Do higher interest rates really cause inflation?

In the comment section of a recent post, people reacted with shock to my claim that, holding the monetary base constant, higher interest rates are inflationary.  I wonder if they thought it was some kind of weird monetarist claim.  Actually I learned this from the Keynesians.

During the long (1950s-70s) debate over the relative effectiveness of fiscal and monetary policy, Keynesians used to criticize the Equation of Exchange as follows:

“Sure if you increase M, and if V is constant, then NGDP will rise.  But V is not constant.  Even worse for the monetarists, more M will tend to reduce V.  Here’s why.  As M rises, nominal interest rates will fall. At a lower nominal interest rate, there is a lower opportunity cost of holding cash.  This makes people hold on to cash longer, and V falls.  So it’s not even clear than more M will lead to more NGDP.  That’s why we need fiscal policy.”

Don’t blame me, blame the Keynesians.  I agree with their reasoning, except the part about needing fiscal policy.  The good news is that monetary policy has only a temporary effect on interest rates.  When they return to normal, so will velocity, and then the QTM holds in the long run.

So if Keynesian theory predicts that high interest rates are inflationary, why do the Keynesians talk as if the opposite is true?  Now we get back to “never reason from a price change.”  Here are some reasons why interest rates might rise:

1.  Tight money by the Fed (M falls)

2.  Fiscal stimulus pushes up interest rates as the government borrows more.

3.  Robust animal spirits among businessmen leads to more investment.

It’s no surprise that examples 2 and 3 are associated with higher NGDP and inflation, but what about number one?

In case one the effect of higher interest rates, by itself, is to raise V, which is clearly inflationary.  However in most cases V will rise by less than M falls.  Thus the overall effect of tight money is not inflationary.  It’s deflationary, because the deflationary effect of lower M overwhelms the inflationary effect of higher interest rates and faster velocity.

This example also shows that the quantity of money is what drives the transmission mechanism for monetary policy, which must be an embarrassment for Keynesians, who sometimes talk as if movements in the fed funds rate drive investment spending.

How could this problem be fixed?  The Fed could switch from OMOs to IOR as a policy tool.  In that case higher interest rates really would be deflationary, just as the Keynesians assume.  So you no longer need to put the money stock somewhere in the model.  Under our pre-2008 system the monetary base earned no interest.  Thus nominal rates became the opportunity cost of holding base money. Higher rates meant less demand for base money, which is inflationary.  With IOR it is just the reverse.  Higher IOR actually leads people to want to hold more base money, because it’s the reward for holding bank reserves.  So if you raise the IOR you increase the demand for reserves and base money, which is deflationary.

I’d guess this is why Michael Woodford is so gung-ho for switching to an IOR approach to monetary policy.  He abhors quantity theoretic models in much the same way a vampire abhors sunlight.  The IOR policy tool would allow Keynesians to claim that the Fed’s target interest rate isn’t just a policy tool; it actually explains the transmission mechanism for monetary policy.  Now they can claim that the quantity of money doesn’t matter (not quite, until we get rid of that pesky currency) and that it’s all about interest rates.

In contrast, I abhor interest rate models in the way a vampire abhors mirrors.  I want to eliminate reserve requirements, IOR, member bank deposits at the Fed, discount loans, and indeed anything that connects monetary policy to the financial system.  I’d like to return to the pre-1914 system where the base was 100% currency and coin.  Except I’d like to have the Fed make base money convertible into NGDP futures contracts, not 1/20.67 oz. of gold.


Tags:

 
 
 

35 Responses to “Do higher interest rates really cause inflation?”

  1. Gravatar of Suvy Suvy
    17. January 2013 at 08:57

    Milton Friedman says that there is a difference between the short term and long term. In the short term, printing money reduces the rate of interest while in the long term, it increases the rate of interest.

