DeLong’s peculiar definition of ‘monetary policy’
Keynesians often live in a sort of alternative universe from me. One that reminds me of my daughter’s “Opposite Day” at school. For them, low interest rates mean easy money, for me it indicates money has been tight. Michael Woodford likes to envision a monetary system without money, where monetary policy is all about interest rates. I like to envision a monetary system without interest rates, to show how changes in the money supply are the essence of monetary policy. And now Brad DeLong argues that a policy of targeting future inflation isn’t really monetary policy:
You can call Federal Reserve policies aimed at the sending of signals that alter the expected rate of future inflation “monetary policy” if you want, but then you lose analytical clarity–because the way such policies work (if they work) is not the “normal” way that “normal” monetary policy works. Normal monetary policy works by shifting the private sector’s asset holdings toward assets that people spend more readily and rapidly, thus boosting spending. Quantitative easing at the zero bound does not do that: it simply exchanges one zero-yield government asset for another. What is does do is to change bond prices, rather by raising the safe short-term nominal interest rate and thus giving people an incentive to spend the money they already have more quickly.
In contrast, I regard policies that change the expected track of the money supply and inflation to be the only reasonable definition of monetary policy.
Before I explain why, let me emphasize that I am not the only one who takes a forward-looking approach to monetary policy. In the bible of new Keynesianism, Woodford argues that what matters is not current changes in the policy instrument, but rather changes in the expected future path of that instrument. He uses interest rates, but the same applies to the money supply. Of course in the long run money and inflation go hand in hand, a higher future expected money supply implies a higher future expected inflation rate, and vice versa. Maybe DeLong doesn’t buy into these newfangled new Keynesian models, so let me illustrate the point with a couple very simple thought experiments:
Consider two hypothetical experiments. In one case the monetary base is doubled, and is expected to remain at the higher level. In the second, the monetary base is doubled, but people expect the money to be pulled back out of circulation one year later. Let’s look at these hypotheticals first from a monetarist perspective, then a Keynesian perspective. (BTW, I’m not sure the second example is that hypothetical, given recent hawkish statements from the Fed it seems close to current policy.)
Now consider the impact on the price level. Monetarists would argue that the first case, a permanent doubling of the monetary supply, would eventually lead to a doubling of the price level. Indeed these sorts of one-time, permanent changes in the money supply are the bread and butter of textbook examples of the Quantity Theory of Money. As we’ll see, there’s a reason why they make this assumption.
Now consider the temporary currency injection. We know that the long run price level doesn’t change. Once the money supply returns to the original level (a year later) prices should also return to the original level. At first you might assume that monetarists would claim that the price level would double, and then fall in half. But consider the real interest rate. Monetarists typically assume that real variables like the real interest rate aren’t much affected by monetary shocks. But if prices doubled, and then were expected to fall in half, the expected real return on currency would be 100% in year two. That is, the purchasing power of money would be expected to double in year two. More likely, almost all of the temporary currency injection would be hoarded and prices would rise by just one or two percent—the risk free real rate of return.
Astute readers might notice that this thought experiment is quite close to the model Krugman used to explain the liquidity trap in Japan. But this shows much more than that; it shows that even in normal times, even when the real rate is slightly positive, it is still approximately true that only permanent changes in the money supply have a significant impact on the price level. Furthermore, I can slightly tweak the model to get even closer to Krugman. Let’s now adopt the Keynesian assumption that easy money temporarily depresses the real interest rate. In that case it’s not hard to imagine a massive and temporary currency injection driving risk free real rates close to zero. And then there would be little or no impact on the price level. Any impact would be washed out by the noise from other types of shocks.
To summarize, it is now true and it has always been true that the monetarist story about monetary shocks assumes those shocks are expected to permanently impact the money supply and the price level. It is hard to draw up any model where monetary policy can significantly impact the current price level without also impacting the expected future price level. And this isn’t just me saying this; Michael Woodford says the same thing.
So here is what DeLong has done. He has defined monetary policy to exclude any policies that can only be effective if they are able to move the expected future price level. By doing so, he has defined monetary policy in such a way that it excludes any policy that might actually be effective. I can see why an old-style Keynesian like DeLong would be attracted to a definition of ‘monetary policy’ that excluded any effective monetary stimulus. If you define monetary policy in such a way that only ineffective policy actions count, then monetary policy will be ineffective. And what’s left? Fiscal policy—big government.
How does he end up so far off course? Look at how he defines “normal” monetary policy. It’s all about swapping one asset for another. No discussion of expectations, although most explanations of really powerful monetary shocks (liquidity traps, currency devaluation, hyperinflation, etc) are heavily dependent on expectations. In fact, monetary policy is not about swapping currency for T-bills. It is about changes in the future path of currency (relative to demand.) Because money doesn’t affect interest rates in the long run, those changes lead (through the excess cash balance mechanism) to a change in the future expected price level. And that has an immediate effect on all sorts of asset prices; including stocks, commodities, and real estate. And if wages are sticky (and they are) this also has an immediate impact on employment.
Expectations are the key to the transmission mechanism. They are what monetary policy is all about.
