Re-Targeting the Fed
The quarter-century that preceded the financial crisis of 2008 was a period of low inflation and relatively stable growth, interrupted by just two relatively mild recessions. Economists call it "the Great Moderation," and many of them argue that it was made possible in large part by the fact that, during those years, the Federal Reserve (and many of the world's other central banks) adopted stable regimes of inflation targeting: deciding on a preferred inflation rate and steering the economy toward it, generally by lowering interest rates when inflation fell below the target and raising interest rates when inflation exceeded the target. By sticking to such an approach throughout the Great Moderation, the common argument goes, the Fed effectively balanced a commitment to growth with a determination to keep prices stable.
But that era ended with a bang. In 2008 and 2009, we suffered the most severe recession since the Great Depression, and in its wake we continue to experience high unemployment and weak growth. Most people think they know the proximate cause of this calamity: There was a housing bubble of unprecedented size, which burst in 2006 and led to a severe financial crisis in 2008, which greatly intensified a recession that had begun in December 2007. All of that put together, they believe, was enough to overwhelm the Fed's inflation targeting; there was little the central bank could do to avert the crash, and it has done what it could since to respond to the crisis.
The available evidence, however, does not quite support this familiar narrative. For one thing, the housing crash did not lead directly to a recession or high unemployment. More than two-thirds of the decline in housing construction occurred between January 2006 and April 2008, and yet, during that 27-month period, the unemployment rate rose only slightly, from 4.7% to just 4.9%. Most of the workers who lost jobs in housing construction were re-employed in commercial construction, exports, and services. It wasn't until October 2009 that unemployment soared to 10.1%, as job losses spread across almost all sectors of the economy. The financial crisis (which of course had roots in the housing bubble) and its associated shocks to the system clearly had much to do with that timing, yet somehow this looming crisis did not set off alarm bells at the Fed until remarkably late in the process. Monetary-policy officials simply failed to see the problem coming and did not react nearly quickly enough; even now, the Fed seems to be flailing.
In other words, the economic crisis not only overwhelmed the Fed's inflation-targeting methods, but showed them to be deeply inadequate. Given the consequences of that failure, America's policymakers now need to ask: Could the Fed have done better?
It could have, had it set its sights not on inflation but on nominal gross domestic product — the sum of all current-dollar (non-inflation-adjusted) spending in the American economy. NGDP growth, which is made up of the inflation rate plus "real" (or inflation-adjusted) GDP growth, would have been a better indicator of the severity of the crisis. Between 2008 and 2009, NGDP declined at the fastest rate since 1938, while inflation (at least the "core" inflation rate, which excludes food and gas prices and which the Fed uses as its key inflation measure) raised no red flags. Because inflation cannot quickly adjust to such sudden drops in spending, this decline in nominal GDP brought about a sharp decline in real GDP — thus, a severe recession.
In an important sense, the sharp drop in NGDP precipitated the crisis: It was the proximate cause, even if the housing crisis was the ultimate cause. This suggests that NGDP is useful not only as a predictor and indicator of trouble, but as a target for monetary policy. Setting a goal in terms of nominal GDP could provide a superior alternative to the Fed's current inflation targets.
Why, then, do economists not place greater emphasis on nominal GDP, even as they focus intently on its two components — real GDP growth and inflation? Indeed, even the basic data are surprisingly hard to find: When quarterly GDP figures are announced by the federal government, the press generally reports real GDP growth, but almost never mentions nominal growth. The same is true in the field of macroeconomics, where professional economists frequently work with real GDP (or "output") and prices, but rarely model NGDP.
The reasons for this behavior have as much to do with the politics of inflation as they do with economics. This suggests that the monetary-policy failure of 2008 reflected not so much a failure of technical macroeconomics as an inability of professional economists to understand how their ideas are filtered through our highly politicized policy arena. As we shall see, inflation targeting can work, but only in a world in which policies explicitly aimed at raising the inflation rate are non-controversial. Obviously, we don't live in that world; in the universe we occupy, a nominal GDP target would be more effective.
By considering the past few years through the lens of nominal GDP data, we can both gain a greater understanding of the crisis from which we are still recovering and see our way to a superior approach to monetary policy.
