Debating Milton Friedman's Legacy For Monetary Policy [View article]
LJK writes: "But all the Fed can do is lead the borrowers to water. It cannot make them drink." And yes, of course the Fed can't meter out inflation in the way your oil company can, if asked, put 371.6 gallons in your tank. These are valid points, but they are an argument for NGDP targeting, not against it.
As I wrote in the above-referenced post on Friedman and NGDP,
"Friedman’s second argument for targeting money growth in a world of variable velocity, and the more important one, is that the Fed should not in any event attempt to stabilize either the P or Q variable individually, or their product, in the short run. He viewed the transmission mechanism from money to the PQ side of the equation of exchange as being subject to long, variable, and unpredictable lags. Trying to fine-tune monetary policy in pursuit of an inflation or real GDP target in the face of those lags would be more likely to destabilize the economy than to stabilize it.
The bottom-line argument for monetarism was always that in a world subject to unpredictable lags and shocks, the best a central bank can do is to provide a simple, transparent, and predictable framework within which market participants can make decisions." http://bit.ly/yZ0sO0
Latest Data Suggest Output Gap Is Closing, But Employment Gap Is Closing Faster [View article]
You are right, there are many measures of the output gap and many interpretations. I was trying to be rather cautious in suggesting one of several that are possible. However, we have to face the possibility that some of the jobs that people held before the recession may not return in the same form or at the same salaries. One possible reason is that there has been a reassessment of the value that is really added by those jobs.
As examples, I would give autoworker jobs, which may come back, but not at the former salary scale. US autoworkers aren't really enough more productive than their competitors elsewhere to justify the old UAW salaries. As another example, there were people working in construction who seemed to be adding value by building houses, but when it turned out the houses were overvalued, the value of that labor dropped, too.
I do agree that some construction workers now bagging groceries may eventually go back to better jobs and some people working part-time involuntarily will eventually find full-time employment. That does not mean, though, that we will necessarily get back to the pre-recession real GDP trend line.
Updated Seasonal Factors Remove Much Volatility From 2011 Monthly CPI Data [View article]
I suppose if you say "it will snow tomorrow" every day, all through July and August when there is no chance at all it will snow, and then one day in December it snows, you will finally be able to say "I told you so" and feel a lot smarter than all those meteorologists.
jhooper: Good question. By law, any central bank (Fed, ECB) is allowed to accept only sound collateral. The problem is, politics now decrees that the ECB has to continue pretending that all euro area sovereign debt is still "sound" even though everyone know it may not be.
"Of course, if those assets have dwindled due to defaults, then the ECB couldn't fully reverse its reserve injections, and the Eurozone could end up with a huge surplus of money."
Of course? I'm not so sure about that. I'm not an expert on the ECB balance sheet or EU banking law, but my understanding is that most ECB assets are term loans to banks with bonds accepted as collateral. When the loans come due, the banks have to repay them. I don't think (please, someone, correct me if I am wrong on this) they have the right, as a typical US homeowner does, just to walk away from the loan because it is "under water" due to a fall in the market value of the collateral. If that is true, then no matter how little the collateral is worth when the loans come due, the banks repay them and the associated monetary liabilities of the ECB are extinguished.
Of course, then the banks would be in trouble and the ECB or some other entity might have to bail them out, but that is a different issue.
On the broader issue of whether the state of the euro is "catastrophic" or not, I recommend this view of the Italian situation from Ed Hugh: http://tiny.cc/kvdcr
Ok, never let down your inflation guard. But maybe time to pivot and raise your deflation guard? As I read the latest data, the US economy faces more of a danger on the deflation side. http://tiny.cc/0bf8w
Profiting From The Euro's Collapse And Looming Inflation [View article]
It is amazing that inflation fears continue. The latest BLS data, if anyone bothers to read them, show that deflation is the more serious threat http://tiny.cc/0bf8w
The Solution For The Eurozone Debt Crisis Is Actually Quite Simple [View article]
There are two major flaws with the idea that the market can enforce fiscal discipline in a currency area like the euro
First, it ignores the time-inconsistency issue. Politicians have a shorter time horizon than the economy. As long as fiscal profligacy before elections gets votes and the pain (higher interest rates) comes after the election, market discipline will be inadequate
Second, it ignores the free-rider problem that is inherent in currency areas. A country tends to reap almost all of the benefits of fiscal profligacy itself, while at least part of the pain (higher interest rates, higher inflation, whatever) is spread to currency area partners. That also means that market discipline is inadequate.
