Barsky and Summers explain why low rates are contractionary

Most people seemed to think my previous post was crazy, and looked for weaknesses.  A few perceptive observers, such as Nick Rowe and Jonathan, noticed that it had the same implication as the IS-LM model.  Some people wrongly assumed I was simply talking about correlation, whereas I was claiming that lower interest rates cause falling NGDP.  Some wondered why that is not reasoning from a price change.

I think the best way to address this confusion is to start with the classic 1988 paper by Barsky and Summers.  They claim that the “Gibson Paradox” is caused by the fact that low interest rates are deflationary under the gold standard, and that causation runs from falling interest rates to deflation.  Note that there was no NGDP data for this period, so they use the price level rather than NGDP as their nominal indicator.  But their basic argument is identical to mine.

The Gibson Paradox referred to the tendency of prices and interest rates to be highly correlated under the gold standard. Initially some people thought this was due to the Fisher effect, but it turns out that prices were roughly a random walk under the gold standard, and hence the expected rate of inflation was close to zero.  So the actual correlation was between prices and both real and nominal interest rates.  Nonetheless, the nominal interest rate is the key causal variable in their model, even though changes in that variable are mostly due to changes in the real interest rate.

Since gold is a durable good with a fixed price, the nominal interest rate is the opportunity cost of holding that good.  A lower nominal rate tends to increase the demand for gold, for both monetary and non-monetary purposes.  And an increased demand for gold is deflationary (and also reduces NGDP.)

Of course that’s just the demand for gold, what about the supply?  It so happens that the supply of gold was fairly stable under the gold standard, rising by about 2% per year, whereas the demand for gold was much more unstable.  Thus changes in the value of gold (which was the inverse of the price level under the gold standard) were mostly caused by shifts in the demand for gold, which were in turn caused by changes in nominal (and real) interest rates.

An interesting question is how these changes impacted the real economy.  I would argue that sticky wages caused the fall in NGDP to result in a fall in hours worked.  A real business cycle proponent might deny that, and claim that whatever caused real interest rates to decline, also caused workers to want to take long vacations.  But anyone who denies the RBC model, and believes wages are sticky, should agree with me; causation goes from interest rates to NGDP to hours worked, even if the initial change in real interest rates was caused by a real shock.  Falling NGDP has an independent effect on hours worked even if caused by a real shock, just as a gunshot wound hurts someone who already has pneumonia.

As far as the claim that this is just IS-LM, I suppose that’s true, but it didn’t stop Barksy and Summers from getting their paper published in the JPE, nor did it prevent Tyler Cowen from calling it an enjoyable paper that addressed an interesting “puzzle”.

The puzzle of why the economy does poorly when interest rates fall (such as during 2007-09) is in principle just as interesting as the one Barsky and Summers looked at.  Just as gold was the medium of account during the gold standard, base money is currently the medium of account.  And just as causation went from falling interest rates to higher demand for gold to deflation under the gold standard, causation went from falling interest rates to higher demand for base money to recession in 2007-08.

There’s no “trick” in my previous post, I meant what I said.  But I’m not surprised that people are confused; after all, didn’t most economists believe the Fed was pursuing an “expansionary” policy in 2008?  Funny how those “expansionary” policies are almost always associated with recessions.  People tend to wrongly equate interest rate movements and “monetary policy”.  Most changes in interest rates reflect changes in the macroeconomy (growth and inflation) not monetary policy. When rates fall, the Wicksellian rate is usually falling faster, which means money is getting tighter in the NK model.

And finally, a word on reasoning from a price change.  Suppose you claimed that low rates should lead to more housing construction, or more investment in general.  That would be reasoning from a price change.  You’d be talking about the impact of the change in a price, on the quantity in the very same (credit) market.  Obviously if low rates are caused by more supply of saving, then the quantity of investment will rise, and if caused by less demand for investment, then the quantity of investment will fall.  But here’s what I’d like to emphasize.  Lower rates will reduce velocity and NGDP regardless of whether they are caused by more supply of saving or less demand for investment.  And that’s because interest rates are not the price of money, they are the price of credit.  So interest rates become a shift variable in the money market.  Lower rates shift the demand for money to the right, which raises the value of money, which is deflationary.  So there’s no reasoning from a price change in that case.  Of course if I didn’t hold the supply of base money fixed, it would be reasoning from a price change.


Tags:

 
 
 

98 Responses to “Barsky and Summers explain why low rates are contractionary”

  1. Gravatar of Nick Rowe Nick Rowe
    22. December 2015 at 06:49

    OK, I’m totally with you now.

    Didn’t know about the relation to Gibson’s paradox. Makes sense.