INFLATION AND GROWTH
Although the term "nominal GDP growth" appears infrequently in the economics literature, its two components — inflation and real growth — underlie much of conventional macroeconomics. Studying the interactions of these factors provides a unique window into the movements of our economy, in both the short and the long term.
Over the long term, real GDP growth has been highly regular, averaging roughly 3% a year for more than a century in the United States. Inflation, however, has been quite variable: Between 1972 and 1981, for instance, nominal GDP growth averaged about 11% due to roughly 8% inflation. That inflation rate was universally deemed to be unacceptably high, and, under Paul Volcker's leadership, the Federal Reserve was able to gradually reduce inflation (and thus NGDP growth). Between 1990 and 2007 — the bulk of the Great Moderation period — the economy generated fairly steady NGDP growth, at just over 5% a year, with inflation at just over 2% and real growth continuing to average about 3%. Thus, over the long term, nominal GDP and inflation have tended to move together while real GDP has been stable.
But in the short term, through the ups and downs of the business cycle, things look quite different. In the short term, nominal and real GDP tend to move together, while inflation is, in the parlance of economists, relatively "sticky" — meaning slow to move. People's choices about spending can change on a dime, but their wages cannot, and neither can prices throughout the economy. That means that a sudden widespread change in spending decisions (which economists call a "nominal shock") will drag real GDP along with it, while inflation lags behind.
So, for instance, between mid-2008 and mid-2009, NGDP fell about 4% (and thus, at -4%, was 9% below its trend rate of 5% growth) as real output fell sharply while inflation remained fairly stable. The same effect can be achieved in a positive direction in the short term — a spike in nominal GDP (achieved, for instance, by monetary or fiscal stimulus) can push real GDP up, though not for long. Almost no one believes that boosting nominal GDP will permanently raise real output: If it could, then a poor country could develop its economy by simply printing more money. Zimbabwe would be the wealthiest country in the world. But for a little while, nominal and real GDP do move together.
These relationships among real GDP, nominal GDP, and inflation yield a standard model of macroeconomics. To recover from a nominal shock, the government can boost nominal spending (or "aggregate demand") through deficit spending or monetary stimulus, and the boost will lead to higher output and, in time, to higher prices. In the long run, however, only the effect on prices will last: Again, it is not possible to make a country permanently richer by printing money. Stimulus is thus short-term medicine, aimed at giving the economy a chance to stabilize and recover from a slowdown.
To avoid nominal shocks in the first place, the government should seek a steady level of nominal GDP growth. It should do this, most economists argue, by using monetary policy to keep output (that is, real GDP) growing steadily and, above all, to keep inflation close to about 2%. Why 2%? Such low inflation is desirable because high inflation rates are believed to hurt the economy. But economists do not believe that inflation is harmful for the reason that most non-economists would cite — the impact of rising prices on consumers. After all, when prices rise, incomes usually rise as well. Instead, many economists would argue that high inflation is a sort of tax on capital: Our tax system punishes savers during high-inflation periods like the 1970s, as people must pay taxes on capital earnings that merely reflect price increases driven by inflation. The result is reduced saving and investment.
So if inflation is harmful, why not aim for a target below 2%, say zero inflation? The country that has come closest to such "price stability" in recent years is Japan, where the Consumer Price Index inflation rate has been at almost zero for more than a decade. Many economists believe this extremely low inflation has hurt Japan in two different ways. First, it is the primary reason that Japan has had almost 15 years of near-zero interest rates on short-term government bonds — a situation often called a "liquidity trap." Such a "trap" makes conventional monetary policy almost impossible: The Bank of Japan is unable to lower rates to spur demand in a downturn. It can resort to some tools other than interest rates, like quantitative easing, but central banks are often reluctant to use such non-conventional policy tools.
The second problem is so-called "wage stickiness." In a competitive market economy, some wages need to fall relative to others. In an economy with mild inflation (which causes all wages to gradually rise), relative wage adjustments can take the form of some workers getting smaller pay increases than others. But if inflation is near zero, it may become necessary to give genuine nominal wage cuts to a significant fraction of workers — and of course workers strongly resist cuts in the current dollar (or yen) amount of their pay.
An inflation rate around 2% is believed to be low enough to avoid most of the downsides of inflation but high enough to avoid the liquidity trap and the challenge of wage stickiness. Inflation at roughly that rate has therefore been considered the ideal target of monetary policy.