The combined effect of time-inconsistency and free-ridership produces a perverse political dynamic that can only be countered by strong policy rules.
Whether the latest attempt at rules announced yesterday by Sarkozy and Merkel will work is another question. Here are some details to watch for that will determine whether they work or not: http://tiny.cc/g07t7
(1) At best, this could "save" the euro but does not address, let alone answer, the question of whether a currency area is suitable for the 17 countries that now share the euro. At this point the burden of proof lies with anyone who argues that a currency union among the 17 is a good idea.
(2) The plan makes it difficult for member countries to issue new debt. Many of them now have large deficits and large output gaps. Those countries would be forced to engage in further austerity measures in already depressed economies. That is not a hopeful strategy. Furthermore, this is not a one-time situation. The whole structure of the new rules would make it difficult to pursue an optimal fiscal policy over the business cycle--a policy of, say, the Swedish kind. ( See here for a description of how Swedish fiscal policy works http://tiny.cc/5nsrj )
(3) As other comments have pointed out, it is not feasible either in the peripheral countries (because of the combination of austerity with loss of sovereignty), or in Germany, Finland, Slovakia, etc. (because of questionable cost-benefit calculation from their national points of view).
Why Is The Greek Tail Wagging The Global Dog? [View article]
"Governments to big to be supported by their economies?" Weak diagnosis. Government of Greece is not especially big by European standards, compared with healthy Northern European countries. US government is not especially big, either historicially or in comparison to other countries, measured by any reasonable yardstick (like cyclically adjusted total government spending). Temporary spike in Federal government spending as percent of GDP during the current strong downturn is a misleading indicator of trends.
The real problem, in my view, is governments all over the world that can't make the decisions they obviously need to make. Anyone who listened to Simpson, Bowles, and Rivlin before the Supercommittee yesterday knows what I mean.
U.S. GDP Growth Stronger In Q3 As Economy Enters Expansion [View article]
Yes, thanks for the long dialog. I should say you have made many good points, and I too, in the past, have had good things to say about inflation targeting. It did not become the world's most widespread central bank policy regime without having some merits. However, in my view, it has weaknesses, too, and those have become more evident in recent years. It will be interesting to see how things evolve.
I read that the Fed staff is putting NGDP targeting on the table as one option in a position paper to the board this month. I do not expect outright adoption of the policy, but it will be interesting to see what kind of comments it draws, after support from such diverse corners of the profession as Goldman Sachs and Christina Romer.
U.S. GDP Growth Stronger In Q3 As Economy Enters Expansion [View article]
You ask several questions that have a fundamentally microeconomic focus--how companies react to changes in relative prices of various commodities, how decisions are influenced by leverage and regulations, how personal work effort might be influenced by the price of energy, and so on. No one would deny that on a day to day basis, economic decisions at all levels are dominated by such considerations.
Monetary policy has much more general and modest aims. It does not expect to determine micro decisions. Instead, it only hopes to provide a relatively stable and transparent macroeconomic background against which micro decisions can be made. The argument in favor of NGDP targeting comes down to the claim that it provides a simpler and more transparent framework. It does not and cannot protect firms and households from external shocks; it only strives not to react in ways that makes matters worse. It claims to be more predictable and transparent because it requires fewer ad-hoc, situational adjustments as unexpected events occur.
You are asking too much of policy. Be willing to let markets for tires, hamburgers, lawn services, etc. do their thing, with monetary policy in the background.
U.S. GDP Growth Stronger In Q3 As Economy Enters Expansion [View article]
Good questions. Let me give a try.
(1) Yes, quality improvements pose problems. People who measure GDP try to correct for them by adjusting the numbers to reflect quality. They don't succeed entirely. The result is that the published GDP numbers are generally thought to understate real GDP growth and therefore to overstate the rate of inflation. That is another another example (better than martinis) of why it is easier to measure one variable (NGDP) than to try to figure out how that is broken down into its two components (price and real output).
