Why are higher interest rates inflationary? (#3b)

In the previous post I pointed out that higher interest rates are inflationary, as they raise velocity.  And yet I forget that just as most people wrongly think economic growth is inflationary, most people wrongly think higher interest rates are contractionary. And they are making the exact same mistake, not holding M constant when they make their observation.

In the short run, a Fed decision to raise rates is indeed often contractionary.  But that’s because they typically raise rates by cutting the money supply, which is more contractionary than the higher interest rates are expansionary.

Don’t believe me? Then how about that famous Keynesian Larry Summers, which established this proposition in a paper he did with Robert Barsky.  They found that under the gold standard, higher interest rates led to booms and rising prices, and lower interest rates led to recessions and deflation.  Before 1913 the Fed didn’t even exist, so movements in interest rates impacted the demand for gold, which impacted NGDP growth.

One commenter asked about the likely September increase in interest on reserves, which is supposed to be contractionary. And it will be.  My claim was that a rise in market interest rates is expansionary, as this makes it more costly to hold base money (in an opportunity cost sense.)  As the demand for base money falls, the price level and NGDP rise.  But a rise in IOR is very different, it increases the demand for base money, which is contractionary.

The key variable here is the opportunity cost of holding base money, which is the market interest rate minus the IOR.  For some insane reason on October 8, 2008 the Fed made that gap negative for the first time since 1940, and we all know what happened next.

PS. For those who can read French, here’s a new article in Atlantico where I am interviewed on the subject of China.


Tags:

 
 
 

51 Responses to “Why are higher interest rates inflationary? (#3b)”

  1. Gravatar of Nick Bradley Nick Bradley
    13. August 2015 at 06:04

    Pretty Spot on here Scott, thanks.

    I think too many people confuse market rates with fed money supply decisions

  2. Gravatar of Brian Donohue Brian Donohue
    13. August 2015 at 06:36

    I heard a good blurb from you on NPR about China last night. INFILTRATOR!

    Also, really enjoying this series.

  3. Gravatar of Doug M Doug M
    13. August 2015 at 07:35

    Interest rates are a price. It is the price of tomorrows dollar today. When the Fed watchers talk about an interest rate target, the Fed changing the money supply until this price reaches its target price.

    When we talk about higher rates being contractionary, it is because M falling. So, when you talk about higher rates holding M constant, what you say isn’t wrong, but it goes against experience and the popular discussion because M is never held constant.

  4. Gravatar of TallDave TallDave
    13. August 2015 at 08:51

    That’s a really fascinating result, and it’s interesting that the Gibson Paradox was well-known in the 1930s.

  5. Gravatar of Tom Brown Tom Brown
    13. August 2015 at 08:51

    Scott, do you speak French?

  6. Gravatar of Tom Brown Tom Brown
    13. August 2015 at 08:52

    Brian, which NPR show was that?

  7. Gravatar of Ray Lopez Ray Lopez
    13. August 2015 at 08:55

    Pied Piper pipes on, a naked emperor deceiving all but schoolchildren.

    Sumner does not define “IOR” (and this is supposed to be a tutorial). Nor does he address that there’s no strong evidence for QTM in econometerics (Bernanke, 2003 finds between 3.2% to 13.2% out of 100%, i.e., nearly nothing). Further, as Philippe correctly pointed out before being cowed into submission by the online bullies here, QTM gets causation backwards (money supply follows what the economic actors do, not the other way around). Sumner further has this gem of a metaphysical quote (in response to Philippe in another thread): “QTM says that money supply increases will be inflationary if exogenous, even if done for many different reasons.”. Reasons are not the problem Sumner. The problem is: who labels a money supply increase as “exogenous” or ‘endogenous’? Do you do this ex ante or ex post? If ex post, your engaging in pseudo-science, as your hypothesis is not falsifiable.

    J’accuse you Sumner, of being a witch doctor. How do you say that in French? Je t’accuse d’être un sorcier-docteur

  8. Gravatar of Tom Brown Tom Brown
    13. August 2015 at 09:04

    Ray, IOR = interest on reserves, paid by the Fed to (some?) Federal reserve deposit holders (such as commercial banks).

  9. Gravatar of Kailer Kailer
    13. August 2015 at 09:16

    “My claim was that a rise in market interest rates is expansionary, as this makes it more costly to hold base money (in an opportunity cost sense.)”