(HT: Alejandro Rodriguez)
Update 10/7/09: I just noticed that Ambrosini made a similar argument exactly one day before I did. I need to actually started reading the other blogs on my blogroll. Apologies to Ambrosini for not noticing this oversight earlier.
Tags:
1. October 2009 at 05:17
Scott – you rock dude, in a nerdy way. You should be the Fed chairman.
1. October 2009 at 05:30
Why can’t you and Brad DeLong simply go with the definitions used by the Fed? I further suggest that you spell his name DeLong, not Delong.
1. October 2009 at 06:06
It’s worth mentioning that this idea is really Gresham’s law in another form. It’s an old idea.
1. October 2009 at 08:52
Hrm, another member of the “t-bill fallacy” club that clings to this perverse notion that the Fed primarily trades t-bills.
1. October 2009 at 08:57
As someone who reads both you and DeLong regularly (and sees more common ground than you might think), I have to admit that I really don’t know what DeLong is saying.
When he says “shifting the private sector’s asset holdings toward assets that people spend more readily” I’m not sure what he’s defining as assets. (If currency is an asset, and it’s the most readily spent of assets, then why would an expansionary monetary policy want to shift people into currency? If they seek to hold more currency, doesn’t that depress other asset values? Isn’t that deflation?) I am confused.
When he says “Quantitative easing at the zero bound does not do that: it simply exchanges one zero-yield government asset for another”… I’m guessing he is talking about a situation where the government is not already running a deficit.
When the government is already running a 1.2 trillion dollar deficit (courtesy of the Bushmeister), then certainly one can call QE either monetary policy or fiscal policy – one can make a semantic case for either. At a pragmatic level, I’m not sure I care.
But here’s a question for Scott (or Bill, perhaps) for clarification: what is the transmission mechanism for QE to inflation expectations, assuming the government is not running a deficit (and is not expected to run a deficit)?
1. October 2009 at 09:48
This idea: “Quantitative easing at the zero bound does not do that: it simply exchanges one zero-yield government asset for another”
Is silly. If I were to go into Texaco and try to pay for my petrol with a T-bill I would be laughed at. T-bills aren’t the medium of exchange. To make a remark like the above you have to model using a barter economy. In a barter economy says law is true in it’s simplest form, and there can’t really be an interest rate.
1. October 2009 at 10:43
As I and others have commented on DeLong’s blog, the statement is nonsense in several ways. It is hardly worth analysing. That said, I find it hard to believe that such reputable academics in the field as DeLong and Krugman can really be so ignorant of the implementation of monetary policy as to think that QE (or even conventional monetary policy) involves buying mainly t-bills, so I can only assume that they make that assumption to make it easier to argue that we are in a liquidity trap.
1. October 2009 at 12:59
I think I get it… took a while.
Let’s posit we’re at the zero bound, we’re in deflation, the government is not running a deficit nor expected to run a deficit.
If QE is deployed to buy private bonds, one can argue this is using government to direct the allocation of resources, and is “fiscal” policy. (Open market purchases of housing bonds, for instance, can be considered sector-specific support, and hence “fiscal”.)
If QE is deployed to buy govt. debt, and the public perceives an expectation the debt will be repaid without monetization, then it’s exchanging one liquid thing for another liquid thing. ASSUMING that the economy is not actually suffering a crude paucity of currency (in the sense of a coin shortage), one can posit this would have little effect. There is no impact on currency’s utility as a store of value.
Scott’s reply is that, by _definition_ monetary policy means changing expectations about the long term supply of currency. So what Brad is describing isn’t really monetary policy.
Brad’s (hypothetical) retort is that without an expectation of monetization, there is no expectation of a change in the long term supply of currency. Monetization cannot occur in deflation without a fiscal disbursement of funds (to inject currency or currency-linked assets into the economy). That’s because everyone believes future debt will be repaid in future dollars, and unlike in an inflationary environment, we believe those dollars will be less valuable (since interest rates are negative).
Scott’s (hypothetical) rejoinder is that, while this is all true – and that effective monetary policy at the zero bound may demand some degree of fiscal action – it is a mistake to believe that it’s the _fiscal_ aspect of the action that would drive the recovery, but rather that _fiscal_ action is merely the enabler that allows monetary policy to work.
IMVHO (_very_ humble opinion) – the deficit is between 1.5 trillion (give or take a few hundred billion).
[going to repost on DeLong’s blod, so everyone can get along and be friends]
1. October 2009 at 14:16
Statsguy:
Scott’s basic idea is if you assume business are influenced by their cost of borrowing (i.e., cheaper borrowing means that the cost of hold inventory is lower which means more goods can be moved at the same cost)–this is essentially DeLong’s model. Then, if you also consider that the cost of holding inventory relates to the inflation-rate. i.e., goods are bought at one price on way day and sold later in a devalued currency, then businessmen do see the ‘real-rate’ in their decision making, not the nominal-rate.
Thus, regardless of a ‘zero’ nominal-rate bound, the real-rate can be arbitrarily negative in this situations. (In reality, given any credible inflation, the easy-money nominal-rate will bounce above zero rather than the real-rate changing as in the prior thought experiment).
