path would have looked like. A loose-money policy would have led to growth significantly above that line. Instead, as the figure shows, spending grew at a level significantly below the pre-crisis rate. It has risen at 3.7 percent per year during the recovery.

Inflation is another indicator of the stance of monetary policy, and it too has signaled tight rather than loose money. Since the end of the crisis, inflation as measured by the Fed’s preferred indicator — the core PCE price index — has averaged 1.5 percent per year. That rate is below the average level of the previous decades, and, as Figure 2 shows, also below the Fed’s target of 2 percent inflation. Both Fed officials and many outside observers call the low inflation rate a “puzzle” given loose monetary policies. But there is no mystery once you drop the assumption that monetary policy has been loose.The crash of 2008 and 2009, however, saw the steepest decline in spending since the Great Depression. In its aftermath the Fed did not pursue a policy of letting spending bounce back to maintain its average growth rate. The red line in Figure 1 shows what a historically normal spending path would have looked like. A loose-money policy would have led to growth significantly above that line. Instead, as the figure shows, spending grew at a level significantly below the pre-crisis rate. It has risen at 3.7 percent per year during the recovery.

Inflation is another indicator of the stance of monetary policy, and it too has signaled tight rather than loose money. Since the end of the crisis, inflation as measured by the Fed’s preferred indicator — the core PCE price index — has averaged 1.5 percent per year. That rate is below the average level of the previous decades, and, as Figure 2 shows, also below the Fed’s target of 2 percent inflation. Both Fed officials and many outside observers call the low inflation rate a “puzzle” given loose monetary policies. But there is no mystery once you drop the assumption that monetary policy has been loose.

Against Higher Inflation

Wednesday, July 5, 2017
David Beckworth

Does the U.S. economy need more inflation? A group of 22 progressive economists has written a letter to the Federal Reserve urging it to appoint a blue-ribbon commission to study whether the central bank should raise its target for inflation above its current 2 percent. Fed chairman Janet Yellen, in her press conference following the latest interest-rate increase, called it “one of the most important questions” facing the organization. 

The economists’ advice shouldn’t be rejected out of hand, but it should be rejected. They make some valid points in their diagnosis of the ills of the current monetary regime. But the Fed can and should address these problems without raising inflation. 

For most people it is probably puzzling that anyone should want inflation to rise. They consider inflation a hit to their standard of living and don’t want more of it. But the extent to which inflation reduces standards of living depends mostly on what drives it. If an oil embargo raises costs throughout the economy, it will indeed tend to raise prices and cut into real incomes. An inflation driven by the central bank’s creation of extra money, on the other hand, should increase prices and wages very close to proportionally. (We’re using the term “inflation” the way most people and economists do, as a rise in the general price level.) This kind of inflation is unpopular mostly because people assume that they have earned their inflated wages but resent paying inflated prices. 

Which is not to say such inflation is costless. If inflation is unpredictable, businesses and households will have difficulty making long-term economic plans — in part because it creates arbitrary wealth transfers. When inflation comes in higher than expected, the bank that provided your fixed-rate mortgage will take a bath; when it comes in lower, you will. Inflation also raises the burden of taxes on capital, since our laws do not adjust for it: It is possible to pay a capital-gains tax on an asset even when it has not appreciated at all in real terms. Then there are the costs of having to adjust prices to keep up with a constantly changing price level: what economists, thinking of the restaurant business as a good example, call “menu costs.” 

When inflation is both low and stable, however, these costs are manageable. Central banks the world over have converged on a 2 percent target rate for inflation during the last few decades partly as a matter of happenstance, but economists have generally considered 2 percent a reasonable number. The costs mentioned above would not rise much higher with year after year of 2 percent inflation than they would with year after year of 0 percent inflation. (And they would be lower still if our capital-gains taxes adjusted for inflation.) 

The slightly higher rate has a few advantages over the lower one. It can make economic transitions smoother. Employees are more hostile to wage cuts of 2 percent with 0 percent inflation than they are to stable wages with 2 percent inflation, even though these outcomes are economically identical. A little inflation therefore makes it easier for employers to lower labor costs by not giving pay raises rather than by laying people off. 