    I’m going to design a very simple model. We start with M, V, P, and Y where all are functions of time. We will say i is the nominal rate of interest, r is the real rate of interest.
    We know that M(t)*V(t)=P(t)*Y(t)
    We also know Fisher’s law that i=r+inflation=r+dP/dt

    If we suppose an economy is producing at full capacity(constant Y) and has a constant V, dM/dt=dP/dt. This will cause an increase in the rate of interest. Not only that, but when prices start to go up, this could very easily reduce the desire for people to hold cash(as they see their cash losing value). This means that V(t) could increase as well. Therefore, dP/dt increases which can cause an even further increase in the interest rate. This can create a positive feedback loop, so it usually his high inflation that causes high interest rates.

    However, it is important to note the role the banking system plays in high inflation. Right now, the borrowers are so indebted that there is no demand for credit, so the supply of credit is not increasing. This means the government can come in and print lots of money and have very little effect on inflation.

    Also, the model above is very simple and does not have any sort of asset markets in it. Increases in the money supply can increase the price of assets and their turnover and have no real impact on output and the price of goods and services. The changes increases in the money supply create in the asset markets should affect the rate of interest, but it is difficult to say exactly how much they will do so because we really do not know. The way I have defined inflation is by changes in the price level of goods and services, if we add assets, it is really difficult to understand how all of these things behave.

  2. Gravatar of Saturos Saturos
    17. January 2013 at 09:07

    Scott, you’re being disingenuous. You know that Keynesians read the ISLM model by saying that investment expenditures are a function of the interest rate, that “real demand” for housing and such will rise as interest rates fall. At the same time higher rates raise the opportunity cost of holding money, and the demand for liquidity as a function of both interest rates and income must equal the supply. Hence the “general equilibrium” of the ISLM model. All of which you know perfectly well, of course. Keynesians reading your post should be no more supportive of the quantity theory of money than before.

    Nice comment on Woodford, though. But surely the quantity of reserves still matters even without currency, indeed it matters even more, as you pointed out to me earlier.

  3. Gravatar of Becky Hargrove Becky Hargrove
    17. January 2013 at 09:17

    The primary reason I agree with you about the financial system is the fact that lending institutions have become somewhat hypocritical about their actual intentions. Whereas NGDPLT encourages incremental economic activity and stability, loans do nothing of the sort, as one is encouraged to either borrow far more than they need or nothing at all. The reality is that anyone of low income needs to reach for sustainability a step at a time and often must do so by economic means that appear “random” by anyone with higher income.

    However, I need a bit more clarification as to fractional reserve lending not being an actual component of NDGP growth. To me that is monumental, but…? How does one actually prove that? As long as anyone believes loans are wealth creation, how could they be convinced that higher interest rates could cause inflation.

  4. Gravatar of TravisV TravisV
    17. January 2013 at 09:21

    Prof. Sumner,

    Yglesias (a hero of mine) advocates that Japan should pursue BOTH monetary stimulus AND fiscal stimulus here:

    http://slate.me/X8qudj

    Why do you think Yglesias believes that fiscal stimulus is necessary? I’m sure he understands the Sumner critique. So why not advocate monetary stimulus and say that the fiscal side is irrelevant?

    Does he think that Japan has serious structural or supply-side problems that tax cuts could help alleviate? Or is it something else? I’m having a hard time understanding.

  5. Gravatar of Suvy Suvy
    17. January 2013 at 09:57

    “Scott, you’re being disingenuous. You know that Keynesians read the ISLM model by saying that investment expenditures are a function of the interest rate, that “real demand” for housing and such will rise as interest rates fall. At the same time higher rates raise the opportunity cost of holding money, and the demand for liquidity as a function of both interest rates and income must equal the supply. Hence the “general equilibrium” of the ISLM model. All of which you know perfectly well, of course. Keynesians reading your post should be no more supportive of the quantity theory of money than before.”

    Too bad this is not what Keynes said at all. These “Keynesians” really need to reevaluate everything they say/believe/propose. They don’t recognize that the supply of credit has very little to do with the amount of credit actually issued, it cannot be practically done. They completely ignore the role of the financial sector and its role in financing investment.
    http://esepuba.files.wordpress.com/2011/10/keynes-the-ex-ante-theory-of-the-rate-of-interest.pdf

    Reducing interest rates also reduces the prospective yield from capital assets. This could actually discourage investment instead of increasing it.