1. October 2009 at 14:23
Statsguy,
I won’t get into this business of what is or isn’t monetary policy.
However, I don’t really agree with the view you present (I understand it isn’t really your view).
Liquid things are not all equal. The market for french-fries is extremely liquid, but they can’t be used as a means of payment (well, not for most purposes). There is a significant difference in liquidity between currency, current account money and T-bills. The latter are significantly less acceptable, normal people haven’t even heard of them. Even stockbrokers and other financiers don’t take them as payment in normal situations, let alone petrol stations.
Now, what’s at issue here isn’t if there is a coin shortage, but rather if there is a general shortage of money, currency and checking account money. If there is then people have to spend less to obtain it. But, more importantly they think that they may have to spend less to obtain it in the future. So they save more now than they would have done otherwise, or hold more money than they would have done otherwise. The uncertainty isn’t necessarily around the price of money, it’s more around the ease with which it can be had in exchange for some other asset, such as a promise to pay back a loan in the future.
And don’t bother posting anything even vaguely critical into the comments on DeLong’s blog, he will delete it.
1. October 2009 at 16:00
Andy, Thanks, I guess.
Kevin. I do use the Fed’s definition you linked to. It is about the “availability of money.” A permanent increase in the money supply is inflationary, whether T-bill yields are zero or not. That is precisely the Fed’s definition, or at least one of them.
As you probably noticed, they intentionally made it so vague that anything could be considered monetary policy.
Current, It is an old idea, but I thought Gresham’s law was “bad money drives out good?”
Jon, Yes, I don’t see why people keep looking at monetary policy that way.
Statsguy, You said;
“As someone who reads both you and DeLong regularly (and sees more common ground than you might think), I have to admit that I really don’t know what DeLong is saying.”
I do sometimes agree with him. He just did a very reasonable piece on the pros and cons of whether money was too easy in 2003. I couldn’t find anything I would disagree with (although of course I would have examined the issue through NGDP growth.) But I assume we agree that now is not the time for the Fed to talk about contractionary policy, for instance.
Regarding transmission mechanisms, I like the monetarist “excess cash balance” approach. I like to carry $X dollars in my wallet. If the Fed doubles currency in circulation, I have more cash than I want. I can get rid of it by spending on goods, services, and assets. But due to the fallacy of composition, society as a whole cannot. They can only get rid of it by driving NGDP up high enough to where I am now in equilibrium with $2*X dollars in my wallet. And the expectation that this will occur over time has an immediate impact on asset prices. That’s basically how I look at monetary policy. BTW, in this story it makes no difference how the money is injected.
Current, Economists went down a dead end street because they misintepreted the liquidity trap. What is so frustrating is that Krugman understands that the key issue is whether the monetary injection is perceived as temporary or permanent. He understands that about liquidity traps, but I don’t know if he understands that that distinction is almost as important when you aren’t in a liquidity trap. But there is nothing special about the present situation, monetary policy is just as effective as in normal times–IF DONE PROPERLY.
Rebeleconomist, I have no idea why he makes these statements either. I assume he believes it, but isn’t the important issue to discuss the question of whether inflation targeting is the solution to our problems, not whether it should be termed “monetary policy?” I mean, if the Fed’s doing it, and they are using their monetary tools to implement it, how can it not be monetary policy?
Statsguy#2, Obviously monetary policy cannot occur in a vacuum. If you drop money from helicopters it is “welfare.” If you buy bonds it reduces the interest the government must pay on bonds held by the non-Fed public. I can’t see how there can be any “pure” monetary policy that doesn’t somehow impinge on fiscal policy.
Jon, I agree.
Current, I think the only way Statsguy’s second point would work was if the liquidity trap was expected to last forever. But in that case the value of T-bills would be fixed at par, and they’d be nothing more than large denomination bills, they’d be part of the currency stock. If the liquidity trap is not expected to last forever, then people will expect prices to rise once cash and T-bills are no longer perfect substitutes. But that expectation will cause asset prices to rise immediately.
1. October 2009 at 20:13
Thank you… these were all helpful answers. Must think more.
2. October 2009 at 02:57
Quick question about the ability of central banks to achieve “temporary” objectives:
This result you achieve about the impotence of temporary money-supply boosts (i.e. prices will fall back – this is anticipated – hoarding – so it never happens) – does it also apply to the first deriviative i.e. inflation?
So can a Central bank credibly say “I will produce 5 years of 5 % inflation and then revert to 2% inflation”?
Or will the same expectations-effects make this impossible to achieve?
Many thanks
Giles
2. October 2009 at 03:23
Well, what Gresham actually said was “that good and bad coin cannot circulate together”.