  2. Gravatar of Benoit Essiambre Benoit Essiambre
    22. December 2015 at 07:21

    I dunno, shouldn’t you distinguish central bank rates, the price of money, from market rates in the credit markets. Isn’t it assumed people mean the former when they are talking about CB policy? I still find there is a wording issue in this whole discussion.

  3. Gravatar of flow5 flow5
    22. December 2015 at 08:01

    “Lower rates will reduce velocity and NGDP regardless of whether they are caused by more supply of saving or less demand for investment”
    ———-

    Maybe income velocity, but not transactions velocity (but only in the short-run). That’s your problem. That’s why you insist on targeting N-gDp as opposed to the correct target – R-gDp.

    Economic prognostications are infallible.

  4. Gravatar of John Handley John Handley
    22. December 2015 at 08:12

    Scott,

    “An interesting question is how these changes impacted the real economy. I would argue that sticky wages caused the fall in NGDP to result in a fall in hours worked. A real business cycle proponent might deny that, and claim that whatever caused real interest rates to decline, also caused workers to want to take long vacations. But anyone who denies the RBC model, and believes wages are sticky, should agree with me; causation goes from interest rates to NGDP to hours worked, even if the initial change in real interest rates was caused by a real shock. Falling NGDP has an independent effect on hours worked even if caused by a real shock, just as a gunshot wound hurts someone who already has pneumonia.”

    I completely agree with you here, I just think it’s very important to point out that a supply shock caused the rate reduction in this first place. At that point, (absent a change in the money supply) nominal rigidities would impact output in their own right.

    “Lower rates will reduce velocity and NGDP regardless of whether they are caused by more supply of saving or less demand for investment.”

    Yes, if the rate changes did not occur because of a money supply change. Most rate changes occur because of a money supply change which is why everyone finds it strange that you’re talking about supply-driven real interest rate changes filtering through to the nominal interest rate. In a New Keynesian framework, what you’re looking at is a reduction in the natural rate of interest which everyone should know is consistent with a contraction.

  5. Gravatar of Ken Simpson Ken Simpson
    22. December 2015 at 09:19

    Count me among the confused.

    “Lower rates shift the demand for money to the right, which raises the value of money, which is deflationary.”

    “demand for money” – Yes, for borrowers
    “demand for money” – No, for savers (or for the Dollar via currency exchange)

    In both cases, more money should be in hand, searching for investments or purchases. More dollars chasing the same amount of goods is inflationary. (Although, more investment in productivity would be deflationary after some time.)

    And, in the case of the potential borrower, they have a choice, in their desire for money, to either borrow it or sell goods or services, or other assets. Low rates make borrowing the easy path. That means they care less about selling goods, which means they would have a tendency to raise the price on those goods, services, or assets.

    My brain has a hard time getting past the point that the Fed lowered rates, to try to make more money available for spending, to try to increase prices and NGDP, averting deflation, which sounds like a sensible policy.

    Finally, “Lower rates shift the demand for money to the right”. But then, higher demand for money leads rates higher again. Where the equilibrium falls depends on many variables, not the least of which are outlook, sentiment, and confidence. Those don’t seem to be in any of the equations.

    Sometimes, watching policymakers try to set prices via rates is like watching a dog chase its tail. Especially when they raise rates by paying banks not to lend, via interest on reserves or reverse repos.

  6. Gravatar of Nick Rowe Nick Rowe
    22. December 2015 at 10:01

    Reading some of these comments makes me depressed. But “Ours the task eternal” is my university’s motto.

    Listen up guys. There is:

    1. The interest rate you get paid for *holding* money.

    2. The interest rate you pay for *borrowing or lending* money.

    These are not the same thing. And they have totally opposite effects on NGDP.

    An increase in 1 reduces V and NGDP (for given M), because it makes holding money more attractive.

    An increase in 2 increases V and NGDP (for given M), because it increases the spread between 1 and 2, so increases the opportunity cost of holding money, which makes holding money less attractive.

    And, if we are talking about the Fed, and the base money it issues, 1 can be broken down into:

    1a. Currency. It pays 0% nominal.

    1b. Positive balances in commercial banks’ checking accounts at the Fed. They pay 0.5%. (That’s the one they just raised)

    1c. Negative balances (overdrafts) in commercial banks’ checking accounts at the Fed. (Not sure what that is, but it’s 0.75% in Canada, so probably the same at the Fed.)

  7. Gravatar of Dan W. Dan W.
    22. December 2015 at 10:27

    Scott,

    The policy implication of your claim is the FED you never reduce interest rates. Yet you claim that in 2008 the FED tightened when it was slow to reduce interest rates.

    BTW, if you search the archives I once remarked that the better FED response would be to raise rates to the desired level and let the financial chips fall where they may. So I am way ahead of the curve than many on this matter!