One of the other reasons central banks preferred 2 percent to 0 percent inflation is that it gave them more flexibility. During a recession, central banks usually cut interest rates in order to stimulate the economy. The higher the interest rate is at the start of the recession, the more they can cut it. Since interest rates are composed of both a real return to savings — the real interest rate — and compensation for expected inflation, a 2 percent inflation target will keep interest rates higher than a 0 percent one.

And that brings us to the case for an even higher inflation target. The supporters note that a higher inflation target would give the Fed even more room to maneuver. International comparisons bolster this case: Countries that have tolerated higher inflation over the past decade have fared better than we have. Australia had some of the same vulnerabilities we did before the Great Recession — a real-estate boom and a surge in household debt — and was more exposed than we were to swings in commodity prices. Yet its economy continued to grow during the crisis. It came into the crisis with a core inflation rate of 4 rather than 2 percent, which meant higher interest rates. Australian monetary authorities could thus cut interest rates more than we did. 

Israel did not avoid the recession, but it had a more robust recovery than we did. Higher inflation played a role there too, albeit a different one than in Australia. After Israeli interest rates fell close to 0 percent, monetary authorities allowed the inflation rate to go as high as 5 percent in 2009. Although this inflation eventually caused interest rates to rise, its immediate effect while interest rates were stuck near 0 percent was to push real interest rates deep into negative territory and thus spur more spending. 

The argument, then, is that both Australia and Israel did better than the United States during the crisis by tolerating higher inflation. Why not tolerate a bit more inflation here, the advocates say, through a higher inflation target? 

The proponents also argue that a higher inflation target is needed because real interest rates have been headed downward for decades. Many economists, especially those who want a higher target, say that slowing productivity growth and an aging population are bringing real interest rates down. 

Over the decades before the Great Recession, inflation averaged 2 percent and the real interest rate roughly 3 percent, making for an interest rate a little above 5 percent. These days it is often estimated that the economy can sustain real interest rates near 1 percent. To get interest rates above 4 percent again would therefore require raising the inflation target to the 3 to 4 percent range. 

These arguments for a higher inflation target are reasonably strong if you accept the premise that keeping inflation stable should be the Fed’s principal task in the first place. There is, however, a superior alternative. That alternative would stabilize the growth of nominal spending: the total amount of dollars spent throughout the economy. The growth rate of nominal spending equals the sum of the rates of inflation and of real economic growth. (In 2015, for example, inflation ran at 1.2 percent and the economy grew by 1.6 percent in real terms, so nominal spending grew by 2.8 percent.) Under a level target for nominal spending, the Fed would commit to keeping total spending growing at a steady rate, say 4 percent. It would commit, further, to correcting for any failure to hit the target. If nominal spending grew by 4.5 percent one year, that is, the Fed would shoot for 3.5 percent the following year. 

This policy would capture the benefits of inflation targeting, such as facilitating long-term economic planning by households and businesses. It would arguably serve that purpose a little better than inflation targeting, since most debt and labor contracts are written in nominal terms. It would be compatible with keeping inflation low and fairly stable. If the economy’s real growth rate over the long term averages 2 percent, then hitting a 4 percent nominal-spending target implies achieving a 2 percent average inflation rate too.

A key difference between the two policies is that a nominal-spending target would allow inflation to fluctuate over the short term in response to movements in productivity. A negative supply shock, such as the oil embargo we mentioned earlier, will push prices up and output down. A strict inflation-targeting central bank would try to keep prices from rising by raising interest rates — at the cost of harming an already weak economy. Something like that happened in 2008: Rising commodity prices led the Fed to refrain from cutting rates in the early months of the recession. A nominal-spending target would have had more tolerance for a short-run increase in inflation. 

