One way to learn macroeconomics is to figure out when money is neutral, super-neutral, and non-neutral, and when it is not.
Money is said to be neutral when a once-and-for-all change in the money supply or money demand has no real effects. Money is super-neutral when a change in the growth rate of the money supply (or demand) has no real effect. And money is non-neutral when a change in the supply or demand for money does have real effects.
My new book begins with an examination of money neutrality, then covers money super-neutrality, and then covers money non-neutrality. No macro model is perfect, but the following claims seem like a good approximation of reality:
1. Money is neutral in the long run, but not the short run.
2. Money is approximately super-neutral in the long run, but not exactly so. This is especially true if there are other distortions such as taxes on nominal investment income.
3. Money is strongly non-neutral in the short run, as monetary shocks affected real wages, real output, employment, real interest rates, real exchange rates, debt defaults, and many other real variables. The short run can last for years.
With this framework, one can avoid a number of fallacies. Thus you should be very skeptical of claims that a particular monetary policy is impacting highly persistent structural issues such as income inequality or chronic trade deficits. But you should also be very skeptical of claims that printing money cannot possibly produce short run benefits, such as recovering more quickly from a recession.
I’m surprised by how many people struggle with the fact that you can’t print your way to prosperity, but you can print your way back to prosperity.
You can think of my new book as an attempt to justify the claims made in this post, and also to flesh out the implications of these claims for recent macroeconomic events in America and elsewhere.
READER COMMENTS
marcus nunes
Sep 29 2021 at 11:00pm
Money, more specifically monetary policy, is very powerful. It is responsible for the patterns observed over many decades. Since it is said that humans “are pattern-seeking, story-telling animals”, it may be a good start for understanding the process.
https://marcusnunes.substack.com/p/macroeconomic-patterns-and-stories
Sven
Oct 1 2021 at 6:54pm
Money is not neutral in the long run. It is just a fallacy. I would recommend to you the book below which explains how money transforms the real economy. Prof. Sumner has not even a clue how money works in the long run.
https://www.amazon.com/dp/B093RWX8FX
MarkLouis
Sep 30 2021 at 7:22am
If money is neutral in the long-run, doesn’t that imply that any real short-term effects are ultimately reversed?
Scott Sumner
Sep 30 2021 at 11:07am
Yes.
Lizard Man
Sep 30 2021 at 11:05pm
What is the evidence that money is neutral in the long run?
Just to give one counter-argument, the fertility rate in the US fell dramatically in response to the Great Recession, which I believe Sumner would argue was the result of monetary policy. So monetary policy had the effect of preventing millions of births over a decade. That is a demographic shock which will likely reverberate for a generation or two. Is a couple of generations still considered the short-run?
Thomas Lee Hutcheson
Sep 30 2021 at 8:32am
Don’t you think that most economists agree with you on this? [I allow that some who “agree” might come to different opinions about the merits of fiscal and monetary policy, possibly illogically.]
Sven
Oct 1 2021 at 7:03pm
Yes. They would. It is an orthodox view about money. However, it is false. Money fundamentally changes production and consumption, and income and wealth.
Sven
Sep 30 2021 at 10:17am
Money is never neutral. Money has varying effects in different time periods.
Prof Sumner, you say that money is neutral or super neutral in the long run. That is a misconception economists could not understand so far.
The expansionary monetary system changes relative values of goods and services. Values are raised artificially. Therefore, income inequality aggravates in the long run. And this long-term pattern is deflationary. Zero lower bound interest rates are not a coincidental outcome. It is not about inflation. It is about economic growth. Economic growth which is enabled by money supply growth is inherently a deflationary process. (For the clarification, I do not claim that economic growth is maintained by money supply growth. It is only possible with money supply growth with accompanying productivity growth.)
And short run as you point out is already non neutral.
I conclude my point with your words.
“You can think of my new book as an attempt to justify the claims above, and also to flesh out the implications of these claims for recent macroeconomic events in America and elsewhere.”
Scott Sumner
Sep 30 2021 at 11:10am
Sven, You said:
“The expansionary monetary system changes relative values of goods and services. Values are raised artificially. Therefore, income inequality aggravates in the long run.”
Here’s your mistake. The change in relative prices only occurs in the short run. Thus, income inequality is not significantly impacted in the long run.
Allen
Oct 1 2021 at 12:33am
And does your book convey the process by which changes in relative prices are reversed?
Sven
Oct 1 2021 at 9:32am
Prof. Sumner,
Unfortunately, you did not get what I want to mean. I will try to explain with a thought experiment.
Let us say we have an economy that has a value of 1.000. In monetary terms, MV=PY=1.000 Let’s say M is 200 and V is 5.