    I thought you weren’t supposed to reason from a price change!

  10. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    13. August 2015 at 09:17

    ‘Interest rates are a price.’

    Well, THEY are priceS.

    ‘It is the price of tomorrows dollar today.’

    No, it’s just the rental price of today’s dollar. What a borrower has to pay to USE someone else’s dollars temporarily.

    The price of tomorrow’s dollar will be the goods and services that will be exchanged for the dollars.

    ‘When the Fed watchers talk about an interest rate target, the Fed changing the money supply until this price reaches its target price.’

    That may be what they think they’re doing. But it’s bad economic logic.

  11. Gravatar of Market Fiscalist Market Fiscalist
    13. August 2015 at 09:18

    I think what is being said is that (ignoring any central bank interventions) a rise in market interest rate is reflective of people wanting to borrow more and lend less, and these things will probably lead to inflation.

    That not quite the same thing IMO as saying that higher interest rates are inflationary. If a bank arbitrarily raised interest rates and nothing else changed then people would borrow less and save more and this would be deflationary.

    So the fact that higher market interest rates and higher inflation are correlated does not really mean that one causes the other, as is implied in the title of the post.

  12. Gravatar of Tom Brown Tom Brown
    13. August 2015 at 09:25

    Ray, I’ll be impressed if you can somehow work

    “C’est le sang de Danton qui t’étouffe!”

    into your criticism (it’s the one thing I remember from French). Feel free to replace “Danton” with something more relevant.

  13. Gravatar of E. Harding E. Harding
    13. August 2015 at 09:58

    “They found that under the gold standard, higher interest rates led to booms and rising prices, and lower interest rates led to recessions and deflation.”
    -That smells a heck of a lot like reasoning from an interest rate change to my laymanish nose.
    “But a rise in IOR is very different, it increases the demand for base money, which is contractionary.”
    -I believe raising interest on reserves is, in the long run, inflationary, as it improves banks’ capital positions, making them more willing to lend. Imagine, for example, that the Fed raised IOR to 120%. I’m pretty sure Neo-Fisherism would be true in that situation.

  14. Gravatar of Scott Sumner Scott Sumner
    13. August 2015 at 10:02

    Doug, Of course M is sometimes held constant. It was held constant from August 2007 to May 2008, while interest rates declined.

    You are right that the public (and much of the economics profession) is hopelessly confused, which is why they misdiagnosed 2008 as easy money.

    Tom, “Non” (Seriously, I took three years in school, but that was 45 years ago.)

    Ray, You said:

    “Sumner does not define “IOR””

    Or did you not read the post carefully?

    Kailer, Interest rates aren’t the price of money, 1/P is the price of money. Interest rates are a variable that shifts the demand curve for money. It’s normal to say a higher price of natural gas will, ceteris paribus, raise the demand for coal.

    But in a sense I agree with you. I’m saying you can’t tell whether higher rates are expansionary or contractionary unless you know whether they were caused by a change in the money supply, or some other factor.

    Market, Nope, I’m claiming a causal relationship, as higher rates reduce the demand for base money, causing inflation.

  15. Gravatar of Scott Sumner Scott Sumner
    13. August 2015 at 10:04

    E. Harding, That would produce another Great Depression, if that’s all they did.

  16. Gravatar of Market Fiscalist Market Fiscalist
    13. August 2015 at 10:08

    ‘Nope, I’m claiming a causal relationship, as higher rates reduce the demand for base money, causing inflation.’

    Wouldn’t that mean that the Neo-Fisherians are right ?

  17. Gravatar of E. Harding E. Harding
    13. August 2015 at 10:17

    @Scott Sumner
    -I’m sure that, at first, it would create a giant bank lending freeze, as banks stopped lending. But, as the rate of base money growth skyrockets to 100%, and depositors get these fantastic rates of return on their deposits, after the initial decline in NGDP (not sure if to the extent of the Great Depression) due to the increased money demand, Neo-Fisherism would start to become true as people spent their deposits, and NGDP growth would eventually skyrocket to over 100%.

  18. Gravatar of Scott Sumner Scott Sumner
    13. August 2015 at 10:18

    Market, I’ve always argued they are right in some cases and not others. The problem I see is that they don’t seem to understnad that distinction. If the Fed raises rates by reducing the monetary base, the total effect is contractionary, not expansionary.