The explanation behind it is very similar to what you’re discussing. Gresham’s law describes commodity money regimes. In the times it described there was no long term alternative to specie currency. However, legal tender laws meant that coins containing different amounts of precious metal but the same face value existed. From time-to-time a monarch would debase the coinage. When that happened people tended to use the debased coinage for transactions. The reasons for this were quite simple. If the debasement was permanent then the price of bullion will eventually become denominated in debased coins. The specie value of a coin would be larger than it’s face value, so the best thing to do with it is to, clip it, or export and have it melted down. If the debasement is temporary then the debased coins must at some point cease to be legal tender, so it is best to store the good coins and use the bad ones until that happens. In this latter case we have the “temporary injection” situation.
What you suggest in the first paragraph I quote above is something I don’t quite agree with. Since if people keep their hoards depends on how long they generally need to hold money for. If it’s a matter of years people will hold interest bearing assets or real assets. But I mostly agree with it and I’ll defer arguing about details until some other time. I agree that if the value of money is going to fall and then rise later it may be best for agents to hold onto a hoard. This is what motivates people to hold hoards of good coins in the Gresham’s law situation.
2. October 2009 at 04:56
I agree.
I don’t quite agree with what you and Statsguy are saying. Money is a short run thing, isn’t a creature of the long run. I don’t hold any money to buy things in 2011. Instead I hold property and I have plans on how to use it. I know that I can liquidate some of that property if I need money in 2011. I hold money for purchases in the short-term future that I can’t plan.
The liquidity trap is also to do with quite short amounts of time. Of course exactly how short depends on the institutional circumstances of the time. (In the Elizabethan times I mention above things were probably quite different and hoarding money may have been a practical means of saving).
You can’t have a “liquidity trap that is expected to last forever”. Such a thing assumes (according to the Keynesian liquidity preference explanation) that speculators hold money because they think that doing so may allow them to make a profit from an opportunity to buy an asset. In such a case there must be such opportunities.
Consider how much sense the following case makes. An investor thinks there will be a recession for 10 years. So he holds his money in cash for 10 years waiting until an opportunities arises to invest it in a rising asset. Clearly such a thing wouldn’t happen. If an investor thought there would be a 10 year long recession he would put his capital into rentable assets with scarcity value.
This is why I don’t really quite agree with your view on the injection of money. If an injection remains in place for a long time scale, into the long term I talked about above, then people will spend it. Because, they don’t need to hoard money at all except in a shorter-run. I think this business of the difference between temporary and permanent injections really refers to the short run.
2. October 2009 at 06:26
Giles, No, it is based on the fact that the expected rate of deflation cannot exceed the real interest rate, so it doesn’t restrict any expected slowdown in the rate of inflation.
Current; You said;
“What you suggest in the first paragraph I quote above is something I don’t quite agree with. Since if people keep their hoards depends on how long they generally need to hold money for. If it’s a matter of years people will hold interest bearing assets or real assets.”
But in the scenario I discussed the real rate of return on cash would equal that of bonds. The nominal interest would be zero
Current; You said;
“I don’t quite agree with what you and Statsguy are saying. Money is a short run thing, isn’t a creature of the long run. I don’t hold any money to buy things in 2011. Instead I hold property and I have plans on how to use it. I know that I can liquidate some of that property if I need money in 2011. I hold money for purchases in the short-term future that I can’t plan.”
This confuses two issues. One is how long people typically hold cash, and the other is the issue of expectations. Yes, people usually spend cash rather quickly (although not in a liquidity trap) but even so, they form price level expectations over a longer period. Individuals and businesses are very interested in what stocks, commodities and real estate will be worth in 5 years. And one reason they are very interested is because new information about changes expected to occur in 5 years, will immediately change the value of those assets. I assumed that if the liquidity trap was expected to end, then people would expect higher prices to result from the monetary injection, and that would raise asset prices right now. Nothing in that argument requires people to hold cash for a long time.
2. October 2009 at 07:57
But, you can’t realistically hypothesis of such a thing. Certainly the short-run rate of interest may fall to zero. But as we expand the period of time which we are considering more types of investment become possible. In a market economy we cannot say that all of them will produce no return, that would deny the existence of scarcity and/or time-preference.
3. October 2009 at 09:40
Current, We are going around in circles here. Yes, I don’t realistic think all rates would be zero, but that is what is required for the liquidity trap hypothesis to hold. I was just making that assumption to show what would happen if it was true. I agree that DeLong’s idea of monetary ineffectiveness is implausible, but you were arguing the other side. One can’t have it both ways, either the liquidity trap will last forever, in which case government bonds are actually cash, or it won’t last forever, in which case there is no problem of monetary policy ineffectiveness.
3. October 2009 at 10:19
Reading about disagreements on the most fundamental princples of macroeconomics brings me to the conclusion that it is a pseudoscience at best, though not every economist is a pseudoscientist.
I think I know why this is. Cosmology is also an observational science, but it is built upon micro principles that fit the data better than any other models in any other fields, period. What would cosmology be without quantum mechanics?
I think the same goes for macroeconomics. It should be built up from micro theory, which must be behavioral. I realize that many consider this too difficult, but it really isn’t. I model brain and behavior all the time, and my results are far more concrete. This will always be the case, but macroeconomics should be further along than it is.
3. October 2009 at 11:08
I’m not intending to argue DeLong’s side, if that’s what you mean by “argue the other side”. Though I’m saying something slightly different from you.