  8. Gravatar of Jose Romeu Robazzi Jose Romeu Robazzi
    22. December 2015 at 10:28

    It would help to be specific. I assume most of the time Prof. Sumner and commenters are talking about the short rate, which the central bank can peg if it is willing to adjust base money supply. But that rate can change even if the central bank does nothing.

  9. Gravatar of Dan W. Dan W.
    22. December 2015 at 10:28

    “… FED should never reduce interest rate”

  10. Gravatar of Jose Romeu Robazzi Jose Romeu Robazzi
    22. December 2015 at 10:35

    @ Nick Rowe
    Liked your comments, but I made the same ones a few months ago (distinction between your 1 and your 2 points) and if I am not wrong, Prof. Sumner disagreed. I maybe very off the mark here, but that is what I recall …

  11. Gravatar of Dan W. Dan W.
    22. December 2015 at 10:50

    Scott,

    You write: “The puzzle of why the economy does poorly when interest rates fall”

    Many would argue that interest rates decline in response to the economy not doing well, not that interest rate declines cause the economy to do poorly.

    Consistent with your thinking would be to observe there is a puzzle that human health does poorly when the doctor shows up. Of course the causation is the other way around. Poor health causes people to see a doctor!

  12. Gravatar of E. Harding E. Harding
    22. December 2015 at 11:01

    “Thus changes in the value of gold (which was the inverse of the price level under the gold standard) were mostly caused by shifts in the demand for gold, which were in turn caused by changes in nominal (and real) interest rates.”

    -Isn’t this the reverse? Changes in gold demand–>Changes in (expected) NGDP–>Changes in interest rates? That’s how I always thought it worked. Why doesn’t it work this way?

    And that’s because interest rates are not the price of money, they are the price of credit. So interest rates become a shift variable in the money market. Lower rates shift the demand for money to the right, which raises the value of money, which is deflationary. So there’s no reasoning from a price change in that case.

    -Again, isn’t this the reverse? But, then again, I was always confused by the money demand curve back in the day. After looking at some explanations of it again a few minutes ago, I’ve become even more confused. Though I might be starting to understand your point.

  13. Gravatar of James Alexander James Alexander
    22. December 2015 at 11:06

    When Bernanke set out to lower interest rates he specifically targeted making mortgage borrowing cheaper to boost housing demand. In his eyes this worked well. Indeed, house prices recovered it raised V and NGDP. But he also consciously neutralised the increase in base money used to bring about this fall in the price of credit about via the introduction of IOR thus reducing V and NGDP. Is there any way of telling which will be the bigger effect before the actions take place? Or is it a matter of “suck it and see”?

  14. Gravatar of Jose Romeu Robazzi Jose Romeu Robazzi
    22. December 2015 at 11:32

    @ James Alexander
    I was going to mention the same thing in the previous post: what if the central bank monetized (long term) private debt?

  15. Gravatar of Market Fiscalist Market Fiscalist
    22. December 2015 at 11:36

    I can think of 2 reasons for interest rates to fall

    1. The money supply increases
    2. People are prepared to hold the existing money supply at a lower rate of interest

    The first would tend to cause NGDP to rise, the second would cause NGDP to fall. But as in both cases the change in the rate of interest is caused by a change to another variable its not clear to me to why “a fall in the rate of interest is contractionary” is any more true than “a fall in the rate of interest is expansionary”.

  16. Gravatar of Kevin Erdmann Kevin Erdmann
    22. December 2015 at 13:18

    I know I’m all housing all the time, lately, but I think this post ties into the housing issue.

    Demand for housing is sort of like demand for gold, so falling rates leads to increased demand for home ownership, and given our conventions for financing home ownership, this means that along with lowering NGDP, lower rates lead to higher mortgage debt.

    This leads to a lot of analysis, which, to my eye, is confused, because the causality gets mixed up, with low rates -> expansion of the financial sector -> “overheating” -> eventual contraction.

    Using James’ comment above, even in 2002-2004, the Fed saw low rates and housing expansion as a sign of accommodation. But, in fact, long term real rates were low throughout the period as part of the conditions you are describing here. This, combined with supply constraints in highly valued locations (similar to a negative supply shock in gold), caused housing to expand because of the sort of causes you outline in this post.

    So, the way out of the context of 2003-2005 (besides removing obstacles to supply) was to create an increase in real growth and real interest rates. But, since policy makers have the causation all mixed up, they thought that we had to bring debt and housing growth down as the causal factor, even if that meant pulling NGDP down, too.