  19. Gravatar of Market Fiscalist Market Fiscalist
    13. August 2015 at 10:20

    I’m still trying to understand:

    “My claim was that a rise in market interest rates is expansionary, as this makes it more costly to hold base money (in an opportunity cost sense.) As the demand for base money falls, the price level and NGDP rise.”

    is the logic this:

    – people hold a stock of bank notes
    – interest rates rise
    – they take some of this stock and deposit it in bank
    – it gets lent out, spent and drives up the price level ?

    Who is going to borrow it though, without driving rates back down again ?

  20. Gravatar of Scott Sumner Scott Sumner
    13. August 2015 at 10:20

    E. Harding, In that case sure, but I was holding the base constant.

  21. Gravatar of Philippe Philippe
    13. August 2015 at 10:23

    Scott,

    in your previous reply to me you said: “What matters is the gap between market rates and IOR.”

    Would it be right to say that whether interest rates are inflationary or not depends on the difference between market rates and the interest rates set by the Fed?

    i.e. high market interest rates are inflationary if they are much higher than the Fed’s interest rate.

  22. Gravatar of Philippe Philippe
    13. August 2015 at 10:45

    E.Harding

    “But, as the rate of base money growth skyrockets to 100%, and depositors get these fantastic rates of return on their deposits”

    IOR doesn’t increase the size of the monetary base, unless the central bank is ‘insolvent’. Normally, the interest on reserves is paid for by government borrowing, i.e. bond sales.

  23. Gravatar of Brian Donohue Brian Donohue
    13. August 2015 at 10:58

    @Tom Brown, Marketplace was the NPR show, I think.

  24. Gravatar of E. Harding E. Harding
    13. August 2015 at 11:55

    “IOR doesn’t increase the size of the monetary base, unless the central bank is ‘insolvent’. Normally, the interest on reserves is paid for by government borrowing, i.e. bond sales.”
    -Cite?

  25. Gravatar of Philippe Philippe
    13. August 2015 at 12:21

    E.Harding

    the central bank pays interest on reserves out of the income it receives from the assets it owns. On net, this means the interest is paid out of income it receives from government bonds.

  26. Gravatar of Doug M Doug M
    13. August 2015 at 12:21

    Patrick Sullivan.

    Sorry, the “price of tomorrows dollar today” is the discount rate 1/(1+y).

    The rental cost of today’s dollar is priced in tomorrow’s dollar. It really is the same thing.

  27. Gravatar of Doug M Doug M
    13. August 2015 at 12:28

    Philippe,
    “Would it be right to say that whether interest rates are inflationary or not depends on the difference between market rates and the interest rates set by the Fed?”

    I think that this is hugely important and poorly understood. Economists throw around the term “interest rates” but seldom define which rate they are talking about. The spreads between these rates dries it all.

  28. Gravatar of LC LC
    13. August 2015 at 13:22

    Wow, I was on the road for couple of days and now there’s an emerging market crisis with a 3% devaluation of RMB? Shocking! I just saw Gordon Chang on Bloomberg talking about imminent crash in China. Bloomberg headlines also have terms like rout in RMB. I sure missed all that hype in China.

    BTW, the Tianjin blast is a real tragedy, but reflects the poor mid-level management practices and lack of basic low and mid-level management talent in China. That’s the biggest drawback to Chinese development I saw.

  29. Gravatar of AlexR AlexR
    13. August 2015 at 14:12

    Kudos. The past three days’ posts are the best pieces on macro I think I’ve ever read. Finally, macro analysis that makes sense. My only objection is that the first post on this topic made a glancing reference to AS/AD analysis. There is no need to ever revert to that conceptual monstrosity given your clearer explanations using the QTM.

  30. Gravatar of Philippe Philippe
    13. August 2015 at 14:15

    Major_Fiefdom is strangely absent.

  31. Gravatar of Benny Lava Benny Lava
    13. August 2015 at 14:47

    These are thoughtful posts with some thoughtful critiques in the comments.