I think I’ve confused two different issues. I’ll rectify that.
Regarding the “long term liquidity trap” this idea goes against the correct version of Say’s law. I think we agree about that.
Regarding temporary injections…. Let’s say we have:
M1 = k1 P1 T1
The central bank then inject money, which is expected to be temporary and not affect prices much. So we have:
M2 = k2 P2 T2
And k1 ~= k2, M2 > M1. So, either P or T must rise, or both must rise. There are limits on T rising so there must be a rise in P too.
This can be looked at another way. Suppose the central bank buy bonds to raise the money supply. The first receivers of money, the former bond owners, have not changed their plans much. They don’t think that the change in money supply is permanent so they continue to hold the same regular money holdings as before. They spend the currency they have obtained. This spending though will result in profits elsewhere for the next receivers of the money. Those people may realise the cause of the higher demand for their goods or property, and understand that it can’t last. But, they may not, and as the money spreads further so it becomes harder to tell. So, some agents will decide to respond by raising their prices to make more money from the new demand (that is they will use Hayekian “price signaling” rather than something more sophisticated). Although they may know that the money injection is temporary they can’t necessarily know that they are seeing an effect of that injection.
So, for a temporary injection prices rise more slowly. But, I think they will still rise. Mises discussed this by using the product of M and P. In this temporary situation we start with money M1 and prices P1. Later we have money M2 and prices P2. In this case M1 P1 < M2 P2, we don’t have M1 P2 = M2 P2 as some early monetarists thought. However, P1 doesn’t equal P2.
4. October 2009 at 06:02
Mike, You said;
“I think the same goes for macroeconomics. It should be built up from micro theory, which must be behavioral.”
Some of the smartest people in the world have tried to do this for decades, and no one has succeeded. It may be too hard. I prefer reduced form models, which simple assume that there is some wage and price stickiness, for reasons that are not entirely clear.
Cosmology is not a good analogy, meteorology is much closer. It is silly to try to emulate cosmology, as the fields are so different. By the way, cosmology has not been able to answer a single basic question, such as the size of the multiverse, its age, what is time, what is matter, what is space, what is causality, or whether the laws of nature are stable over time. So it’s not as successful as people think. We know almost nothing about the multiverse. Indeed some respected scientists think that current theory implies that our universe is probably just someone’s dream. I think the theory is called something like “Bolzman’s brain.”
Current, You said;
“Regarding temporary injections…. Let’s say we have:
M1 = k1 P1 T1
The central bank then inject money, which is expected to be temporary and not affect prices much. So we have:
M2 = k2 P2 T2
And k1 ~= k2, M2 > M1. So, either P or T must rise, or both must rise. There are limits on T rising so there must be a rise in P too.”
I don’t follow. If the currency injection is temporary, it will be hoarded. In that case k will fall as fast as M rises.
Prices may rise a tiny bit, but the maximum they could rise is the real rate of interest. And that’s not very high in an economy flooded with cash. Right now the real rate of interest is probably negative, so prices wouldn’t rise at all.
4. October 2009 at 11:56
I don’t see why you think that would be the case. If the change is expected to be temporary then why should people increase the stock of money that they normally keep?
5. October 2009 at 06:11
Current. Suppose the money supply doubles for ten minutes, and then contracts again. And you know this will occur. The QT says the price of houses will double and then fall in half. If you have a $100,000 house it should rise to $200,000 and then it will be expected 10 minutes later to fall back to $100,000.
Does this sounds farfetched? Would you pay $200,000 for a house that you expected to be worth $100,000 ten minutes later? Clearly not. Now lets make it 20 minutes, then 40 , then 80, then 160 . . . Exactly where does my argument break down? I claim that depends on interest rates. The expected rate of deflation can’t exceed the real interest rate. And that limits the short term upside from temporary currency injections. Over ten minites (or 10 days) the real interest rate is essentially zero. So if temporary currency injections don’t cause inflation, they must be hoarded.
What motivates the hoarding? Nominal rates fall enough so that people or banks are willing to hold larger real cash balances for short period.
5. October 2009 at 07:04
I see your point, however I still don’t quite agree.
Notice that in your last two sentences you are proposing the view that a loosening of money is related to a fall in nominal rates. This is a view you have criticized on other occasions.
Let’s say that agents realize that the currency injection is temporary. So, they do not change their current plans. Let’s suppose, just for now, that the interest rate doesn’t change. If that is the case then we have the following situation. The first receivers of money from the injection will continue to hold the same stock of money that they held previously. They will spend the money and pass it on to others, who will also likely spend it. In time this will causes some markets to tighten and some sellers to make more profit and change their plans.
The only condition where this would not take place is if the reduction of interest rates that it causes offsets the effect. For this to happen the fall in interest rates would have to cause the first receivers of money to increase the amount they demand for holding by exactly the amount of the increase in supply.
I’m not saying the monetarists are right. I’m just saying that your own view is a simplification of the situation.