  6. Gravatar of Ritwik Ritwik
    17. January 2013 at 10:05

    Holding M constant and increasing V = pushing IS curve upwards. Fiscal policy pushes up the rate of interest. Monetary policy raising interest rates would be pushing LM curve inwards. The two are not the same.

    C’mon Scott. To criticize the model, at least understand it fully first. Yes, let’s not reason from a price change. The IS/LM doesn’t.

  7. Gravatar of Peter Peter
    17. January 2013 at 10:10

    “Now they can claim that the quantity of money doesn’t matter (not quite, until we get rid of that pesky currency) and that it’s all about interest rates.”

    So how negative would the IOR have to be to get, say, 10 trillion in NGDP from 1 single unit of base money (with zero currency)? Do they really think this is possible? And that it would be a good system?

  8. Gravatar of Arthur Arthur
    17. January 2013 at 10:15

    “Except I’d like to have the Fed make base money convertible into NGDP futures contracts, not 1/20.67 oz. of gold.”

    I feel this sentence is extremely important for reasons yet unknown to me.

  9. Gravatar of marcus nunes marcus nunes
    17. January 2013 at 10:17

    A story where interest rates don´t make an appearance!
    http://thefaintofheart.wordpress.com/2011/09/15/an-alternative-script-%E2%80%93-one-where-interest-rates-don%C2%B4t-make-an-%E2%80%9Cappearance%E2%80%9D/

  10. Gravatar of Ron Ronson Ron Ronson
    17. January 2013 at 10:52

    I am skeptical about this.

    Higher interest rates will indeed cause people to be less inclined to hold balances in cash but will that not be factored into the original supply of loanable fund curve ?

    Even if its true why is it a problem in a world with an active CB?

    In effect I think it is being said that higher interest rates will cause the demand for money to decrease (ie the curve to shift not just a move along the curve). This would be inflationary but with either IT or NGDPT in place the CB would adjust monetary policy to decrease the money supply in response to this change in demand and prevent inflation and maintain monetary equilibrium.

  11. Gravatar of Geoff Geoff
    17. January 2013 at 11:04

    Maybe it’s just me, but I think the causation is the other way around.

    Instead of higher interest rates being inflationary (which I guess means if interest rates fell, then inflation would fall), I think the correct thing to say is that higher inflation is what makes higher interest rates…but not always. Not always because real rates could rise.

  12. Gravatar of Andrew Andrew
    17. January 2013 at 11:47

    Most of us were taught IS-LM by saying that output is determined by the Keynesian cross (which creates the IS equation), and interest rates are determined in the money market (the LM equation).

    But the way of thinking Scott prefers is to say that output (nominal at least) is determined by the LM equation (holding interest rates constant) while the interest rate is determined in the loanable funds market holding Y constant (the IS equation). If you map the LM equation in M-Y space, you get an increasing function between money and output (L(Y,r1)=M). Then draw the same function but with a higher interest rate (L(Y,r2)=M), and you find that outputs higher for a given M. So thus a higher interest rate is “inflationary.”

    When you describe the mechanisms in words, they sound like different models, but they’re not. They’re both IS-LM.

  13. Gravatar of Becky Hargrove Becky Hargrove
    17. January 2013 at 11:57

    The only thing IOR does for transmission is completely strip out the gears. There’s got to be a better way than surrender and why should any Keynesian just give up, roll over and raise that flag. Apparent gain with IOR transmission in the short run is nothing but total loss in the long run when people get too scared of debt. Best not to go too far down the path where the arbitrary and destructive nature of IOR is completely forgotten.

  14. Gravatar of Doug M Doug M
    17. January 2013 at 12:23

    “Fiscal stimulus pushes up interest rates as the government borrows more.”

    Evidence seems to show the opposite. When we had a surpluss rates were high. Now that we have record deficts, rates are at an all time low.

    However, the government is not the only one who borrows money. We can look at the Treasury / Corporate spread, and see that currently has no borrowing advantage over high quality corporations. Typically, they Tsy is 0.50% percent more expensive. in 2000, Treasuries were 1.40% more expensive than AA corporates.