  32. Gravatar of Jason Smith Jason Smith
    13. August 2015 at 15:11

    While I agree that under certain conditions rising interest rates and inflation go hand in hand, I’m not sure the source of rising interest rates and rising price level are related. The rates appear to rise due to the “other stuff” and the price level goes up due mostly to monetary expansion …

    http://informationtransfereconomics.blogspot.com/2015/08/are-higher-interest-rates-inflationary.html

  33. Gravatar of Andrew_FL Andrew_FL
    13. August 2015 at 16:30

    The effect you’re talking about basically works through the demand side of the loanable funds market, right?

    When people talk about a Wicksell effect, as opposed to the effect you’re talking about here, they usually have in mind interest rates that are lowered relative to what they would otherwise be by injecting new money into loan markets on the supply side-and raising them by doing the opposite. Such an effect of course presumes that M is increased (decreased).

    Would you contend that the demand side effect on interest rates is more important, or plays the larger role in interest rate movements, most of the time?

  34. Gravatar of emerich emerich
    13. August 2015 at 19:30

    In the Atlantico interview you clearly made many points you’ve been making on the blog, e.g., that devaluation is not a negative for global growth because devaluation is an expression of monetary expansion, leading to growth that benefits the neighbors of the devaluing country. But in your last comment you seemed to be saying that a “currency war” would reprise the currency wars of the 1930s, with the implication that these were central to the Great Depression. But that seems inconsistent with your earlier points. Did I misunderstand your final answer?

  35. Gravatar of Ray Lopez Ray Lopez
    13. August 2015 at 19:49

    A lot of confusion would be avoided if you simply understood–as the data suggests–the following model:

    –money is neutral, except, as Rowe says, during hyperinflation (and btw Brazil with high teens inflation for 40 years post WWII did not have ‘hyperinflation’)

    –when the money supply is fixed, as in the gold standard of the 19th century, prices are stable (in the long run, defined as over 10 years), hence a growing population and/or economy means deflation

    –when the money supply is not fixed, as in fiat currencies, prices rise in the long run. Money is neutral however and neither in the short term nor in the long term does the central bank have any real effect on the economy, except in a trivial manner (Bernanke et al (2003) suggests 3.2% to 13.2% out of 100% effect)

    Simple and the data supports this model.

    It’s like Copernicus and the heliocentric model of the Renaissance: if you try and use Ptolemy’s model, you have to do fancy stuff like epicycles to make the model work, analogous to Sumner’s undefined ‘exogenous’ ex post fudging; if you accept money neutrality, everything is simple, like the earth going around the sun. Occam’s Razor.

  36. Gravatar of David de los Angeles Buendia David de los Angeles Buendia
    13. August 2015 at 20:48

    Dr. Sumner,

    You linked a paper discussing Gibson’s Paradox where higher interest rates are associated with periods of economic growth and lower interest rates with periods of economic contraction during a two hundred year period dominated by a gold standard. The same relation holds today under a fiat money system.

    If one examines the relationship between the Effective Federal Funds Rate (EFFR) and Gross Fixed Capital Formation in United States© (CFR – USAGFCFQDSMEI) there is a strong and statistically significant relationship. Between the 1st Quarter 1956 and the 4th Quarter 2014 there were 236 Quarters. The mean Year over Year Percent Change (Y/Y%C) in the quarterly EFFR was 6.5 % (Std Dev = 47) and the mean Y/Y%C for CFR was 6.4% (Std Dev. = 5.8). The Median values for each respectively were 3.8% and 6.6%. The Pearson Product Moment Correlation (PPMC) was 0.55 p<0.0001). So there is a very strong, statistically significant correlation between the change in value of EFFR and CFR[1].

    A similar relationship exists between the EFFR and total debt. Between the 2nd Quarter of 1954 and the 1st Quarter of 2015, there was a weak but statistically significant correlation between the year over year percent change in the EFFR and the year over year percent change in the total debt in the United States economy. The Pearson Product Moment Correlation was ER = 0.213 P = 0.0009 (N = 239). When the EFFR rate increased, total debt increased approximately at the same time and same direction some of the time.

    Now what does this mean? Do business create and expand capital in response to rising interest rates? More likely it is that interest rates rise as demand for credit increases because of capital formation, as business expands they need more credit which drives up interest rates. Interest rates are at an all time low because of all time low demand for credit.

    [1] http://capformeffr.tumblr.com/

  37. Gravatar of ssumner ssumner
    14. August 2015 at 05:36

    Philippe, It depends what you mean by “interest rates set by the Fed.” If you mean fed funds rate, then no, those are market rates. If you mean IOR then yes.