5. October 2009 at 09:15
Dr. Sumner,
You replied:
“Some of the smartest people in the world have tried to do this for decades, and no one has succeeded. It may be too hard. I prefer reduced form models, which simple assume that there is some wage and price stickiness, for reasons that are not entirely clear.
Cosmology is not a good analogy, meteorology is much closer. It is silly to try to emulate cosmology, as the fields are so different. By the way, cosmology has not been able to answer a single basic question, such as the size of the multiverse, its age, what is time, what is matter, what is space, what is causality, or whether the laws of nature are stable over time. So it’s not as successful as people think. We know almost nothing about the multiverse. Indeed some respected scientists think that current theory implies that our universe is probably just someone’s dream. I think the theory is called something like ‘Bolzman’s brain.'”
My team has made significant progress developing deterministic micro models of brain and behavior. It isn’t hard to imagine these and other models making significant impacts on macroeconomic theory, though no one on my team has the macro expertise to begin. We think we’ve answered many of the most fundamental questions about brain and behavior, including the mathematics of human motivation; explicit models for learning; explicit models for the actions and interactions of neurotransmitters; what mood is, as we operationalize it; how many emotions there are and how they function and are defined; among other things.
We operate from the perspective of the economics of inclusive fitness, or passing genes, and all the models are mathematical. We hope to begin publishing soon. I would share more, but this is obviously not the forum for this. However, those interested can email me at m_sandifer@yahoo.com.
With respect to cosmology not being a good analogy, but meteorology being much closer, I disagree.
First, the physical laws determining weather models are actually very well understood. There are nonlinear dynamical elements to understand, but the principal problem in modeling weather is a problem of measurement. It is very hard to collect enough data on key variables to allow for more precise modeling and predictions.
I think the parallels between cosmology and macroeconomics are more compelling in this context. In cosmology we have gravity and other fundamental forces determing the shape of space-time and the lines of information transmission in the universe. These are often easy to measure, given the apparent behavior of observed mass and energy. Likewise, in macroeconomics, the existence of money as a medium for measuring financial incentives gives macroeconomists a significant advantage vis-a-vis meteorologists. And while many of the newer ideas in cosmology seem ridiculously speculative, to say the least, we shouldn’t forget the tremendous progress made in cosmology in decades past.
If we measure progress in cosmology against what may be the actual scope of the universe, it can be seen as a science in its infancy. However, compared to the understanding that existed at the beginning of the 20th century, much more progress has occured than in macroeconomics over the same period. We discovered that the universe is at least trillions of light years in expanse, that the observable universe is expanding, that the speed of light is variable as the universe flattens out, much about the energetic dynamics of stars, much data to support the existence of black holes, predictions of slight differences in the cosmic background energy resulting in the clumps of matter we see today, dark matter and energy, much to confirm the cosmological constant, evidence for super high density stars such as white dwarfs and pulsars, quasars, and many other counter-inuitive concepts.
However, as far as I can tell, macroeconomics hasn’t fared nearly as well. From my perhaps primitive perpective, many of the debates of the early 20th century have not been resolved. There are still some adherents to Say’s Law. There are still very fundamental questions about the effects of fiscal versus monetary stimulus. There still seems to be far too many who fail to understand that crowding out is not a worry when there are significant output gaps coupled with deflation. There are fundamental questions about the effects of trade surpluses versus deficits under varying circumstances. ]
As I see it, the difference in the development of cosmology and macroecomics is that the former is heavily informed by micro theory.
5. October 2009 at 13:33
Say’s Law, when properly stated is certainly correct. When properly stated though it describes a long-run trend, not a short run one. What it tells us is that there will always, eventually, be recovery from a recession, as long as the government doesn’t intervene to fix prices. (I can explain why if you’re interested).
The problem with Say’s law is that it doesn’t tell us about the shorter run. It doesn’t tell us what happens in the time before some important prices can be changed. All this is the interesting stuff.
I agree with you that macro needs a micro foundation.
6. October 2009 at 08:25
Current,
Say’s law works in barter systems, but only in the long run at best in monetary ones. But, even if Say’s law holds up over the long-term, in the short run it doesn’t necessarily have much in the way of policy implications.
6. October 2009 at 10:01
Mike,
That’s fairly much correct. If there is an increasing demand for money then recovery does not occur straight away. Say recognizes this himself, but says that whenever more money is needed it can always be provided by banks of issue. (This is the basic free-banking claim which I won’t go into here).
Well, that’s what I thought until recently. However, Say’s law is still important on two counts. Firstly, it shows that early Keynesians were wrong. For there to be a perpetual recession and fall in AD there must be an infinite demand for liquidity. Any theory which contains this as a component must require revision.
Secondly, it is useful for understanding the process of recovery. Say’s law stated in it’s more correct form is “A man who applies his labour to the investing of objects with value by the creation of utility of some sort, can not expect such a value to be appreciated and paid for, unless where other men have the means of purchasing it. Now, of what do these means consist? Of other values of other products, likewise the fruits of industry, capital, and land. Which leads us to a conclusion that may at first sight appear paradoxical, namely, that it is production which opens a demand for products.”
Or, as Hutt put it “the demand for any commodity is a function of the supply of noncompeting commodities”.