Of course, the Federal Reserve cares about more than just inflation. Its "dual mandate" is to provide stable prices and high rates of employment. During the Great Moderation, the Fed carried out this mandate by raising interest rates (tightening policy) when inflation or real output (or both) rose above the Fed's target, and by cutting interest rates when either or both fell below the target — a policy referred to as the "Taylor Rule," after its chief exponent, Stanford's John Taylor. But although this approach takes in elements of both inflation and real GDP, it is still essentially an inflation-targeting mechanism — or what economists call "flexible inflation targeting." Under the Taylor Rule, for instance, the short-term interest rate is adjusted by one and a half times changes in the inflation rate plus half the so-called "GDP gap" (the difference between measured real GDP and the estimated maximum amount the economy could produce while maintaining reasonable price stability). This policy insures relatively stable inflation, and also moderates the business cycle.
So although it pursues a dual mandate, the Federal Reserve views the manipulation of inflation through interest rates as the key role of the central bank. And this approach seemed to work for a long period of time. What, then, went so spectacularly wrong in 2008?
THE CRISIS RECONSIDERED
The Federal Reserve made two crucial and revealing errors in its response to the recession and economic crisis of 2008 and 2009. First, it failed to "target the forecast" — that is, to implement policies that could be expected to actually produce the results the Fed wanted to achieve. This was the consequence not of a failed attempt to use the Fed's power, but rather of an explicit reluctance to use that power at all.
Consider the following analogy. Suppose a passenger is chatting with the captain of a ship that is scheduled to reach New York in two days. The captain mentions that, due to wind and currents, he expects to make landfall in Boston, not New York. The passenger might ask why the captain doesn't adjust the steering, so that the city he expects to reach (the forecast) is the same as the city he had initially hoped to reach (the target). To most people, this would seem like common sense. Thus one might be surprised to learn that, while a few prominent economists (like Lars Svensson of Princeton University) have advocated a policy by which the Federal Reserve would similarly embrace and adjust policies as needed to reach its forecast, most economists take a kind of "wait and see" approach. When the economy drifts off course, the Fed makes an adjustment — but not necessarily one that is expected to lead to on‑target nominal growth. Rather, the Fed waitsand sees how its policy moves affect the economy, and then makes further adjustments as needed, in an effort to avoid overreacting to short-term data.
In late 2008, this incremental approach produced a very weak and ineffectual policy response to a rapid and deep decline in nominal spending — a policy response that was clearly understood by the markets to be grossly inadequate in a fast-changing economic environment. Although we have no futures market for NGDP expectations, the sharp and simultaneous plunge in stock prices, commodity prices, and Treasury-bond yields provides a strong indication that investors foresaw a sharp slowdown in spending.
But this was not the Fed's only, or greatest, error. The central bank's sluggish response to early indications of trouble might help account for the severity of the decline in economic activity in late 2008, but it can't really explain why the economy remains in the doldrums nearly three years later.
Before considering what went wrong, it would be helpful to compare the current recovery to the recovery from the other post-war recession that featured double-digit unemployment: the recession of the early 1980s. That recession was really two related slowdowns — one in the first six months of 1980, and then another in the 16 months from July 1981 to November 1982. Six of the 12 quarters in that period saw negative real GDP growth (including -7.9% in the second quarter of 1980 and -6.4% in the first quarter of 1982, two of the worst quarterly performances since the Great Depression). Unemployment also spiked higher than it had at any time since the Depression, reaching a peak of 10.8% in November 1982. But that recession was followed by an equally strong rebound: During the first six quarters of the "Volcker recovery," starting at the end of 1982, NGDP rose at 11%, RGDP rose at 7.7%, and (therefore) inflation was about 3.3%.
The recession of 2007-09 was of a roughly similar magnitude, featuring five quarters of negative real GDP growth, including an astounding contraction of 8.9% in the fourth quarter of 2008. As noted earlier, unemployment during this downturn peaked at 10.1%, in October 2009. But the recovery from this latest recession has been nowhere near as impressive as that of the early 1980s. During the first six quarters of our current "recovery," NGDP rose at a rate of 4.3%, RGDP rose at 3.0%, and inflation ran at about 1.3%. In the first two quarters of this year, NGDP growth slowed even further, to 3.4%. Our "jobless recovery" is thus no mystery: Unemployment remains high because the economy is not really recovering at all. Real GDP growth is roughly equal to trend, and hence we are not closing the output gap created by the severe decline of 2008 and 2009. We fell into a deep hole, and we started digging sideways.