  15. Gravatar of Doug M Doug M
    17. January 2013 at 12:44

    “higher rates raise the opportunity cost of holding money, and the demand for liquidity as a function of both interest rates and income must equal the supply.”

    I understand that this is straight from Keynes, but it is wrong nonetheless. Cash pays interest. What could be said is that a steep curve (high interest rate differential) increases the cost of holding money.

  16. Gravatar of Felipe Felipe
    17. January 2013 at 14:07

    Scott, you are cheating. An active central bank cannot raise the interest rate above expectations while at the same time maintaining the monetary base at expectations. And this is the only way that higher rates could cause (negative) inflation. The other two are inflation causing a rise in the interest rate (hmm, maybe I am from the concrete steppes).

    In other news, Martin Feldstein thinks monetary stimulus in Japan is folly.

    http://www.project-syndicate.org/commentary/the-wrong-growth-strategy-for-japan-by-martin-feldstein

  17. Gravatar of ssumner ssumner
    17. January 2013 at 14:45

    Suvy, Yes, higher inflation causes higher interest rates. But then the higher interest rates caused by higher inflation cause still higher inflation. Or to be more specific, a further one time increase in the price level.

    Saturos, No I’m not being disingenuous, and your comment is actually incorrect. Yes, an increase in RGDP does increase the real demand for currency, and that does absorb some of the excess currency. But not enough. All it does it make it so that prices rise by less than NGDP.

    That is why you need to look at velocity, which is not subject to your criticism. So let’s say you have M go up 10%, and V go down 4%. Then NGDP rises 6%. because RGDP rises, inflation is less than 6%. But it still rises, this is a demand shock after all. Say RGDP and P both rise by 3%. Then M will have caused 10% inflation, RGDP caused negative 3% inflation, and lower interest rates caused negative 4% inflation (via lower V). QED.

    And suppose I had argued that higher interest rates raise NGDP instead of inflation, don’t you think people would have been equally shocked? So the real income change here doesn’t explain the paradox.

    My goal wasn’t to make Keynesians support the QTM–I’m not sure I do (at least in the simplest form–stable V.) Indeed I said I agreed with Keynesians that V isn’t stable.

    Woodford doesn’t think the quantity of reserves matters, just the interest rate.

    Becky, I’m not saying loans aren’t important, I just don’t want them mixed up with monetary policy.

    Travis, I think he just wants to use all tools available. If the BOJ was super aggressive he’d probably back off from favoring fiscal stimulus. But they aren’t.

    Doug, I hope you aren’t serious about the “evidence” that deficits lower interest rates. We run surprluse sduring booms and deficits during recessions. The business cycle effec tcompletely overwhelms the fiscal policy effect. I’m sure hospital admissions are correlated with death, but I doubt hospitals cause death.

    Ritwik, You might want to reread my post–I think it went in one ear and out the other. Yes, I do understand the distinction between interest rate increases caused by tight money and those caused by fiscal stimulus.

    Felipe, You said;

    “Scott, you are cheating. An active central bank cannot raise the interest rate above expectations while at the same time maintaining the monetary base at expectations.”

    I never ssaid they could.

    Everyone. Lots of people aren’t taking the time to read the post carefully, and think about what it means. I’m not saying that a tight money policy by the Fed that raises interest rates will cause higher inflation. Just the opposite, it will cause lower inflation.

    My post is completely consistent with IS-LM, which also predicts that tight money will cause lower inflation. There is nothing at all controversial here–Keynesians agree that higher interest rates raise velocity. Keynesian used to criticize Milton Friedman for forgetting that higher interest rates raise V. I make the same claim and people act like I don’t know IS-LM!

    The floggings (I mean posts) will continue until understanding sets in.

  18. Gravatar of dtoh dtoh
    17. January 2013 at 15:02

    Scott,
    “As M rises, nominal interest rates will fall. At a lower nominal interest rate, there is a lower opportunity cost of holding cash. This makes people hold on to cash longer, and V falls.”

    At normal interest rates, people hold the amount of cash they need for transactions. The interest rate has very little impact on V (only the shoe leather cost). At the ZLB, MB becomes dual purpose and is used as a financial asset. When it’s a financial asset, a drop in rates raises the real price and there is an increased exchange of financial assets into real goods and services ergo higher AD.