    Of course, now we have even lower interest rates and lower NGDP growth, as a result. You mentioned in the comments of the previous post that there doesn’t seem to be a strong relationship between housing expansion and real interest rates, since both were declining in 2007 and 2008, when we would expect home prices to rise when rates were falling. But, since monetary policy was working in this upside-down fashion, the explicit policy of the Fed could only push things into disequilibrium. Since policy wouldn’t relent, the Fed kept tightening credit and money until finally housing broke, and home prices disconnected from a ratex equilibrium by late 2007. In fact, that disconnect between home prices and long term real interest rates should be a bright red flag that by then things had gone terribly wrong.

    Today, interest rates and growth remain very low, which would lead to even more demand pressure for home ownership, and our policy response is to enforce an extreme ongoing negative supply shock in housing. As market prices for housing inevitably rise as a result of natural pressures in the face of this supply deprivation, it appears that the consensus policy view is that we will need to pull those rising prices (and the meager new supply that is associated with them) down with monetary contraction, which comes first through higher short term rates, but eventually creates lower rates and lower NGDP.

  17. Gravatar of TravisV TravisV
    22. December 2015 at 13:52

    Larry Summers: I think the Fed is making these 4 mistakes right now

    http://www.businessinsider.com/larry-summers-4-part-fed-criticism-2015-12

  18. Gravatar of Kevin Erdmann Kevin Erdmann
    22. December 2015 at 14:32

    Hm. Just saw your econlog post, which is sort of telling the same story as my comment above….

  19. Gravatar of wiretap wiretap
    22. December 2015 at 14:42

    I feel like you have said this before somewhere. Or at least I took it to be implied by a previous post, possibly your response to misguided comments by Bernanke about stimulating the economy by Fed policy aimed at lowering long-term rates.

  20. Gravatar of TravisV TravisV
    22. December 2015 at 15:44

    “Ted Cruz backers plan $1 million push on gold standard”

    http://www.cnn.com/2015/12/22/politics/ted-cruz-gold-standard

  21. Gravatar of Benjamin Cole Benjamin Cole
    22. December 2015 at 16:18

    As my late Uncle Jerry used to say, “Even if it is true, I still don’t believe it.”

  22. Gravatar of Major.Freedom Major.Freedom
    22. December 2015 at 18:04

    Sumner, your flawed argument and confusions are not rescued by the Gibson Paradox.

    The Gibson Paradox is only the empirical observation that interest rates are positively correlated with prices.

    You can’t use this as basis for your very different causative argument. This is funny because in your opening paragraph you said your initial post was not a correlation argument, but a causative one. But then you cite a correlation theory to ground your causative claim? Say what?

    Your confusions have you reversing cause and effect.

    The reason interest rates and prices were correlated under the gold standard, and why this correlation could for certain periods persist after the end of Bretton Woods and into fiat, is because nominal interest rates are the effect, not the cause, of the difference between aggregate spending (i.e. revenues) and investment spending (i.e. costs). Interest rates are constrained to profit rates. The higher are business profits, the more interest they will require to lend and the more interest they are able to pay to borrow.

    In other words, interest is the difference between two classes of spending, present spending and past spending. Borrowing and lending consists of two classes of expenditures, the initial principle and the principle plus interest.

    The faster the money supply growth, such as in a fiat standard, the more will aggregate spending increase above investment spending, and thus the higher nominal profits and interest rates will become. The slower the money supply growth, such as in a gold standard, the less will aggregate spending increase above investment spending and thus the lower nominal profits and interest rates will become.

    This is of course ceteris paribus. It assumes away the temporary liquidity effect that can arise on the basis of inflationary credit expansion which can for a time depress interest rates, until the above pressures on the difference between aggregate spending and investment spending push rates back up (usually once the liquidity effect is reduced or eliminated by central banking system reducing the extent of credit expansion). This is of course the source of the business cycle.

    Prices in a gold standard may exhibit empirical characteristics of a random walk, but this primarily due to the fact that productivity increases often roughly matched or slightly exceeded the rate of increase in the money supply and volume of spending in a gold standard, such that prices sometimes fell, sometimes rose, and sometimes remained stable. But this does not in any way imply that all of a sudden interest rates are now the causal factor! The correlation between interest rates and prices is NOT evidence of interest rates being a cause for prices (or NGDP).

    The reason you have confusions on this is that you are unable or unwilling to think outside the New Keynesian box. The flawed core of the theory makes it literally impossible to correctly describe actual market processes, and actual causal relationships.

    It seems, rather ironically, that the theory of money is your biggest weakness.

    ——————————-

    Nick Rowe, the reason there is so much confusion in economics is because students are taught fallacious doctrines in school from their professors, and carry it with them to blogs, the media in general, and [shudder] government.