Also less than mysterious is the reason for that slow real GDP growth: Nominal growth has been much slower in the past few years than during the 1983-84 recovery. It is almost inconceivable that the economy could grow at the 7.7% real rate of 1983-84 with nominal spending growing at only 4.3%, which is below even the trend rate of nominal GDP growth during the Great Moderation. There is an ongoing debate about whether the current problem is a matter of demand (i.e., not enough nominal spending) or whether it is "structural" (i.e., workers' lacking the right skills for the available jobs). But we will never find out until we actually see the sort of nominal spending that would be required to ignite a robust recovery.
This is something that should be kept in mind by those who argue that our problems are "structural." Advocates of the structural view sometimes lose sight of the fact that we do not have enough nominal spending to launch a real recovery even if the economy had no structural problems at all. Whatever our long-term problem, our immediate problem is poor NGDP growth. Just as a sharp drop in NGDP was the proximate cause of the crisis, persistently low NGDP is the proximate cause of our slow recovery. Monetary policy that focused directly on nominal growth rather than viewing it only indirectly through the lens of inflation would identify this problem more clearly, and would thus be better positioned to address it. The Fed's second error, in other words, is that it has been focused on the wrong problem.
Many economists would contest this description of the policy failure, on two different grounds. First, they would argue, prominent macroeconomists did understand that there was a demand shortfall in 2008, but did not think that monetary policy was capable of fixing the problem, especially once short-term interest rates approached zero, leaving the Fed "out of ammunition." This is the "liquidity trap" view.
But the Fed itself never claimed to be "out of ammunition," even after rates hit zero. Indeed, Chairman Ben Bernanke has repeatedly stressed that the Fed still has many options for boosting demand, and he has proved the point with two rounds of "quantitative easing." Indeed, it is hard to see how a fiat-money central bank would ever be left unable to boost nominal spending. That would logically imply it was unable to raise the rate of inflation — that is, to "debase the currency," which it can always do. There is no example in history of any fiat-money central bank that tried to create inflation and failed. In 1933, for instance, Franklin Roosevelt was able to create substantial inflation by devaluing the dollar against gold, despite near-zero interest rates. Some economists (most notably Paul Krugman) cite the example of the Bank of Japan over the past decade and more to suggest that a central bank can try and fail to raise inflation. But the Japanese twice tightened monetary policy in an environment of zero inflation (in 2000 and 2006), so it would be hard to claim that they were trying to create inflation.
Indeed, by just hinting at quantitative easing in a series of speeches last year, Fed officials were able to drive up stock prices, elevate inflation expectations, and lift commodity prices. After the past few years, there can no longer be any serious disagreement about whether monetary policy can be effective once rates hit zero. Clearly, it can be. The only questions are how the Fed understands its goals and how aggressive it is willing to be in pursuing them.
Second, some critics might argue that the policy of inflation targeting should have been able to address the demand shortfall in 2008-'09 yet failed to do so, and that it is therefore not clear why nominal GDP targeting would have done any better. This of course raises the question of why inflation targeting itself did not succeed. A substantial decline in aggregate demand will tend to reduce both inflation and output to below the central bank's targets — so if the Fed uses a flexible inflation target like the Taylor Rule, such reductions should trigger monetary stimulus, which would eventually help put the economy back on track. We have seen above that this policy was not sufficiently forward-looking in late 2008, and that the Fed was too slow to react and not aggressive enough in targeting its goals. But what is the problem today? Why is inflation targeting not producing a more aggressive Fed policy now? And is there reason to think that NGDP targeting would do any better?
THE LIMITS OF INFLATION TARGETING
The Federal Reserve's continuing failure to address our economy's woes points toward four major problems with inflation targeting itself.