  19. Gravatar of No school of thought to belong to? | Catalyst of Growth No school of thought to belong to? | Catalyst of Growth
    17. January 2013 at 15:13

    […] Update: Scott continues to bury himself into incorrect monetary mechanics: Under our pre-2008 system the monetary base earned no interest. […]

  20. Gravatar of DOB DOB
    17. January 2013 at 15:26

    Felipe is right on the money. I’ve updated my post.

    { Interest Rate Target, IOR, Monetary Base }: pick two, the market gets to set the third! Scott is implying that the central bank moves the IR up, IOR stays at 0%, and the monetary base stays the same! Higher rates means less demand for base money, which just means there is less base money.. M goes down, V goes up by roughly the same proportion leaving NGDP just about unaffected

    Scott, you’re now saying that this isn’t what you meant, but then what was the purpose of confusing everyone?

    Keynesian used to criticize Milton Friedman for forgetting that higher interest rates raise V. I make the same claim and people act like I don’t know IS-LM!

    Sorry but you didn’t write that “higher interest rates raise V”, you wrote:

    Let me repeat, higher interest rates are inflationary. [..] Higher interest rates are inflationary. Repeat 100 times.

  21. Gravatar of Eliezer Yudkowsky Eliezer Yudkowsky
    17. January 2013 at 15:29

    *Cough.* Hospitals actually cause the **** out of death. Not saying I dispute the overall point but it’s a terrible metaphor considering iatrogenic fatality figures.

  22. Gravatar of Shaun Peterson Shaun Peterson
    17. January 2013 at 15:38

    I once read a great post by Michael Pettis that made the exact same point about Chinese savings rates. He noticed as rates rose, savings decreased, therefore pushing up spending, economic activity, and inflation. Kind of like the Paradox of Thrift principle, but what happens when it reverses.

    Good post.

  23. Gravatar of Doug M Doug M
    17. January 2013 at 15:58

    Hospitals do cause death!… and not just because most Americans die in hospitals. Adverse reactions to medications and hospital related infection are the two big iatrogenic factors.

    Regarding deficits and interest rates…Yes, I am having a little bit of fun with you, but I have seen equally high quality analysis presented in this blog.

    I actually did have a point. Treasury borrowing about 1/3 of total borrowing in the US. Corporate borrowing, and consumer borrowing (mostly mortgages debt) are also close to 1/3 each. There is some debate as to what degree Government borrowing crowds out other forms of borrowing, it is safe to say that a moderate increase in Government borrowing, leads to only a small increase in total borrowing.

  24. Gravatar of Geoff Geoff
    17. January 2013 at 18:21

    Dr. Sumner:

    “Suvy, Yes, higher inflation causes higher interest rates. But then the higher interest rates caused by higher inflation cause still higher inflation. Or to be more specific, a further one time increase in the price level.”

    Since this comment essentially answered my questions, I say thank you.

    I don’t understand how higher interest rates have their own independent force on GENERAL prices.

    I know what you’re thinking though, that with higher interest rates investors would invest more in capital and labor and so on, and less in money itself (as would occur more often with lower interest rates, as money and loans become more and more similar), but then the initial stage higher or lower interest rates, since they are a function of inflation, would simply be rolled up with any subsequent inflation, in that inflation begets inflation, not inflation begets higher interest rates which begets more inflation.

    Imagine inflation rates are stable, and then imagine that without any change in the inflation rate, REAL interest rates rise or fall. Rates on loans would then change. Yet, according to the theory that interest rates themselves affect inflation, there should be a rise in prices with higher real interest rates, and a decrease in prices with lower real interest rates.

    This is a “test”, I think, of whether interest rates by themselves can affect general price levels. Do you agree with this?

  25. Gravatar of Justin Merrill Justin Merrill
    17. January 2013 at 20:44

    Scott, are you familiar with the work of Knut Wicksell?

    High nominal interest rates are not the cause of inflation in themselves; a steep yield curve leads to high NIMs by banks, leads to increased bank credit, leads to higher inside money supply, leads to inflation.