    There is no such thing as interest rates paid on the basis of “holding money”. The money you deposit into your checking account is a loan with infinitesimal maturities. The money you deposit is in fact a loan to the bank. The bank doesn’t bury your money. They keep a portion of it to facilitate day to day withdrawals and transfers, lend the rest at interest, and remit a small portion of that interest to you. That is not payment for you holding money! Try holding money in a safety deposit box, or under your mattress, and see what interest you are paid.

    It is literally impossible, in the context of the economy as a whole, for everyone to “hold” more money, unless there is an increase in the supply of money. If 100 million people all tried, and succeeded, in increasing their cash balances, then guess what? That success requires other people to reduce their cash balances, that is to say, it requires an offsetting fall in the demand for money holding on the part of other people. Dollar for dollar, with no exceptions. This is mathematically necessary in the absence of inflation of the money supply.

    When you think of periods in which “the demand for money holding increases”, you are not actually thinking of a world of people holding more money. You are actually thinking of a world where the length of time between earning money and spending money increases on average, that is to say, a world where the velocity of money decreases. With a decrease in velocity and unchanged M, that reduces aggregate spending. THAT is what puts downward pressure on prices and interest rates. This is not a case of interest rates lowering inexplicably and then that puts downward pressure on prices. It is the same cause of the downward pressure on prices that causes interest rates to fall.

    Please for the love of economic science, drop the whole IS-LM monstrosity, and the whole NK framework, as it is preventing clarity in the minds of students all over the world. The abuse has to stop. Just because your profs abused your generation, there is no reason to continue the madness.

  23. Gravatar of Nick Rowe Nick Rowe
    22. December 2015 at 18:45

    MF: Here’s a fun post I wrote, just for you:

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2015/12/what-is-a-cashless-economy-anyway.html

  24. Gravatar of Major.Freedom Major.Freedom
    22. December 2015 at 21:20

    Nick:

    “…an exogenous increase in the rate of interest paid on cash will cause each individual to plan to accumulate cash balances, by spending less cash than he expects to receive from others.”

    That is not possible to actuslly achieve without an increase in the supply of money.

    You can plan to hold more cash all you want, indeed everyone can plan to hold more cash, but any increase you or any subgroup of the population are able to acquire, is exactly offset by other people not in that group purposefully reducing their cash balances.

    Even if every single individual tried to increase their cash balances, this would have the effect of a drop in aggregate spending and it is that which reduces interest rates. It is no counter to this to say “Well in my hypothetical communist scenario, there is no market for lending and borrowing, and interest and money are state monopolies.”

    Even if there is a central bank that pays interest in a world where lending and borrowing were not outlawed, it would not be able to monopolize all interest payments. When there is private ownership of the means of production, and there are profits to be made, the rates on loans in the market will be a function of profits. Right now the Fed is paying IOR, but this did not absorb all savings into the central bank such that no other interest bearing lending or deposits took place.

    In the market, there is no such thing as “exogenous” interest. It is all endogenous. Profits and interest can persist even with a completely fixed money supply. As long as capitalists draw funds out of their firms to spend on their own consumption, net of wage payment financed consumption, the difference between total revenues and costs need not ever fall to zero. And even if for whatecer reason it did, market forces would reverse that trend.

    I don’t find much use out of overly contrived models. Those that assume away so many real world factors so as to be an imaginary story are best left as like you said, fun pieces to read.

    There is no equilibrium in the real world. I never use that conceptualization as an actual descriptive mechanic, but rather only as a mental tool from which to understand how the world differs from it, and in so doing, understand the market a little better. The market is inherently always in disequilibrium, thankfully so.

  25. Gravatar of Nick Rowe Nick Rowe
    23. December 2015 at 04:15

    MF: “That is not possible to actuslly achieve without an increase in the supply of money.”

    Yes, I know that. I wasn’t born yesterday. Just like it’s not possible for everybody to buy more land. But their individual attempts to do something that is impossible in aggregate cause something (P, or Y, or r) to adjust until they change their plans. As Hayek said.

  26. Gravatar of Dan W. Dan W.
    23. December 2015 at 05:08

    Kevin,

    Loose credit standards increase the demand for credit, and with all other things being equal, will increase interest rates. Tighter credit standards will have the opposite effect.

    Ample evidence suggests credit standards were too loose in the early to mid 2000s. Loans were made that could not be repaid, no matter the monetary policy or NGDP level. It was common sense that credit standards would tighten and it is logical that such tightening would depress housing loans specifically and real estate activity in general.

    What is your opinion on credit standards in the 2000s? Were they too loose? Did they become too tight?

    Question to all monetarists: Do you accept that credit standards can be too loose and that the extension of “bad credit” can distort economic activity, until loans start to fail and the problem is manifest? If you reject the idea that credit standards can be too loose then are you not taking the position there should be no credit standards at all? I find that position to be irrational.