Quality mismeasurement can create a problem for inflation targeting. If the reported real GDP numbers overstate inflation (because they understate real growth in the form of quality improvement), the Fed might unnecessarily tighten policy and slow real growth to fight an illusory inflation problem.
(2) "But Fed policy must be anticipatory. " This is a key point, for sure. Because of lags in data and in the effect of policy actions, the Fed always has to anticipate. They are like the captain of an oil tanker, whose ship takes a long time to turn or stop, navigating in the fog without radar. But unlike the tanker captain, they don't have an accurate chart, just forecasts that are always subject to error. So how to proceed?
The standard approach is to target the inflation forecast. The Fed's economists put together the best forecasting model they have and they say "we think 2012 inflation will be 2.6 percent if you do nothing. We think you can bring it down to 2% (assume that is the target) if you raise the Fed funds rate by 75 basis points." The Fed raises rates, but, since the inflation forecast is only accurate to within, say, + or - 2, by the time 2012 comes around, inflation might actually be, maybe, 2.7% despite the attempt to hit the 2% target. Then they have to tighten more. So it is not really true that "if I just target inflation I can hit that target with 70% certainty." Maybe you can hit it + or - 2% with 70% certainty, but that's about all.
What about NGDP targeting? One approach to NGDP targeting would also be to base your policy on the same kind of forecasting model. Of course, your forecast still might be wrong and NGDP growth might come in at 4 or 6% even if you targeted 5%, so you would have to correct.
Same problem as with inflation targeting, only you are better off in one important respect. You can miss your inflation target either because of a shock to the demand side of the economy, or a shock to the supply side. In the case of NGDP targeting, you only have to worry about the demand side shocks. Supply side shocks won't affect total NGDP, just the way they are split between inflation and real growth. So the size and frequency of misses should be smaller with an NGDP target than an inflation target, and as a consequence, the need for after-the-fact corrections would be less frequent.
There is another way to work the NGDP target that is potentially more accurate. That method involves establishing an NGDP futures market to help your forecasting. It is a little complex, but there is a good, nontechnical discussion here: http://bit.ly/sJJtBy
(3) "As long as inflation is low, why should the Fed EVER be trying to put a cap on GDP growth?" There is one important circumstance where they should do so. Suppose the economy is in a phase where productivity is increasing rapidly. Other things being equal, faster productivity growth would cause lower inflation, or even mind deflation. There is a danger that the Fed will push too hard to keep inflation up to its target by keeping interest rates too low.
When that happens, we get the strange case where the economy begins to overheat (in the sense that real output grows above its long run potential and labor markets become very tight), but there is little or no price inflation because productivity gains are holding prices down. The combination of low inflation, overly expansionary monetary policy and abnormally low interest rates leads to asset price bubbles, like the stock market bubble that burst in 1929, the dot.com bubble, and the subprime housing bubble. I think I mentioned earlier that Hayek pointed to this problem in his debates with Keynes in the 1930s.
There is a good short summary of some of these points in this post that discusses Goldman's recent endorsement of NGDP targeting: http://bit.ly/t4hkfB/
U.S. Inflation Data: Scant Fuel For Inflation Fears In September CPI Report [View article]
When productivity increases you keep the same NGDP target, lower the price, and produce more stuff. E.g., before productivity gain, you expect to produce 100 widgets at $10 each. After productivity gain, you produce 110 widgets at $91 each, or 100 at $91 and spend the remaining $90 on something else. So no, you don't have to raise prices to hit the target, instead quantity goes up.
With regard to NGDP, it is one variable, not two. The whole point is that the Fed doesn't worry in advance how it is split up. If you want another lame example, it is like dividing up a quart of martini's among 8 guests. The host can easily control the amount of liquid he mixes, and then pour each guest 4 ounces of martini. It would be harder, not easier, to control how many inches of martini to pour in each glass, because some guests might have short wide glasses and some tall skinny glasses. Inflation targeting is more complex because it would be like trying to measure the width of each different shaped glass and then figuring out how many inches to pour in each glass. Think height of glass=inflation, width=real output, volume of liquid = NGDP. The volume is the basic variable you focus on, let the others come out how ever they do.