A positive supply shock, on the other hand, will pull prices down. In that case, an inflation-targeting central bank, whatever inflation rate it has chosen, will be tempted to lower interest rates to keep inflation from falling. If it does it will overstimulate the economy. This appears to have happened in 2002–04, when a productivity boom made the Fed fear deflation and it responded by holding interest rates too low for too long. Under a nominal-spending target, on the other hand, the Fed would have allowed productivity to reduce inflation. 

A key argument for a higher inflation rate, you will recall, is that real interest rates have declined as the economy’s productive potential has. But those interest rates can move rapidly. The San Francisco Fed uses a measure of the “natural” real interest rate — the rate justified by the economic fundamentals — that fell from 2.12 percent at the end of 2007 to 0.34 percent a year later. That was a rapid decline tied to the business cycle. There is no reason in principle that the real interest rate could not rise rapidly as well. If it did, the argument for a higher inflation target would go into reverse: A lower one would be justified. But changing the target with every change in productivity would nullify the advantage of stability.

A nominal-spending target, on the other hand, would not need to alter in response to changes in productivity. Changes in productivity would change the composition of nominal spending — the more productivity grows, the higher the ratio of real economic growth to inflation — but not the total. 

As for Australia and Israel, it’s worth noting that in both countries nominal spending grew at a steady pace. Spikes in inflation may have been helpful, that is, only insofar as they enabled the stabilization of nominal spending. 

We have not yet mentioned one peculiarity of the argument for a higher inflation target. Even though Yellen signaled her openness to that argument, we have actually been below 2 percent inflation for eight years — which is to say, before, during, and after the Fed’s formal adoption of that target. In some quarters the Fed’s failure to hit its target has led to doubt about whether the Fed even has the capacity to raise inflation any more, in which case its capacity to increase nominal spending would also have to be questioned. This doubt is unjustified: The Fed has repeatedly refrained from taking steps that would have increased inflation further, and taken steps that reduced it. 

A nominal-spending target would make for a stabler macroeconomic environment than an inflation target, for the reasons we have described, and would not require enduring higher inflation rates on average. But no target is going to work as well as it could if the Fed is unwilling to take it seriously. And no blue-ribbon commission is going to supply that willingness. 

Yes, the Fed's Errors Made Recession Worse

Monday, February 8, 2016
David Beckworth

Bold theses should receive skeptical reactions, and ours did. We argued in the New York Times that, contrary to what just about everyone believes, the financial crisis and the Great Recession that blew up the American economy in 2008 were not the necessary consequences of a housing bust. They would not have happened if the Federal Reserve had responded appropriately to increasing economic weakness over the course of that year. Barry Ritholtz (a Bloomberg View colleague), Edward Conard, Mike Konczal and Paul Krugman are among those who criticized our argument. Here we respond.

Things We Did Not Say

Since some criticisms were directed at arguments we didn’t actually make, we should clarify a few things.

First, we are not saying that the right Fed policy would have kept any recession from happening. As we noted, the recession began in December 2007, before the Fed mistakes we discussed: failing to cut interest rates between early April and early October 2008, and even spending much of that time suggesting rate increases were on the way. Our argument, rather, is that these mistakes turned what could have been a mild recession into a “great" one.

Second, we aren't saying that better Fed policy could have prevented serious financial turmoil. Again, we explicitly note that financial stress began before the Fed’s worst errors. Our argument is the errors made that stress much worse.

Third, we aren’t ignorant of the fact that the Fed lowered interest rates between October 2007 and April 2008. We stated it in our op-ed. Again, we were discussing mistakes made after this period.

Fourth, our article did not say that uncertainty about the value of mortgage-backed securities caused the decline of housing prices.

How the Fed Mattered 

By missing these points some of our critics misconstrue our views and make invalid arguments against them. They note, as though it contradicts our story, that Bear Stearns collapsed in March 2008. But that’s entirely consistent with our argument: Stress in the financial sector pre-dated the Fed’s errors, but that stress did not have a severely negative effect on the broader economy. Neither inflation expectations nor nominal spending declined at that time; they declined later, when expected future interest rates rose relative to the natural rate.