As Jerry O’Driscoll persuaded me, what Say has said directs our attention to what happens in a recovery. For example, suppose that you live in a place in the US where property prices have fallen greatly. So, as these prices have fallen the ability of those who owned the houses to demand other local goods has also fallen. However, there can never be “general overproduction”, overproduction is always of a set of goods compared to demand. This shows how misdirection of production results in a lower level of prosperity for the general community. Sticky price exacerbate this problem of decline by generating unemployment.
Say gives an example of a merchant given the choice of locating in a minor town or in one of the great cities of Europe. He points out how even though there may be more competition in the cities it may be better to locate there because there are more residents who make more products and have more to sell. The same sort of thing that applies to geography applies also to other markets. It may be better for a car company to make mass market cars even though it’s very competitive, or it may not.
Consider something like the stimulus program from this point of view. In what places and what markets it stimulates production may not be the same places and markets where it can be sustained. Hopefully businesses will understand this and not invest poorly on account of the temporary situations it creates.
6. October 2009 at 13:17
Current,
You replied:
“For there to be a perpetual recession and fall in AD there must be an infinite demand for liquidity.”
What do you mean by “perpetual” in this case?
“For example, suppose that you live in a place in the US where property prices have fallen greatly. So, as these prices have fallen the ability of those who owned the houses to demand other local goods has also fallen. However, there can never be “general overproduction”, overproduction is always of a set of goods compared to demand. This shows how misdirection of production results in a lower level of prosperity for the general community. Sticky price exacerbate this problem of decline by generating unemployment.”
First, are you claiming that the above occurs only in closed economies, ro do you take a wider view?
Second, since prices are sticky, unemployment occurs and deflation is self-feeding(sans intervention), then what what good does Say’s law do?
With respect to stimulus programs, I can buy that their effectiveness is limited by lags and constraints that large debts may incur, but I thought the point was to essentially increase money velocity with a sometimes generous multiplier. As my macroecomomics professor used to say, it doesn’t really matter what the money is spent on as long as it’s spent.
6. October 2009 at 16:57
Current, You said;
“Notice that in your last two sentences you are proposing the view that a loosening of money is related to a fall in nominal rates. This is a view you have criticized on other occasions.”
I see why you say this, but those earlier statements referred to permanent increases in the money supply. In fact I have argued the recent rise in the MB is ineffective partly because it is viewed as temporary. But I still say you must stand back from the simple monetarist story, and think about price expectations. Asset prices won’t rise if they are expected to fall back quickly. It doesn’t mean temporary currency injections always have zero effects, but it limits the effect more than you’d think if you are just considering the “excess cash balance” monetarist mechanism (which I fully agree is very important in the long run.)
Mike, All I can do is repeat that very smart people are trying to integrate micro and macro, but it is hard. The macro economy is very complex (like weather) and economists don’t agree as to what simplifications get at the core issues. We all have a micro model of the apple market that we understand pretty well, but we have trouble scaling it up because issues that seem trivial at the micro level become important at the macro level (like price stickiness.)
I agree with both of you about Say’s law, it’s best thought of as a long run relationship. But that is still very important for subjects like growth theory, which is about the long run.
7. October 2009 at 12:20
What I mean is that it can’t last forever. Liquidity traps are temporary phenomena due to temporary circumstances.
You’ll understand what I mean if I explain the underpinning micro argument.
Suppose you have an investor/speculator and a wage worker. Both are types of consumer. I’ll consider both separately though the argument is similar for both.
At the start of period 1 there is a disaster, a “real shock”. Some large set of assets are destroyed, a natural disaster for example. Our investor decides that this threatens his security, so he sells some assets and holds more currency. Similarly, our worker spends less and holds more currency. Both show liquidity preference.
Now, with an inelastic supply of money these preference lead to the social effects you described earlier. The amount of money used for transactions must fall, and recession ensue.
So, at the beginning of period 2 we have less transactions than before, a depressed economy. Now, this wider recession may make our investor hold more cash, and our consumer do the same. For the same reason as before, uncertainty about the future.
This may happen for several further periods, depending on how long a period is. However, as time passes the future becomes less uncertain. Investors realise the future will be bad. As a result they invest in very safe investments such as land rather than continue to earn nothing from cash. Similarly, workers realise the future will be bad and plan for it, which entails reducing the amount of money they normally hold. So in time the impasse is breached, it depends on how long uncertainty lasts in the population of economic actors.
I’ll see if I can find an article to explain this better. I found a good one once, but I had a look in my books an downloads but I can’t find it right now.
8. October 2009 at 08:33
I’m not arguing the “pure monetarism” position here. My point is that even with a temporary injection you can’t eliminate a change in prices, though it’s temporary nature my make it smaller. As I described above, agents who receive new money will spend it and cause tightening in other markets. This will lead to changes in pricing and profits and then to changes in expectations. Some of these will be annulled by the knowledge that there will be a reversal of the policy. But, they can’t all be annulled.
11. October 2009 at 07:17
Current, I’m not sure we disagree, my only point is temporary currency injections have less impact than people surmise. A doubling of the money supply might easily cause prices to rise 1/2%, or even less. Most people don’t realize that.