The first problem is well understood by macroeconomists: The optimal policy response to inflation depends on whether price increases are caused by greater aggregate demand (that is, an increase in nominal spending) or by an adverse supply shock (such as a suddenly reduced supply of food or oil due to a natural disaster or war). Policymakers generally try to "accommodate" temporary changes in food and oil prices, as these don't reflect changes in aggregate demand and so generally don't result in persistent changes in the rate of inflation. For instance, inflation in mid-2008 was quite high due to rising oil prices, but by early 2009 the CPI was actually falling for the first time in more than half a century. The Fed understands this distinction and generally directs its attention to "core inflation," which excludes food and energy prices.
By itself, however, the problem of supply shocks should not result in the sort of policy failures that we have witnessed since 2008. After all, during 2009 and 2010, real output and core inflation were both below the Fed's implicit targets, yet the central bank still did not act aggressively enough to address the problem. The Fed's wait-and-see approach was part of the reason, but the remaining flaws of inflation targeting — which are frequently overlooked by macroeconomists — were crucial factors.
The second major problem with inflation targeting has to do with how the Consumer Price Index, which is the standard measure of inflation, is calculated — a problem brought into stark relief by the housing crisis of the past few years. As of mid-2009, the rate of NGDP growth over the previous 12 months was about -4%, as noted above. In contrast, core inflation was running about 1.5%, only slightly below the Fed's (implicit) 2% target. In fact, between mid-2008 and mid-2009, the housing component of the CPI rose even faster than the overall index. That's right: During the greatest housing-price crash in American history, government data showed the cost of housing rising, even relative to other goods.
This is largely because the government relies on a flawed "rental equivalent" estimate for housing costs, which in turn distorts the entire CPI measure. Until 1983, the Bureau of Labor Statistics measured housing prices based on direct ownership costs like home purchase values, mortgage-interest rates, and property taxes. But because interest rates and housing values were changing rapidly, the BLS became concerned that this measure was providing an inaccurate measure of inflation — making inflation seem more jerky and uneven than it was. And because a home is a long-term investment as well as a consumer good, the agency also worried that it was giving too much weight to considerations tied to the investment component — factors that did not relate to the immediate state of prices in the real economy. Thus it sought to separate the investment component of housing from the consumption component through a calculation of the rental value of homes; it has since measured the prices of homes based on what the cost of renting them would be. This has meant that the CPI and home values have grown increasingly disconnected. And because housing inflation accounts for 39% of the core CPI, the official inflation rate fell much less during the housing crash than the actual decline of prices in the economy — because the cost of renting homes stayed relatively flat, even as their resale values plummeted. So the Fed did have its eye on inflation, but was receiving faulty inflation signals. During a period of rapidly declining aggregate demand, NGDP would have provided far more timely and accurate data on the need for monetary stimulus than price indices composed mostly of sticky prices.
By mid-2010, however, it was clear that the recovery was faltering. Even the inflation indicator was providing a relatively clear signal of the need for more stimulus, as the core CPI inflation fell to 0.6%, far below the Fed's implicit target. Beginning in late August 2010, Fed officials started sending signals that additional stimulus was coming, and in early November the Fed announced a new round of quantitative easing (which came to be known as QE2). But even before the announcement, the prospect of QE2 generated a great deal of controversy. Many people seemed bewildered by news stories that Ben Bernanke thought inflation was too low and was trying to engineer an increase in the "cost of living." The typical American tends to see inflation as a problem, not a solution, so the policy became a lightning rod for criticism — and pressure to end it was growing before QE2 had even begun.
This episode exposed the third fatal flaw in inflation targeting: the politics of inflation. In principle, an inflation target means that, while the Fed will sometimes need to restrain inflation to keep it under control, the rest of the time it will need to boost inflation to keep it from falling below the target. Indeed, if its errors are "unbiased," the Fed should spend about half of its time trying to increase the inflation rate. But clearly the American public has not bought into this view of the matter. Americans view inflation as a process that reduces their living standards, because they take their own nominal incomes as a given when thinking about the impact of higher prices. Inflation is thus highly unpopular.
Yet according to standard macroeconomics, only supply-side inflation (caused by shortages or other supply shocks) reduces living standards; inflation generated by the Fed actually raises living standards. Consider an example: As noted above, NGDP has been rising at about 4.3% during this slow recovery. If the Fed were able to boost NGDP growth to, say, 6.3%, then part of the extra nominal spending would show up as inflation and part would show up as higher real GDP, or real output — since stimulus can't move inflation as quickly as it can move output. Therefore, with real output growing, more people would have jobs and those already working would be able to work more hours. Income from capital would also increase.