  26. Gravatar of Felipe Felipe
    18. January 2013 at 05:12

    Scott:
    “I never ssaid they could.”

    Which is why you are cheating ;). Lets go back to your original claim:

    Suppose a central bank sees inflation rising too fast. They might decide to raise interest rates. But higher interest rates don’t lower inflation, they raise inflation. That’s because as interest rates rise, velocity rises. Higher velocity raises inflation, it doesn’t lower it.

    And then you post this where you say you meant assuming the monetary base is constant. No surprise people got confused, since (as you agreed) the central bank cannot raise interest rates while keeping the base constant.

    Just to clarify, I do agree with you. I just think you were too confusing.

  27. Gravatar of ssumner ssumner
    18. January 2013 at 07:39

    Felipe, I love it when cheaters accuse me of cheating. This is what I said:

    “Let me repeat, higher interest rates are inflationary. They increase velocity. If you don’t believe me, check out interest rates and velocity during any extremely high inflation episode. When rates rise, inflation usually rises. Higher interest rates are inflationary. Repeat 100 times. But the thing the Fed uses to generate higher short term interest rates””a reduction in the money growth rate””is deflationary. To reduce inflation you must reduce money growth. Interest rate targeting (or exchange rate targeting) is simply a tool. It is changes in the money growth rate that actually drive macro nominal
    aggregates up and down.”

    Funny that you just quoted part of my post. 🙂

  28. Gravatar of ssumner ssumner
    18. January 2013 at 07:46

    dtoh, Many studies show that you are wrong. I’ll do another post.

    DOB, Higher interest rates are inflationary, holding M constant. I stand by that claim.

    In the cases where the higher rates are caused by tight money, the deflationary impact of the tight money overwhelms the inflationary impact of the higher interest rates.

    Eliezer and Doug, Even as I typed that I sort of knew I wouldn’t get away with it.

    Justin, Yes.

  29. Gravatar of Floccina Floccina
    18. January 2013 at 09:45

    That is a bit scary.

    I’d like to return to the pre-1914 system where the base was 100% currency and coin.

    I totally agree but also allow banks to print their own currency.

  30. Gravatar of Bob Murphy Bob Murphy
    18. January 2013 at 21:43

    I’d like to return to the pre-1914 system where the base was 100% currency and coin. Except I’d like to have the Fed make base money convertible into NGDP futures contracts, not 1/20.67 oz. of gold.

    I’m with you halfway, Scott.

  31. Gravatar of dave dave
    22. January 2013 at 23:44

    I am confused with the statement above. Does higher interest rate cause inflation lower? if central bank wants to increase the interest rate, it is only because the expected inflation tends to be high in the future. Until the inflation is realized, the interest rate would be decreased if the ex-post inflation is lower than the expected inflation. That is to say, the higher interest rate is fighting with the expectation of the high inflation. It did not say that the high interest rate goes with a high inflation. basically, the high interest rate only means the demand for money is lower, spend less, low inflation. While, maybe I am wrong.

  32. Gravatar of Höhere Preise durch höhere Zinsen? « Aus dem Hollerbusch Höhere Preise durch höhere Zinsen? « Aus dem Hollerbusch
    23. January 2013 at 08:27

    […] was ihr zumindest kurzfristig auch gelingen wird, so daß der Gesamteffekt unbestimmt bleibt. (Scott Sumner, den ich hier paraphrasiere, meint, daß in der Regel der deflationäre Effekt der Geldmengenreduktion überwiegen […]

  33. Gravatar of flow5 flow5
    22. March 2015 at 10:54

    High rates are both a cause and an effect of high rates of Vt.

  34. Gravatar of Tella Oluwatoba Ibrahim Tella Oluwatoba Ibrahim
    24. January 2018 at 12:51

    A misleading relationship between money supply, interest rate and velocity of money.

  35. Gravatar of Tella Oluwatoba Ibrahim Tella Oluwatoba Ibrahim
    24. January 2018 at 12:53

    A misleading link of the variables (money supply, interest rate and velocity of money).

Leave a Reply