Now let’s say our economy grows by %1 by the introduction of a new invention. Subsequently, it becomes 1.010 assuming V is constant in the short term. (Most probably V will fall a little bit by an increase in money demand) So, MV=PY 202.5=1.010
Now assume that our fictional economy grew with an invention that did not exist before. And the value of the new invention is determined by its producer. The producer will set a price for that product after deducting the costs. Since this producer is the only one in the market, in other words, it is a monopoly in the beginning; it can set a high price beyond the cost of the product. And hence, it would set an optimal price that maximizes profit. Pi.Qi would be the total quantity of product with its initial price that is sold in the market. (Pi is the initial price and Qi is the initial quantity) This is how it works in the real world that we live in. It is expected that the price of this product would fall in the long run by the introduction of new producers to the market. However, the prices of products do not recede to the perfect competition level.
On the other hand, if the money supply does not increase in a parallel course to growth, the price of the new invention would be determined at a lower level than the level that is set in the expansionary monetary system. Both Pi and Qi would be lower. And with the introduction of new producers to the market, the price of this product recede to perfect competition level after a period of time.
I hope you will understand my reasoning. It is a very different approach compared to existing schools of economics.
Real-world Example;
Now imagine you are Steve Jobs. You invented iPhone. You are setting IPhone’s price at an optimal level that maximizes Apple’s profit since the monetary system we have allows it. As you can also observe in the real world Apple makes huge sums of profit. Or Microsoft sells Windows and Office with high values for years. We can extend it to whole product market. There is no perfect competition in the real world as classicals and neoclassicals claim. Fundamental postulates of classical economics were actually for the constant money supply system.
The expansionary monetary system changes the relative values of goods and services in the long run (I admit it also changes relative values in the short run, however, I consider this phenomenon insignificant). And it creates a growth-oriented economy. Consequently, it is called capitalism. And capitalism is not a market economy.
Don Geddis
Oct 3 2021 at 8:20pm
Your example is very sloppy, even with the simple math.
You didn’t complete your example. What do you think P and Y are?
OK, that is a change in the real economy. So Y increases. (From what to what?)
No, you’ve already made many mistakes. What do you think M, V, P, and Y are, after the new invention? You never mentioned anything about the money supply changing, so M must still be 200. This means that V must increase, in order for nominal GDP to increase to 1010. You said V is constant; this is already wrong.
Unless you think M increases? But that’s a central bank choice, determined by monetary policy, and you didn’t mention anything about it. What’s important is that the change in M has nothing at all to do with the 1% new technology invention. They are completely independent.
You’ve asserted this multiple times, but you’ve given no explanation at all for why any relative prices would change. Your example was about a change in the real economy (the development of a new invention). That has nothing at all to do with this claim, which would be about an economy where the real factors are unchanged, but only the money supply changes. In your example, increase M from 200 to 250 (but don’t change Y at all). How does this change “the relative values of goods and services”? You don’t even offer a hint of any explanation.
Sven
Oct 4 2021 at 9:26am
Dear Don Geddis,
The explanation might look like incomplete. Since this is a comment section of a blog post, I tried to explain with minimum words as possible. I wrote a whole book how monetary system changes the real economy. I put the link above. If you are interested, first two chapters are enough to understand the logic.
However, I will try to explain it again here. I hope it will answer your questions.
So, our economy (GDP) has a value equals to 1.000. This is not nominal. This is real value. Quantity equation is normally used to understand nominal price changes as opposed to an increase in money supply. However, we can also use it to understand real growth changes. That means it is an equation to figure out changes.
So at To which means when GDP is 1000 P is constant. In my example, M is increasing in parallel to an increase in Y. In other words, real output grows year to year. So, M, while at T0, is 200 shifts to 202 at T1. Now M(202).V(5) =P.Y 1.010
I take the V is constant but in the real world V falls at some degree. However, the main point here is the output growth rate. So, if the output growth rate (which is determined by productivity growth) is 1%, central bank adjusts money supply growth according to output growth rate. If V falls little bit, central bank can increase money supply growth more than 1%. This is how it works in the real world. I abstracted inflation in this example. So, inflation rate in this economy is zero.
As I said in the example, we have a new invention. Let say it is Iphone. Iphone contributes 10 point to our GDP. This 10 point is determined by Iphone’s price(x)quantity. Let say our IPhone is $1. So, 1×10=10 Iphone is sold in the market. And its cost is $0.5 per Iphone. Hence, Apple made $5 profit.
Now let us imagine how a constant money supply system would work with a new invention. Our GDP is 1.000 again at T0 in this example. We invented Iphone. At T1 our economy again must be at 1.000 since we cannot increase money supply. Therefore, Apple HAS TO set Iphone’s price at a lower level since there will not be the same demand level at the same price level. And the demand for some other products currently sold must fall little bit even though the price is determined at a lower level.