  27. Gravatar of Alejandro Alejandro
    23. December 2015 at 06:29

    This is excellent. How do u link this argument with oversupply and overcapacity or malinvesment in cmdty mkts? The direct argument is higher rates will speed up prices as supply normalizes (or unprofitable projects shut down) but not sure the impact in NGDP.

  28. Gravatar of ssumner ssumner
    23. December 2015 at 06:38

    Thanks Nick.

    Benoit, You said:

    “I dunno, shouldn’t you distinguish central bank rates, the price of money, from market rates in the credit markets. Isn’t it assumed people mean the former when they are talking about CB policy?”

    I’m trying to convince people that market interest rates don’t reflect central bank policy, they reflect the economy.

    John, Don’t you mean “real shock” rather than supply shock. It need not be a shock that impacts equilibrium output. It need not shift the SRAS or LRAS curve, it might only shift the AD curve. I do understand that it might affect SRAS, and indeed that may be the most common case, but it’s not necessary.

    Ken, You said:

    “In both cases, more money should be in hand”

    You are confusing the credit and the money markets. I’m assuming a fixed money supply.

    You said;

    “My brain has a hard time getting past the point that the Fed lowered rates, to try to make more money available for spending, to try to increase prices and NGDP, averting deflation, which sounds like a sensible policy.”

    That’s a widely held view, but wrong. From August 2007 to May 2008 the Fed did nothing to lower interest rates, they fell in the free market, on their own accord. The Fed simply adjusted it’s target rate to market conditions.

    Nick, That’s helpful.

    E. Harding, Many things can cause gold demand to change, but interest rates are one of the most important.

    James, Hard to know because so much depends on the future expected path of policy. That’s why I keep saying that NGDP growth expectations ARE the policy.

    Market, Look at my gunshot analogy. The other factor may also impact NGDP, but the change in interest rates has an independent effect.

    Kevin, Yes, but falling EFFECTIVE demand for home ownership is one thing that caused lower rates. By effective I don’t just mean how much people wanted to buy, but rather how many qualified after banks tightened standards.

    Wiretap. Yes, pretty much all my posts are now repetitive, but I try to add a new wrinkle each time.

    Dan, Moral hazard creates excessive bank lending—that’s always true.

  29. Gravatar of Market Fiscalist Market Fiscalist
    23. December 2015 at 07:41

    I can see something like:

    – Some exogenous factor causes interest rates to fall
    – These lower rates cause an increase in demand for money
    – This reduces NGDP

    But I’m struggling to see what those exogenous factors might be that wouldn’t ultimately boil down to changes in the supply and demand for money – can you give an example of such an exogenous change in interest rates ?

  30. Gravatar of Ray Lopez Ray Lopez
    23. December 2015 at 08:16

    I largely agree with Sumner, who essentially is capitulating from some of his earlier views that the Gold Standard is unworkable {flamebait but essentially true}

    But more specifically Sumner does make some mistakes, which of course I am happy to correct.

    Sumner: “The Gibson Paradox referred to the tendency of prices and interest rates to be highly correlated under the gold standard. Initially some people thought this was due to the Fisher effect, but it turns out that prices were roughly a random walk under the gold standard, and hence the expected rate of inflation was close to zero” – not true, inflation was not zero, but went up and down in the 19th century. The referenced Barsky et al paper is flawed since the time period is too long. It stretches from 1831 to 1913, when in fact it should be much shorter (most people don’t live that long). Keynes comment about stormy seas comes to mind; and the analogy is that if a period is long enough, nothing matters (as everything goes to zero due to the laws of entropy).

    Sumner: “Lower rates shift the demand for money to the right, which raises the value of money, which is deflationary.” – no, not true at all as Ken Simpson pointed out. In fact, the IS-LM model, which Sumner is using, specifically allows only one way to shift the LM curve (as opposed to traveling up or down the upward-sloping LM curve), and that is by an increase (or decrease) in money (from the central bank). LM = only increase in money shifts curve down and decrease in money shifts curve up; all other things like taxes, spending, investment are irrelevant to LM curves (at least this is how Blanchard presents the curves). The IS curve is a different story: you can have shifts right or left due to taxes or government spending, or exogenous changes to investment, consumer spending or imports/exports. The only way to ‘save’ this confusing quote by Sumner is to assume that by ‘lower rates’ Sumner means ‘decreased money supply’ and ‘higher rates’ = increased money supply, but in the IS-LM model that is a confusion of terms. Another way to ‘save’ this quote is to assume that Sumner is using some sort of shorthand for reaching a conclusion, rather than quoting the IS-LM model properly. But that’s muddled logic.