Now suppose that, instead of saying inflation was too low in mid-2010, Bernanke had announced that we needed to boost the incomes of Americans in order to have a healthy recovery — and that the Fed would therefore try to boost the growth rate of national income from 4.3% to 6.3%. This message would have sounded much more appealing to the average voter than a call for higher inflation.
One might object that there is something disreputable about the Fed's trying to deceive the public by working to engineer higher inflation under the guise of "more income." A call for boosting income, however, would have been not only more popular but also more truthful. Ben Bernanke does not really want higher inflation; it would be more accurate to say that he wants more aggregate demand and expects such higher demand to result in somewhat higher inflation rates. Unfortunately, however, the Fed has chosen a language to communicate its intentions that is both deeply unpopular and profoundly misleading. Inflation targeting gives the public the wrong impression, and the resulting political reaction impedes the Fed's ability to carry out its work.
If we stopped talking about inflation targeting and started talking about NGDP targeting, we could greatly simplify the policy debate. Do we want more demand, or not? Most Americans surely think that more demand would be a good thing right now, but very few people want to see more inflation. To the Federal Reserve, these two effects are simply two sides of the same coin. But because the Fed expresses its aims in terms of inflation, its work is understood as a matter of managing inflation, and therefore Fed policies aimed at boosting inflation are politically problematic.
This dynamic contributes to the fourth flaw in inflation targeting — that it creates the strong impression that the job of the central bank is to control inflation while the job of fiscal stimulus is to produce economic growth and more employment. This view is a function of economic ignorance, but it is encouraged by the way the Fed defines its role, and it contributes to profound confusion about economic policy.
Almost any article on monetary stimulus will cite someone discussing the policy's impact on inflation. And almost any discussion of fiscal stimulus will include someone weighing the benefits of growth against the problems of deficits. There is a popular misconception that we should talk about fiscal and monetary stimulus using completely different terms and concepts, but from an economic perspective, there is no justification for the divide.
The standard economic model says that both fiscal and monetary stimulus boost aggregate demand (nominal spending). Whether that shows up in the form of higher inflation or more real growth depends on supply-side factors — like whether there are large numbers of unemployed workers or factories sitting idle that might respond to greater demand, or whether there is little such "slack" (such that new demand would lead mostly to higher wages and prices). If there is a difference between the two policy tools, it is that fiscal stimulus would actually be the more inflationary policy, because it consists of government spending (which is often less efficient, and hence may reduce aggregate supply and create shortages that boost prices), while monetary stimulus enables more private-sector spending. Indeed, free-market economists should be especially partial to monetary stimulus on these grounds.
Recent events in Britain provide a perfect example of the confusion generated by drawing this sort of false dichotomy between monetary and fiscal policy. The government of Prime Minister David Cameronhas been sharply criticized for its policy of fiscal austerity. The recovery from the recent recession has been even weaker in Britain than in the United States, and there are fears that budget cuts will lead to a double-dip recession. At the same time, the press has been highly critical ofthe Bank of England for allowing inflation to rise far above the 2% target. But these criticisms cannot both be correct: Either Britain needs more aggregate demand or it does not. If it needs more, then the inflation rate in Britain needs to rise even higher, because the Bank of England needsto provide even more monetary stimulus. If inflation is too high and Britain needs less aggregate demand, then Britons should desire fiscal austerity that would slow the economy. The press seems to believe in some sortof policy magic whereby fiscal stimulus can create growth without inflation and monetary tightening can reduce inflation without affecting growth.
The way to clear up this muddle is to stop talking about the effect of monetary stimulus on inflation and the effect of fiscal stimulus on growth, and to start talking about how each affects nominal spending. Then we can focus on the fundamental issue: Does Britain need more NGDP growth or not? Most observers would agree that Britain needs both more NGDP growth (because its recovery from the last recession has been anemic) and less government spending (because its deficits and debt threaten its credit and potential for growth). The only way to achieve both goals is to combine monetary stimulus with fiscal austerity. And that can occur only if the Bank of England is free to focus on NGDP growth, rather than on inflation.