    LC, I agree.

    Thanks Alex.

    Jason, I’m saying higher rates make inflation rise by more than you’d predict by monetary expansion alone.

    Andrew, Not sure I follow your question, but I do think that rising interest rates usually lead inflation to be higher than you’d expect from monetary policy alone.

    emerich, I believe that devaluations helped countries recover from the Great Depression.

    Ray, Why have prices fallen in Japan despite large increases in the (fiat) money supply?

    David, I agree.

  38. Gravatar of Market Fiscalist Market Fiscalist
    14. August 2015 at 06:14

    I still don’t see an answer to the question posed in the blog title “Why are higher interest rates inflationary?”

    Assume interest rates rise but M stays constant.

    If I’m a net saver:
    – I spend less as I will be able to spend more in the future if I save more out of income at the higher rates.
    – I hold smaller cash balances as the cost of liquidity has increased. But I don’t spend these reduce balances I deposit them at the bank.

    So spending and velocity fall.

    If I’m a net borrower:
    – I borrow less at the higher rates.

    This also leads to less spending and less velocity

    Can someone explain where the increased velocity comes from ? Are there other factors that I am missing that drive the increased velocity and inflation?

  39. Gravatar of Ray Lopez Ray Lopez
    14. August 2015 at 07:22

    Sumner asks me: “Ray, Why have prices fallen in Japan despite large increases in the (fiat) money supply?” –

    First, because people in Japan don’t believe –wrongly IMO–that their government is in trouble, so they hoard money and don’t buy things (an Asian specialty btw). Mass hysteria is the term of art. The same reason the German population post WWI believed that the German military did not lose the war but were cheated into surrendering (this caused problems that lead to WWII).

    Second, prices have fallen for 20 years but that’s not the “long run”. Give it time, like the French Revolution it’s too early to tell.

  40. Gravatar of ssumner ssumner
    15. August 2015 at 05:50

    Market, That’s a near perfect example of where Keynesian thinking leads you astray. You can’t (in aggregate) put money in the bank, as the banks also don’t want to hold base money when interest rates rise. No one does. The only way for people and banks to get rid of this hot potato is by spending the extra money until NGDP rises enough so that people do want to hold the money.

    Also, claiming you will “spend less” when interest rates rise is to engage in reasoning from a price change. Historically people tend to spend more when interest rates are high, and vice versa.

    By the way, there are literally 100s of money demand studies that confirm that V tends to rise as interest rates rise.

    Ray, Mass hysteria? The Japanese, or you?

  41. Gravatar of Market Fiscalist Market Fiscalist
    15. August 2015 at 10:17

    Scott,

    Thanks for the reply however I am still am not fully convinced.

    We’re talking here about a change in r unaccompanied by a change in M , right ?

    Most likely this would be driven by a genuine change in market conditions (people want to borrow more and lend less). In this case the change in V drives the changes in r – which is the opposite of what you are claiming.

    So we must be talking about a (theoretical) situation where the banks randomly raise r above market clearing levels while M, the money demand curve, and everything else stays the same.

    In this theoretical case people want to hold lower cash balances at those higher rates. So they take their money to the bank. At that point either

    – The banks have to lower interest rates back to the equilibrium level again so that the loan market clears
    or
    – They take the deposits anyway at the higher r and hold the extra cash themselves and presumably make a loss of these holding
    or
    – They refuse to take some of the money people try to deposit and people then have the choice of either continuing to hold these cash balances or spending them.

    Only the third of these scenarios will (possibly) lead to increased V. But these results can only be obtained if we have banks acting in a non market-clearing way. And even then I don’t see the increase in V as a sure-thing as people may just go back to holding larger cash balances once they realzie banks won’t take their money.

    Also why is claiming people ‘will “spend less” when interest rates rise is to engage in reasoning from a price change’ ? We’re assuming here a arbitrary increase in r while everything else stays the same. This must surely lead to lead lower present spending and increased saving for future spending ?

    (We WOULD be arguing from a price change if we assumed that the rise in r was due to change in market conditions – but we’re expressly ruling that out here , as in that case the change in V drives the chnage in r)

  42. Gravatar of Market Fiscalist Market Fiscalist
    15. August 2015 at 16:46

    To state it more simply.