    Mind you I don’t agree with the implicit assumptions behind the entire IS-LM curve, the sloping nature of the two curves, the fact that money can be changed by a central bank rather than production (Fed leads in the IS-LM model, but in reality, since money is largely neutral, the Fed follows production), but that’s a topic for another day.

    @MF – Rowe is right; you are wrong. Analogy: just because for every buyer there is a seller does not mean there is no upward or downward pressure on the price of a stock; you are ignoring rates of change and focusing on static conditions. Google Zeno’s paradox for another analogy.

  31. Gravatar of Ray Lopez Ray Lopez
    23. December 2015 at 08:25

    @myself- I just read “Market Fiscalist”‘s reply just above mine, and in more pithy prose he or she essentially is making my more verbose argument. Indeed Sumner is making unclear, unconventional explanations of the standard IS-LM model. It’s as if he’s saying a man died not because of blood loss shock due to a gunshot but ‘lead poisoning’ as a shorthand expression of shock.

  32. Gravatar of Britonomist Britonomist
    23. December 2015 at 09:27

    I feel like you just used the wrong definition of ‘ceteris paribus’ originally. For instance you can only say high rates are expansionary if you wholly and entirely disregard ‘ceteris paribus’: yes a positive demand shock will raise rates by raising expected inflation, or some kinds of positive supply side shocks will raise the ‘marginal product of capital’ (if such a concept is really meaningful, it probably isn’t) – causing rates to rise. However with ceteris paribus you’re supposed to assume that none of these positive shocks happen, and to only analyze the response to a rate rise exogenously without any demand or productivity shock to force a raise in the first place.

    What would happen for instance if the Bank of England suddenly decided to raise interest rates to 5% for no reason, not because of positive signals about inflation or a from a seriously overheating economy but just because it could, would it be expansionary?

    1) Government debt interest would become very hard to service, forcing a fiscal consolidation and further austerity: contractionary

    2) Private debt for many would become unserviceable, leading to many bankruptcies, foreclosures etc… Contractionary

    3) Borrowing would become much more expensive, leading to a drop in any kind of leveraged investment and lending to the private sector. Contractionary.

    4) Safe assets would become very attractive to a portfolio manager yielding a high safe return, causing a portfolio re-balancing away from stocks and commodities towards government gilts. Contractionary.

    5) Deposit interest rates would rise, in almost any modern macro model with a dynamic budget constraint, this reduces consumption (I don’t care about silly old fashioned models that insist on equating ‘savings’ with investment’, they’re not the same in a modern broad money based economy, simply paying back bank debt or leaving more money in a bank deposit does not have to lead to a corresponding rise in investment – and even if you do rely on old models, what happened to the paradox of thrift?)

    6) Interest income would increase for some people. Expansionary.

    I’d say on balance, that’s contractionary ceteris paribus.

  33. Gravatar of John Handley John Handley
    23. December 2015 at 10:01

    Scott.

    “Don’t you mean “real shock” rather than supply shock. It need not be a shock that impacts equilibrium output. It need not shift the SRAS or LRAS curve, it might only shift the AD curve. I do understand that it might affect SRAS, and indeed that may be the most common case, but it’s not necessary.”

    In my mind there’s really no distinction between ‘real shock’ and ‘supply shock’. Any real shock is caused by either a shift in preferences (e.g. a sudden desire to hold more money) or a change in fiscal policy (e.g. a shock to government spending). I think generally my problem with your last two posts comes down to wording. You frame it in such a way that a deliberate reduction in the nominal interest rate, ceterus paribus, will result in lower output. What’s really happening is that a negative *real* shock hits the economy (say, for example, a shock to money demand) which causes the interest rate to fall and output to fall. The lower interest rate, because of the money demand function, serves to also impart some damage through nominal rigidities. That, to me, is very different from lower nominal interest rates being contractionary. They don’t cause the contraction in the first place.

    A deliberate reduction in the nominal interest rate by the central bank *can* be contractionary, but not at all for the reasons implied by your post; if the lower nominal interest rate is caused by expectations of tighter policy, then it is reasonable to suggest that what occurred was a contractionary move in monetary policy. If, on the other hand, the lower interest rate represents a temporary loosening of monetary policy (an increase in the size of the money supply relative to the next period), then the move in monetary policy should be seen as expansionary. I don’t for a minute have the same confusion as some others who think that monetary policy works through manipulating the real interest rate; that’s just a side effect of monetary policy.