Let's switch this discussion to today's post if you want to continue it.
U.S. Inflation Data: Scant Fuel For Inflation Fears In September CPI Report [View article]
If we get an increase in productivity growth or new employment with constant NGDP growth, then we let the rate of inflation go down. It is OK even to go into negative inflation under these circumstances. Productivity-driven deflation is not likely to trigger the Japanese-style downward spiral; that comes from demand driven deflation.
The superiority of NGDP targeting (or similar) is an old point. It was one of the main points made by Hayek in his critique of Keynes in the 1930s. With inflation targeting, productivity growth requires increasing the rate of NGDP growth to keep the rate of inflation constant. That risks causing bubbles, as in dot.com, subprime mortgages, etc.
You have hit on one of the most important arguments in favor of NGDP targeting.
"It's already clear that this Fed has caused inflation without any increase to income or employment." Not clear to me. Say, we now have 3.5 percent inflation and 1% RGDP growth = 4.5% NGDP growth. How can you be sure that if the Fed had slowed NGDP growth to 3%, we would now have 2% inflation and real growth? The evidence is to the contrary; part of the slowdown in NGDP would most likely have come out of RGDP and part out of inflation. For example (I'm just using these numbers as illustrations) we might have 0.5% real growth and 2.5% inflation, or 0 growth and 3% inflation. So it is not "clear." you have to look at the data. The data say that usually changes in NGDP growth induce (or reflect) changes in both P and RGDP components..
Debating Milton Friedman's Legacy For Monetary Policy [View article]
As I wrote in the above-referenced post on Friedman and NGDP,
"Friedman’s second argument for targeting money growth in a world of variable velocity, and the more important one, is that the Fed should not in any event attempt to stabilize either the P or Q variable individually, or their product, in the short run. He viewed the transmission mechanism from money to the PQ side of the equation of exchange as being subject to long, variable, and unpredictable lags. Trying to fine-tune monetary policy in pursuit of an inflation or real GDP target in the face of those lags would be more likely to destabilize the economy than to stabilize it.
The bottom-line argument for monetarism was always that in a world subject to unpredictable lags and shocks, the best a central bank can do is to provide a simple, transparent, and predictable framework within which market participants can make decisions." http://bit.ly/yZ0sO0
Latest Data Suggest Output Gap Is Closing, But Employment Gap Is Closing Faster [View article]
As examples, I would give autoworker jobs, which may come back, but not at the former salary scale. US autoworkers aren't really enough more productive than their competitors elsewhere to justify the old UAW salaries. As another example, there were people working in construction who seemed to be adding value by building houses, but when it turned out the houses were overvalued, the value of that labor dropped, too.
I do agree that some construction workers now bagging groceries may eventually go back to better jobs and some people working part-time involuntarily will eventually find full-time employment. That does not mean, though, that we will necessarily get back to the pre-recession real GDP trend line.
Updated Seasonal Factors Remove Much Volatility From 2011 Monthly CPI Data [View article]
PIIGS Update: Not Catastrophic [View article]
PIIGS Update: Not Catastrophic [View article]
Of course? I'm not so sure about that. I'm not an expert on the ECB balance sheet or EU banking law, but my understanding is that most ECB assets are term loans to banks with bonds accepted as collateral. When the loans come due, the banks have to repay them. I don't think (please, someone, correct me if I am wrong on this) they have the right, as a typical US homeowner does, just to walk away from the loan because it is "under water" due to a fall in the market value of the collateral. If that is true, then no matter how little the collateral is worth when the loans come due, the banks repay them and the associated monetary liabilities of the ECB are extinguished.
Of course, then the banks would be in trouble and the ECB or some other entity might have to bail them out, but that is a different issue.
On the broader issue of whether the state of the euro is "catastrophic" or not, I recommend this view of the Italian situation from Ed Hugh: http://tiny.cc/kvdcr
A CPI Christmas Present [View article]
Profiting From The Euro's Collapse And Looming Inflation [View article]
The Solution For The Eurozone Debt Crisis Is Actually Quite Simple [View article]
First, it ignores the time-inconsistency issue. Politicians have a shorter time horizon than the economy. As long as fiscal profligacy before elections gets votes and the pain (higher interest rates) comes after the election, market discipline will be inadequate
Second, it ignores the free-rider problem that is inherent in currency areas. A country tends to reap almost all of the benefits of fiscal profligacy itself, while at least part of the pain (higher interest rates, higher inflation, whatever) is spread to currency area partners. That also means that market discipline is inadequate.