15. October 2009 at 04:56
I like that De Long and Krugman both pick and choose their definitions of fiscal and monetary policy to suit the story at hand. One moment you have De Long casting all of the “extraordinary” monetary actions as fiscal policy, the next they both whine about $800 billion in stimulus being only about 50% of what’s needed. But when you count fiscal policy the way De Long does haven’t we had approx $2.5 trillion in fiscal policy? (0.8 ARRA boondoggle, 0.7 TARP, 1+ Fed QE) So maybe they wouldn’t call the non-ARRA stuff “stimulus”, but I shudder to think about the state of the economy without it. This lack of consistency points to a very broken analytical framework.
15. October 2009 at 07:24
I agree if you’re talking about CPI and a temporary injection. However, if the injection is over a long enough time scale then it will be spent and will start having an effect on profit and loss accounts. The price increases will begin in the prices of assets. Even with anticipation this can’t be avoided.
17. October 2009 at 07:00
Thruth, That is a very good point. I wish I had mentioned it.
Current, Over a long time scale I agree.
7. May 2011 at 23:58
[…] 三ã¤ã®ç´ æを使ã„ãŸã„。一ã¤ã¯ã€é‡‘èžæ”¿ç–ã¯ã‚¹ãƒ´ã‚§ãƒ³ã‚½ãƒ³ã®ã€Œäºˆæƒ³ã‚’目標ã«ã™ã‚‹ã¹ãã€ã¨ã„ã†è€ƒãˆã 。政ç–ã¯NGDPã‚‚ã—ãã¯ã‚¤ãƒ³ãƒ•ãƒ¬çŽ‡ã®äºˆæƒ³ãŒç›®æ¨™ã¨ä¸€è‡´ã™ã‚‹ã‚ˆã†ã«è¨å®šã•ã‚Œã‚‹ã¹ãã§ã‚る。第二ã¯ã€ã€ŒADã¯ä»Šã®é‡‘èžæ”¿ç–ã®è¨å®šã‚ˆã‚Šã‚‚政ç–ã®äºˆæƒ³å°†æ¥çµŒè·¯å…¨ä½“ã®å¤‰åŒ–ã«ã¯ã‚‹ã‹ã«å½±éŸ¿ã•ã‚Œã‚‹ã€ã¨ã„ã†ãƒžã‚¤ã‚±ãƒ«ãƒ»ã‚¦ãƒƒãƒ‰ãƒ•ã‚©ãƒ¼ãƒ‰ã®è°è«–。(ã“ã®ãƒã‚¹ãƒˆï¼ˆé‚¦è¨³ï¼‰ã¯ã“ã®è€ƒãˆã‚’大ææ…Œã«é©ç”¨ã—ãŸã‚‚ã®ï¼‰ã€‚一時的ã ã¨äºˆæƒ³ã•ã‚Œã‚‹é‡‘èžæ”¿ç–ã¯ã»ã¨ã‚“ã©å…¨ãADを変ãˆãªã„。第三ã¯ãƒãƒãƒ¼ãƒˆãƒ»ã‚ングã®æ´žå¯Ÿã§ã€ã€Œãƒ•ã‚©ãƒ¯ãƒ¼ãƒ‰ãƒ»ãƒ«ãƒƒã‚ングãªIS-LMモデルã§ã¯ã€ç·©å’Œçš„ãªé‡‘èžæ”¿ç–ã«ã‚ˆã£ã¦ã‚¤ãƒ³ãƒ•ãƒ¬æœŸå¾…ãŠã‚ˆã³æˆé•·æœŸå¾…ãŒå¢—ã™ã¨åç›®ãŠã‚ˆã³å®Ÿè³ªé‡‘利ã¯å対ã«å‹•ãã€ã¨ã„ã†ã‚‚ã®ã 。ã¤ã¾ã‚Šãƒžãƒãƒ¼ã‚’引ãç· ã‚ã‚‹ã“ã¨ã«ã‚ˆã£ã¦åç›®ãŠã‚ˆã³å®Ÿè³ªé‡‘利を引ã下ã’ãŸã‚Šã€ãƒžãƒãƒ¼ã®ç·©å’Œã§åç›®ãŠã‚ˆã³å®Ÿè³ªé‡‘利を引ã上ã’ã‚‹ã“ã¨ãŒã§ãる。ã“ã‚Œã¯å˜ã«ç†è«–çš„ãªå¥½å¥‡å¿ƒã«ã‚ˆã‚‹ã‚‚ã®ã§ã¯ãªã„。ã“ã®ãƒã‚¹ãƒˆï¼ˆé‚¦è¨³ï¼‰ã§ã¯2007å¹´12月ã®é€†ã‚µãƒ—ライズãŒ3ã‹æœˆã‚‚ã®ã®çŸæœŸå›½å‚µã®åˆ©å›žã‚Šã‚’事実ã¨ã—ã¦ä¸‹è½ã•ã›ãŸã“ã¨ã‚’示ã—ã¦ã„る。 […]