This is the critical point for relative values of goods and services. We can put astronomical price tags for a Swiss Watch or an Iphone in this expansionary monetary system. If we had a constant money supply system, Apple would not determine Iphone’s price at the current level. And would not make huge profits.
P.S.
“You didn’t complete your example. What do you think P and Y are?
OK, that is a change in the real economy. So Y increases. (From what to what?)
Sven
Oct 4 2021 at 9:43am
P.S.
Iphone’s price let’s say would be 0.6 instead of $1 at T1 in constant money supply system. And with competition it would fall to 0.5, which means its cost, after some period of time.
Don Geddis
Oct 4 2021 at 3:51pm
(Sven: for some reason, I don’t see any Reply button on your latest comment, so I’ve had to reply to my own comment instead.)
That claim is simply not correct. “1000” = MV = PQ is a nominal quantity, not a real quantity. M is nominal, and P is nominal. The products are nominal GDP, not real GDP.
You never once said that, and that assumption changes everything. And it is not a reasonable assumption to make. Changes in M are completely arbitrary (a choice of the central bank), and there is no reason at all to assume that they will exactly match changes in Y. That’s not a realistic scenario (and is certainly not something that anyone could guess from your example).
You should have said that, originally. Again, that assumption changes everything.
No, that is false. Velocity V could change instead. In which case nominal GDP could be higher than 1000, even with a constant money supply.
Yes, that’s fine. If NGDP remains fixed at 1000, but real output Y increases, then indeed that implies that the price level P must decrease. We agree.
No, that’s completely wrong. You still have yet to show even a single relative price change. All you have shown is that the price level as a whole would decrease. That is an absolute price change, not a relative price change. With a lowering of the price level P, Apple’s costs would also lower. And (real) profits would remain unchanged (at a lower price level).
False. The price level P is easily approximated by a basket of goods; look at the BLS annual computations of inflation (e.g. CPI). Y can be reported by the real goods exchanged in all transactions. P and Y are easy to distinguish.
No, you have it wrong again. Measuring year to year changes gives you nominal GDP, not real GDP. You can only estimate real GDP, by taking measured NGDP, and then estimating likely inflation in order to discount NGDP to try to compute RGDP.
I don’t think you understand the difference between nominal and real (which is the key to all of macroeconomics and monetary policy).
Sven
Oct 6 2021 at 8:56am
First of all, My example is a hypothetical reasoning that is consistent with the real world dynamics. You should consider this when you read it.
“That claim is simply not correct. “1000” = MV = PQ is a nominal quantity, not a real quantity. M is nominal, and P is nominal. The products are nominal GDP, not real GDP.”
If inflation is zero nominal and real values are the same.
“You never once said that, and that assumption changes everything. And it is not a reasonable assumption to make. Changes in M are completely arbitrary (a choice of the central bank), and there is no reason at all to assume that they will exactly match changes in Y. That’s not a realistic scenario (and is certainly not something that anyone could guess from your example).”
I said before this is a blog post comment section I cannot write everything in my hypothesis here. I wrote a whole chapter and subsequent chapters to analyse the phenomenon.
What I tried to explain here how real values change as opposed to a productivity growth. So, I abstract inflation in my hypothetical reasoning. This is how it is done in economics profession when you try to explain things.
“No, that is false. Velocity V could change instead. In which case nominal GDP could be higher than 1000, even with a constant money supply.”
Yes, Velocity rises in the short term. That’s the way I explained it in my book. However, after velocity rises or in other words total expenditures increase, interest rate rises as well. And higher interest rate crowds out other expenditures in the second term. Subsequently, as a response the demand for Iphone goes down.
“No, that’s completely wrong. You still have yet to show even a single relative price change. All you have shown is that the price level as a whole would decrease. That is an absolute price change, not a relative price change. With a lowering of the price level P, Apple’s costs would also lower. And (real) profits would remain unchanged (at a lower price level).”
I don’t know how you got which I said the whole price level would decrease. I didn’t mean that at all. What I meant here is that in an expansionary monetary system Apple can set Iphone’s price at a high level compared to other products. This is the relative price divergence in economy. We call “normal profit” in economics. If the values of other products are determined at normal profit and Iphone’s value is determined at a high profit level, this would be a relative price change.
“False. The price level P is easily approximated by a basket of goods; look at the BLS annual computations of inflation (e.g. CPI). Y can be reported by the real goods exchanged in all transactions. P and Y are easy to distinguish.”
As you say it is an annual computation. What I meant here you cannot write P and Y separately in quantity equation. Because goods and services are not homogeneous. There is no one price. If you can do it, I would like to see it in a sample quantity equation.
“No, you have it wrong again. Measuring year to year changes gives you nominal GDP, not real GDP. You can only estimate real GDP, by taking measured NGDP, and then estimating likely inflation in order to discount NGDP to try to compute RGDP.