The combined effect of time-inconsistency and free-ridership produces a perverse political dynamic that can only be countered by strong policy rules.
Whether the latest attempt at rules announced yesterday by Sarkozy and Merkel will work is another question. Here are some details to watch for that will determine whether they work or not: http://tiny.cc/g07t7
What To Do With Europe [View article]
(1) At best, this could "save" the euro but does not address, let alone answer, the question of whether a currency area is suitable for the 17 countries that now share the euro. At this point the burden of proof lies with anyone who argues that a currency union among the 17 is a good idea.
(2) The plan makes it difficult for member countries to issue new debt. Many of them now have large deficits and large output gaps. Those countries would be forced to engage in further austerity measures in already depressed economies. That is not a hopeful strategy. Furthermore, this is not a one-time situation. The whole structure of the new rules would make it difficult to pursue an optimal fiscal policy over the business cycle--a policy of, say, the Swedish kind. ( See here for a description of how Swedish fiscal policy works http://tiny.cc/5nsrj )
(3) As other comments have pointed out, it is not feasible either in the peripheral countries (because of the combination of austerity with loss of sovereignty), or in Germany, Finland, Slovakia, etc. (because of questionable cost-benefit calculation from their national points of view).
Why Is The Greek Tail Wagging The Global Dog? [View article]
The real problem, in my view, is governments all over the world that can't make the decisions they obviously need to make. Anyone who listened to Simpson, Bowles, and Rivlin before the Supercommittee yesterday knows what I mean.
U.S. GDP Growth Stronger In Q3 As Economy Enters Expansion [View article]
I read that the Fed staff is putting NGDP targeting on the table as one option in a position paper to the board this month. I do not expect outright adoption of the policy, but it will be interesting to see what kind of comments it draws, after support from such diverse corners of the profession as Goldman Sachs and Christina Romer.
U.S. GDP Growth Stronger In Q3 As Economy Enters Expansion [View article]
Monetary policy has much more general and modest aims. It does not expect to determine micro decisions. Instead, it only hopes to provide a relatively stable and transparent macroeconomic background against which micro decisions can be made. The argument in favor of NGDP targeting comes down to the claim that it provides a simpler and more transparent framework. It does not and cannot protect firms and households from external shocks; it only strives not to react in ways that makes matters worse. It claims to be more predictable and transparent because it requires fewer ad-hoc, situational adjustments as unexpected events occur.
You are asking too much of policy. Be willing to let markets for tires, hamburgers, lawn services, etc. do their thing, with monetary policy in the background.
U.S. GDP Growth Stronger In Q3 As Economy Enters Expansion [View article]
(1) Yes, quality improvements pose problems. People who measure GDP try to correct for them by adjusting the numbers to reflect quality. They don't succeed entirely. The result is that the published GDP numbers are generally thought to understate real GDP growth and therefore to overstate the rate of inflation. That is another another example (better than martinis) of why it is easier to measure one variable (NGDP) than to try to figure out how that is broken down into its two components (price and real output).
Quality mismeasurement can create a problem for inflation targeting. If the reported real GDP numbers overstate inflation (because they understate real growth in the form of quality improvement), the Fed might unnecessarily tighten policy and slow real growth to fight an illusory inflation problem.
(2) "But Fed policy must be anticipatory. " This is a key point, for sure. Because of lags in data and in the effect of policy actions, the Fed always has to anticipate. They are like the captain of an oil tanker, whose ship takes a long time to turn or stop, navigating in the fog without radar. But unlike the tanker captain, they don't have an accurate chart, just forecasts that are always subject to error. So how to proceed?