It is likely that Britain faces very unpleasant short-term tradeoffs, regardless of what happens with nominal spending. Monetary policy cannot solve real structural problems, and massive growth in government over the past decade has hurt the supply side of the British economy. Britain is likely to end up with an unfavorable inflation-to-output split regardless of what monetary policy it chooses — but at least NGDP targeting can provide a stable spending environment in which the Cameron government can pursue fiscal reforms. As those reforms reduce government spending, private spending will need to increase. Under this approach, the key is to remember that, if the Cameron government tries to shrink the state at the same time the Bank of England is trying to reduce aggregate demand, the British may end up with a double-dip recession that discredits the policy of fiscal austerity.
Indeed, monetary policy has always been the Achilles heel of conservative economics. Conservatives tend to favor low inflation, which is generally a laudable goal. But as we saw in America during the 1930s and in Argentina during the late 1990s, tight money can discredit free-market policies and open the door to left-wing governments if it is disconnected from an understanding of the proper aims of monetary and fiscal policy. There are times when higher inflation is necessary, and if it does not materialize, the public may choose an explosive growth of government instead. To discern such circumstances, we need to look not at inflation but at nominal spending. NGDP targeting is a way of making the world safe for laissez-faire capitalism. It becomes much easier to say we will not bail out General Motors if we have a monetary policy that assures that the failure of GM will not reduce aggregate spending, but will instead result in resources being re-allocated to other parts of the economy. It is easier to shed jobs in declining sectors if jobs are being created just as rapidly in booming sectors.
The knee-jerk opposition of many conservatives to monetary stimulus — and to inflation under any circumstances — arises from the confusion created by inflation targeting. It is a risky attitude, and one especially ill-suited to this moment. Explicitly targeting nominal GDP rather than inflation would greatly alleviate that confusion, and help make for a political environment friendlier to a more effective monetary policy.
FINDING THE TARGET
Given these theoretical advantages of nominal GDP targeting, how ought it to be implemented in practice?
Most simply, the Federal Reserve should begin by adopting an approach of "level targeting" of nominal GDP. This doesn't mean keeping NGDP level, but rather targeting a specified trajectory, such as a 5% NGDP growth path, and committing to make up for any near-term shortfalls or excesses. Thus, if NGDP grew by 4% one year, the central bank would cut rates or engage in quantitative easing until its models yielded an expectation of 6% NGDP growth for the following year.
Level targeting is especially important when the economy is beset by severe shocks, as in late 2008. If markets had understood that any near-term shortfalls in NGDP growth would be made up for over the following several years, then asset prices (stocks, real estate, etc.) would have declined much less sharply, which would have moderated the decline in late 2008. Promising stable NGDP growth in later years along a pre-determined trajectory would actually make near-term NGDP much less unstable.
Another approach — which would be more radical, but perhaps also more effective — would limit the Fed's role to setting the NGDP target, and would leave the markets to determine the money supply and interest rates. This would mitigate the "central planning" aspect of the Federal Reserve's current role, which has rightly come under criticism from many conservatives. To give a simplified overview, the Fed would create NGDP futures contracts and peg them at a price that would rise at 5% per year. If investors expected NGDP growth above 5%, they would buy these contracts from the Fed. This would be an "open market sale," which would automatically tighten the money supply and raise interest rates. The Fed's role would be passive, merely offering to buy or sell the contracts at the specified target price, and settling the contracts a year later. Market participants would buy and sell these contracts until they no longer saw profit opportunities, i.e., until the money supply and interest rates adjusted to the point where NGDP was expected by the market to grow at the target rate.
It might be helpful to compare this idea to the old international gold standard. Under that system, the U.S. government agreed to buy and sell unlimited gold at $20.67 per ounce. This kept gold prices stable, and the money supply adjusted automatically. Unfortunately, however, stable gold prices did not always mean a stable macroeconomic environment. Putting NGDP futures contracts on the market along a similar model would likewise create a stable price for those contracts, hence stabilizing expected NGDP growth. And stable NGDP growth would be more conducive to macroeconomic stability than a stable price of gold, especially in a world in which rapidly growing demand from Asia might distort the relative price of gold.