Konczal says that the Valukas report on the collapse of Lehman Brothers in September 2008 does not indicate that a looser Fed policy would have staved it off. But here’s what the report says in its introduction: “There are many reasons Lehman failed, and the responsibility is shared. Lehman was more the consequence than the cause of a deteriorating economic climate.” We agree with that assessment. Our view is that raising expected future interest rates was a contractionary move, taken at an especially unfortunate time, and contributed greatly to that climate.

Krugman thinks the behavior of long-term real interest rates contradicts our thesis. They rose in the middle of 2008, but not, he says, catastrophically, as they should have if the Fed were really running a much-too-tight policy. Krugman is incorrect about the implications of our account. We would expect the Fed’s contractionary mistakes to have led to an increase in the risk premium. It did. We would also expect it to reduce the prospects of economic growth and thus lead to a decline in long-term real interest rates adjusted for the risk premium. Again, that’s what happened.

Yes, the Fed Was Wrong

The critics offer different reasons for thinking we are too hard on the Fed. Inflation was “showing unsettling signs of picking up” in 2008, writes Ritholtz, and it’s a “fundamental error” on our part to dismiss the concern the Fed had at the time. But we’re not just applying hindsight: Market expectations of inflation, as measured by TIPS spreads, were declining rather than rising. And those expectations turned out to be correct.

Conard says that monetary loosening in 2008 would not have spurred more lending and would have punished savers. But the decline of lending was an important symptom of the economic crisis, not the cause. Higher nominal income and higher expected nominal income would have increased asset values, which would have increased lending. And savers would have been better off with the higher interest rates they would have received once the economy had strengthened.

Konczal suggests that when Fed officials warned during the spring and summer of 2008 that inflation was rising, they may have helped the economy by increasing the expectations that this would happen. But the Fed’s most powerful way of shaping economic expectations is not by speculating about the future but by indicating what its policy will be. The Fed was signaling that monetary policy would tighten in the future: a contractionary move.

And this brings us to another point. Our critics say or imply that our story just can’t be true: that it’s implausible that the combination of a failure to cut interest rates and some rhetoric about future monetary tightening, even if these were ill-advised, had such disastrous effects. We believe they are thinking about monetary policy too mechanically.

Under most circumstances, the difference for the economy between cutting interest rates and not cutting interest rates would be small. In most circumstances, the difference between taking antibiotics and not taking them would also be small: if, for example, you’re not sick. In certain circumstances, however, the difference is profound.

At a moment of great uncertainty, the Fed signaled that it was more likely to take action to throttle inflation -- a threat markets did not believe existed -- than to prevent a panic. It kept signaling it as that panic grew. And more than signaling: Even after Lehman Brothers collapsed, the Fed’s first policy change was a contractionary one. It started paying interest on excess reserves.

Terminiology, Motives, and Other Distractions

Several of our critics suggest that our argument amounts to saying that the Fed should have prevented a severe recession by taking quicker action, and that we are wrong to say it caused the collapse by tightening money. In truth, we pointed to Fed errors of both omission and commission, but the difference between these types of mistakes does not strike us as especially important in this context. If you don’t turn the ship’s wheel when you’re headed for an iceberg, is that “just” a failure to act?

What matters here is the underlying reality, not the words used to describe it. If our critics come to agree with us that a better Fed policy would have led to a better outcome in 2008, we can agree to disagree on terminology.

Krugman questions our motives. He says that like Milton Friedman before us, we are blaming the Fed for an economic calamity to bolster our laissez-faire ideology. The ideological stakes are not as high, though, as he implies. Even if we are right about the Fed’s responsibility for the Great Recession, it does not mean that, for example, everything was fine in the financial industry and that increasing regulation on it was a mistake. It just means that it is less important to get financial regulatory policy right than we would otherwise have thought, and more important to get monetary policy right.

It is true that, like Milton Friedman, we have views about public policy; it is true that these views are related to one another; and it is true that our political views may have affected our understanding. We daresay all these things are also true of Krugman. Whatever the motives, the arguments have to stand or fall on their own merits. Those merits are strong enough to withstand the criticisms that have been lodged against them.