The standard approach is to target the inflation forecast. The Fed's economists put together the best forecasting model they have and they say "we think 2012 inflation will be 2.6 percent if you do nothing. We think you can bring it down to 2% (assume that is the target) if you raise the Fed funds rate by 75 basis points." The Fed raises rates, but, since the inflation forecast is only accurate to within, say, + or - 2, by the time 2012 comes around, inflation might actually be, maybe, 2.7% despite the attempt to hit the 2% target. Then they have to tighten more. So it is not really true that "if I just target inflation I can hit that target with 70% certainty." Maybe you can hit it + or - 2% with 70% certainty, but that's about all.
What about NGDP targeting? One approach to NGDP targeting would also be to base your policy on the same kind of forecasting model. Of course, your forecast still might be wrong and NGDP growth might come in at 4 or 6% even if you targeted 5%, so you would have to correct.
Same problem as with inflation targeting, only you are better off in one important respect. You can miss your inflation target either because of a shock to the demand side of the economy, or a shock to the supply side. In the case of NGDP targeting, you only have to worry about the demand side shocks. Supply side shocks won't affect total NGDP, just the way they are split between inflation and real growth. So the size and frequency of misses should be smaller with an NGDP target than an inflation target, and as a consequence, the need for after-the-fact corrections would be less frequent.
There is another way to work the NGDP target that is potentially more accurate. That method involves establishing an NGDP futures market to help your forecasting. It is a little complex, but there is a good, nontechnical discussion here: http://bit.ly/sJJtBy
(3) "As long as inflation is low, why should the Fed EVER be trying to put a cap on GDP growth?" There is one important circumstance where they should do so. Suppose the economy is in a phase where productivity is increasing rapidly. Other things being equal, faster productivity growth would cause lower inflation, or even mind deflation. There is a danger that the Fed will push too hard to keep inflation up to its target by keeping interest rates too low.
When that happens, we get the strange case where the economy begins to overheat (in the sense that real output grows above its long run potential and labor markets become very tight), but there is little or no price inflation because productivity gains are holding prices down. The combination of low inflation, overly expansionary monetary policy and abnormally low interest rates leads to asset price bubbles, like the stock market bubble that burst in 1929, the dot.com bubble, and the subprime housing bubble. I think I mentioned earlier that Hayek pointed to this problem in his debates with Keynes in the 1930s.
There is a good short summary of some of these points in this post that discusses Goldman's recent endorsement of NGDP targeting: http://bit.ly/t4hkfB/
U.S. Inflation Data: Scant Fuel For Inflation Fears In September CPI Report [View article]
With regard to NGDP, it is one variable, not two. The whole point is that the Fed doesn't worry in advance how it is split up. If you want another lame example, it is like dividing up a quart of martini's among 8 guests. The host can easily control the amount of liquid he mixes, and then pour each guest 4 ounces of martini. It would be harder, not easier, to control how many inches of martini to pour in each glass, because some guests might have short wide glasses and some tall skinny glasses. Inflation targeting is more complex because it would be like trying to measure the width of each different shaped glass and then figuring out how many inches to pour in each glass. Think height of glass=inflation, width=real output, volume of liquid = NGDP. The volume is the basic variable you focus on, let the others come out how ever they do.
Let's switch this discussion to today's post if you want to continue it.
U.S. Inflation Data: Scant Fuel For Inflation Fears In September CPI Report [View article]
The superiority of NGDP targeting (or similar) is an old point. It was one of the main points made by Hayek in his critique of Keynes in the 1930s. With inflation targeting, productivity growth requires increasing the rate of NGDP growth to keep the rate of inflation constant. That risks causing bubbles, as in dot.com, subprime mortgages, etc.
You have hit on one of the most important arguments in favor of NGDP targeting.
"It's already clear that this Fed has caused inflation without any increase to income or employment." Not clear to me. Say, we now have 3.5 percent inflation and 1% RGDP growth = 4.5% NGDP growth. How can you be sure that if the Fed had slowed NGDP growth to 3%, we would now have 2% inflation and real growth? The evidence is to the contrary; part of the slowdown in NGDP would most likely have come out of RGDP and part out of inflation. For example (I'm just using these numbers as illustrations) we might have 0.5% real growth and 2.5% inflation, or 0 growth and 3% inflation. So it is not "clear." you have to look at the data. The data say that usually changes in NGDP growth induce (or reflect) changes in both P and RGDP components..