    I can that interest rates may increase

    1. Because M has decreased. In this case higher rates will be associated with deflation
    2. Because other things (such as people’s desire to lend and borrow money and hold cash balances) have changed in the economy. In this case higher rates will be associated with inflation.

    I’m struggling with the ideas of rates changing for neither of these reasons. But if you accept this possibility (as a thought experiment) the rate increase is then inflationary only if you have banks “rationing” the deposits they allow so that a HPE effect is needed to make price adjustments equilibrate people’s cash holdings.

  43. Gravatar of ssumner ssumner
    16. August 2015 at 05:12

    Market, You said;

    “Most likely this would be driven by a genuine change in market conditions (people want to borrow more and lend less). In this case the change in V drives the changes in r – which is the opposite of what you are claiming.”

    No, you have the causation wrong, it runs from interest rates to V. Yes, rates change for a reason, such as a sudden burst of good investment opportunities, or in general a boom in the economy, but then causation goes from rates to V.

  44. Gravatar of Market Fiscalist Market Fiscalist
    16. August 2015 at 06:02

    “rates change for a reason, such as a sudden burst of good investment opportunities, or in general a boom in the economy”

    Thanks for that clarification – it seems you also agree that something real is causing rates to change.

    But won’t some things (such as a desire to hold less cash balances) first increase V and then only later be reflected in higher rates ? While other things (more borrowing to fund investment) likely be seen first in higher rates followed then by higher V ? Overall V and r will increase together and it is hard to say which is “causing” which.

    BTW: I agree with 99% of what I read on this blog and only quibble this point as a learning opportunity as I grapple with this fascinating subject.

  45. Gravatar of Market Fiscalist Market Fiscalist
    16. August 2015 at 06:16

    I just thought of a different way to look at it it. Take a CB-induced rise in interest rates via a decrease in the money supply. Looked at in your way we would have to see 2 separate effects

    – A deflationary effect due to the smaller M
    – An inflationary effect due to the increased r

    Its pretty obvious how a HPE affect drives the first bit.

    But if nothing else has changed apart for the decreased M then while it is true to say “the increased r causes the demand for cash balances to fall” – to me anyway this movement along the demand for money curve feels like it is all part of the initial HPE (in response to the fall in M), and not a separate one (in the opposite direction) caused by the increase in r.

  46. Gravatar of Willy2 Willy2
    16. August 2015 at 12:18

    Sheer nonsense.
    – Hihger interest rates are actually VERY DEFLATIONARY and rates going lower are INFLATIONARY.

    – What A LOT OF people overlook is that the rates are not directly connected to inflation but that between inflation & rates a thing called “Wages” can be found.

  47. Gravatar of ssumner ssumner
    17. August 2015 at 04:57

    Market, If you draw a supply and demand curve for money, with 1/P on the vertical axis, then when Ms shifts to the left (contractionary), the Md also shifts left, but not as far. These two effects are logically separate.

    Willy, How do higher rates affect velocity? What happened to interest rates in 2009, the first year of deflation in the US since 1955?

  48. Gravatar of Market Fiscalist Market Fiscalist
    17. August 2015 at 05:43

    ‘ If you draw a supply and demand curve for money, with 1/P on the vertical axis, then when Ms shifts to the left (contractionary), the Md also shifts left, but not as far. These two effects are logically separate”

    OK, its the Md shifting left I’m not getting. I just see a shift along the Md. Let me think about that some more. Do you have any posts that explain why the Md shifts ?

  49. Gravatar of Market Fiscalist Market Fiscalist
    18. August 2015 at 06:24

    OK, I think I see what you mean.

    If you add a third dimension to that chart to show interest rates, then when M changes the Dm would change to reflect both the change in M and the change in r, and in the 2 dimensional chart would look like a shift in the Dm curve.

  50. Gravatar of ssumner ssumner
    18. August 2015 at 10:20

    market, Sounds right, but I’m not good at thinking in three dimensions!

  51. Gravatar of Willy2 Willy2
    18. August 2015 at 14:57

    Higher velocity don’t affect rates.
    – In the 1990s (money) velocity rose but rates fell (on average).
    – From about 2001 up to now (money) velocity continued to drop while rates did the same.

Leave a Reply