Some critics might argue that NGDP targeting would be too easy on inflation. And it is true that looking at nominal GDP targeting through the prism of the past few years might make such targeting seem less "hawkish" than inflation targeting — a backdoor method of allowing excessive inflation. But the argument cuts both ways. Nominal GDP targeting would produce lower than average inflation during a productivity boom. Indeed, one criticism of inflation targeting is that, because central banks focus on consumer prices, they allow asset bubbles to form, which eventually destabilizes the economy. Nominal GDP targeting cannot completely eliminate this problem, but it can impose more monetary restraint during periods of high-output growth than can inflation targeting.
Others might acknowledge that NGDP targeting could help stabilize output, but would nonetheless point to a serious cost: greater inflation variability. In fact, however, many of the problems generally associated with inflation are actually linked to NGDP volatility. For instance, inflation is said to raise the effective tax rate on capital, as most tax systems don't index taxes on interest and dividends to inflation. But the nominal interest rate may be more closely correlated with NGDP growth than with inflation, meaning that the tax distortion is better explained by high NGDP growth than by high inflation. Or, to take another example, deflation is often blamed for high unemployment — but, once again, the problem is actually caused by falling NGDP growth, as lower inflation resulting from productivity gains does not create unemployment. Low inflation, meanwhile, is often thought to make a liquidity trap more likely; in fact, it is low NGDP growth that best accounts for the risk of hitting the zero-rate bound. Interest rates fell to zero when Japan experienced mild deflation, but not when China experienced mild deflation. The difference was NGDP growth: It has been near zero for Japan since 1993, while Chinese nominal growth has not fallen below the 5% to 10% range, even during the East Asian crisis of the late 1990s.
Moreover, George Selgin has shown that, contrary to conventional wisdom, unexpected inflation is not unfair to lenders as long as NGDP growth is on target. If inflation were to rise sharply during a period of stable NGDP growth, it would mean that a real shock had depressed the economy. When real shocks occur, it is only fair that debtors and creditors share the loss. Suppose lenders made lots of foolish loans to the housing sector, leading to a subsequent fall in real GDP as housing construction plummeted and workers had to be retrained for other sectors. In that case, NGDP targeting would lead to a period of above-normal inflation, and lenders would bear some of the burden for this misallocation of capital, even in the absence of outright defaults. That would be appropriate.
So the idea that NGDP targeting might allow for higher inflation and thus fail to prevent the problems caused by inflation is mistaken. In fact, many of the problems we associate with inflation are actually caused by fluctuations of nominal GDP, and would be better addressed by a policy aimed at keeping NGDP at some pre-determined target.
POLICY AND POLITICS
In the end, there is no escaping the fact that the Federal Reserve is a political institution. For most of the 20th century, the politicization of monetary policy meant easier money and thus more inflation. In the current policy environment, the politicization of the Fed means tighter money, and so lower inflation; for all the complaints about easy money, especially on the right, we have a policy that is actually too tight to maintain even 2% core inflation. Either way, though, the inevitable politicization of monetary policy has combined with the Fed's overly narrow focus on inflation to produce monetary policy that is not well suited to produce the kind of consistent growth our economy needs.
Nominal GDP targeting is, of course, not the only possible solution to the problems bedeviling monetary policy today. But the solution that offers the most economically plausible alternative — a higher inflation target, between 3% and 4% — is not politically viable. Nominal GDP targeting, on the other hand, offers both a politically and economically sensible alternative, which would be far better equipped to advance stable growth, to overcome the politics of inflation, and to help the Fed avoid discrediting itself. It offers a single target that effectively combines both facets of the Fed's dual mandate, and so should be attractive to those on both the left and the right who argue that the requirement to simultaneously address inflation and unemployment makes it impossible for the Fed to tackle either very well.
This does not mean, of course, that NGDP targeting can address our broader economic predicament. Monetary policy is not an answer to structural policy problems. It cannot make up for a failure to control the growth of public spending or the explosion of entitlement costs. But it can do a great deal to help create a stable environment of consistent growth, and thus make the difficult structural reforms ahead both more effective and easier to tolerate. At the moment, our monetary policy risks making the hard task facing fiscal reformers all the more challenging. A Federal Reserve with its eyes firmly set on the right target would greatly ease their way.