I don’t think you understand the difference between nominal and real (which is the key to all of macroeconomics and monetary policy).”
I wrote a book about how macro economy works. You claim that I don’t know the difference between nominal GDP and Real GDP. This not a good way of discussion.
Don Geddis
Oct 7 2021 at 1:57pm
Sven,
Unfortunately, this discussion is probably getting too long and complex to be appropriate for this comment thread.
You claim that your hypothetical example is realistic; I don’t believe you. You keep saying that you’ve written a book; perhaps true, but that doesn’t mean that your ideas are correct. Your book could simply be wrong.
I’m frustrated that every time you make a claim, and I show what is wrong with the claim, you then add an additional assumption that you hadn’t mentioned before, which changes the entire example. (E.g. “inflation is zero“, velocity falls, and now “interest rate rises as well“. All major new changes to your original example.)
You seem not to understand the difference between nominal and real. (Y is real. P*Y=NGDP is nominal.) You claim that you do, but you haven’t shown it in these comments. You’ve been sloppy about relative price changes vs. changes in the absolute price level. Even “inflation is zero” doesn’t actually make sense in your example. Since you hypothesize a new technology (iPhone), it’s impossible to precisely define “inflation” between T0 and T1.
I can’t see any easy way to make this conversation productive, in the limited space we have available here. Best of luck.
Sven
Oct 8 2021 at 10:10am
Book is here;
https://www.amazon.com/dp/B093RWX8FX
https://www.academia.edu/50822011/The_Theory_of_Capitalism
Whether you are interested or not is your decision.
Yes. I think the conversation should end here. No prospect for improvement.
Philo
Sep 30 2021 at 12:12pm
These concepts are trickier than they seem. With the neutrality of money, there will be no real effect at later time t1 if the money supply suddenly doubles at earlier time t0 and . . . [here we insert some sort of *ceteris paribus* clause]. Now, if this *ceteris paribus* clause is: “everything else *real* stays the same at t0,” the thesis of money neutrality is virtually trivial: there is no real change at t0, so why should there be any later real change? How about, instead, using simply: “everything else stays the same at t0”? No, that would be incoherent, for “everything” includes both real and nominal values, and they cannot both stay the same. For example, suppose that at t0 there is in place a progressive income-tax regime? When the money supply doubles and everyone’s nominal income doubles, is it real tax rates or nominal ones that stay the same (it cannot be both)?
I suspect the conceptual scheme you are working with is not sharp enough.
Scott Sumner
Sep 30 2021 at 10:26pm
These models are always an approximation of reality. It’s worth noting, however, that income tax rates are indexed to inflation.
David S
Sep 30 2021 at 12:59pm
“The short run can last for years” is something that Fed should have engraved on the wall of their meeting rooms. Unless I’m badly mistaken about current Fed policy, the “print your way back to prosperity” seems to be working, but the next 18 months are going to require some careful moves.
As you note in your book, Hume had this mostly figured out a long time ago.
Andrew_FL
Sep 30 2021 at 7:20pm
An expansion of money in the broader sense which is brought on innovations in payment systems and which facilitates gains from trade is “non neutral” in the sense that there are long term real effects of such a change, although these effects are very different from the business cycle effects of kind of “change in the quantity of money” we usually think of when thinking about whether “money is neutral”
On the other hand, if a loose monetary policy temporarily shifts the rate of interest below the natural rate, there will, temporarily, be irreversible capital investments in projects/ventures which are more “roundabout”-removed from final consumption-in the sense of Hick’s “Average Period”, or equivalently, of longer Macauley Duration, then such irreversible investments will need to be liquidated, and the resources used by them cannot be recovered, permanently reducing the long term growth rate of the economy, in the bust phase of the business cycle after interest rates have returned to their natural rates. Money is not neutral in the long run in this case, either, although it is in the sense that the effects are opposite from short run effects.
But in the age of MMT it may be “morally” true and correct to emphasize the more important point: you cannot make a society permanently richer just by printing money. In that sense, money is neutral in the long run.
Scott Sumner
Oct 1 2021 at 12:05pm
Investment shocks don’t change the long run growth rate, which is determined by technological change.
Andrew_FL
Oct 2 2021 at 10:45am
You believe this because Chicago has taught you capital as Frank Knight’s Crusonia Plant
Christophe Pella
Oct 25 2021 at 11:58am
Thank you for the clear reminder. What can we say of monetary policy and asset prices in the long run? I guess there’s no debate about the short-run effects of monetary policy, but is money also neutral for asset prices in the long run? Prima facie, Japanese equities do not seem to have benefitted from Japanese QE very much, yet some argue that monetary policy drive long-term asset prices.
Comments are closed.