    If a central bank decided to increase inflation today (but not in the future) and the real interest rate falls, then you’d see the conventional wisdom play out. If a central bank increased inflation today and the real interest rate rose, then the monetary policy would still be expansionary because output would increase. Of course, there are a few reasons why the real interest rate should fall if monetary policy becomes more expansionary: first the the consumption channel – because real interest rates are a function of expectations of future consumption growth and consumption should revert to steady state, regardless of monetary policy, a higher real interest rate means that consumption is expected to grow quickly relative to trend, which can only happen if consumption is currently below its steady state (unless real interest rates are going to be depressed for a long period after the boom, I guess). Second is the capital stock – because the real interest rate is a proxy for the marginal productivity of capital, which is decreasing in capital, higher real interest rates are consistent with less capital and therefore less current investment so as to sufficiently reduce the capital stock. In both cases, a high real interest rate is consistent with low levels of demand.

    In the same way, a lower real interest rate should be consistent with a supply-side boom in output. A money demand shock would have the opposite effect, but now all you’ve really said is that an excess demand for money is a bad thing and its signal is low interest rates with a large money supply. I suppose you’ve even helped to point out the liquidity trap argument; it turns out that no money supply would have satiated money demand in 2008, so a recession was bound to happen and so were low rates. The problem is that every shock is not a money demand shock and only money demand shocks should work this way. Normally, the interest rate should fall because the central bank increases the supply of money and money demand needs to correspondingly increase. It’s the increase in the money supply that is expansionary, but the interest rate will still fall given the monetary expansion.

  34. Gravatar of Carl Carl
    23. December 2015 at 10:45

    I’m still hopelessly lost…
    Falling interest rates decrease the opportunity cost of holding money, causing more people and institutions to hold onto it, thereby deflating NGDP. So, if I were to notice tomorrow that the price of credit (i.e. interest rates) were falling it would be okay to conclude that NGDP would fall, but not that any individual market (e.g. housing construction) will be depressed. Can’t I even say that an individual market is more likely to be depressed than if interest rates were rising? You know, rising tides and rising ships and all that.

  35. Gravatar of Britonomist Britonomist
    23. December 2015 at 11:04

    Here’s an interpretation, are you assuming that interest rates are endogenous and that (base) money is the exogenously determined variable? While everyone else is assuming the base is endogenous and it’s the interest rate which is exogenous?

  36. Gravatar of Kevin Erdmann Kevin Erdmann
    23. December 2015 at 11:27

    Dan W., there was surprisingly little change in the characteristics of the typical homeowner during the 2000s. The tightening of credit standards and the collapse of housing starts since 2007 have been much sharper than any previous movements in the other direction.

    Scott, I’m not sure why you said “but falling effective demand for homeownership is one thing that caused lower rates”. I agree with this. This is a core factor in my comment. So, I would keep your statement, but without the “but”. Some combination of monetary and regulatory policy, expectations, and market reactions created a disequilibrium in housing markets where prices were overwhelmingly defined by access (or lack thereof), not by inherent supply and demand for housing or ownership. One of the core elements of a functional economy is broad access. The sharp decline in access to housing that began around 2007 marks a sharp deviation from that ideal, and economic healing and recovery will necessarily be marked by the convergence of inherent demand and effective demand. Regulatory and monetary policy remain explicitly focused on preventing that convergence.

  37. Gravatar of Kevin Erdmann Kevin Erdmann
    23. December 2015 at 11:33

    One sign that this convergence is happening will be if real long term interest rates rise. I agree that the slight decline in interest rates in the 2000s was a product of the factors you are describing. But, I think the sharp fall in treasury rates since 2007 is an anomaly. Rates would still be lowish without the housing problem. But, we have a situation where home ownership is returning something like 4-5% in real returns, after costs and depreciation, and the total value of that market is at least as large as the market in treasuries. So, there is really an average nominal “low risk” long term interest rate around 4%-5%, which is divided between those with access to housing, who earn more, and those without access, who earn less. The implication of low treasury rates in this context may be slightly distorted compared to an open access market.

  38. Gravatar of Gary Anderson Gary Anderson
    23. December 2015 at 12:55

    Banks want to lend when they can make money. Low interest rates actually tighten things up so I agree with Scott. But, and this is what Scott may be ignoring, that low interest rates are a result of demand for bonds. And demand is massive, and becoming more massive as time passes. Demand increases as bonds are used as collateral in clearing houses. I think we are stuck, and Greenspan’s plan all along was to stick us with low interest rates forever.

    Why economists like Sumner and Erdmann refuse to want to study supply and demand of long bonds is beyond my comprehension. That is an issue for economists.

  39. Gravatar of ssumner ssumner
    23. December 2015 at 14:09

    Market, If you draw a diagram of the market for money with interest rates on the vertical axis, then yes, any change in interest rates must be due to either a change in the supply or demand for money.

    Here’s what I meant. Suppose oil is discovered in a country, and this makes interest rates rise. This will then raise V, and if M is fixed it will also raise NGDP.

    Ray, You said:

    “not true, inflation was not zero, but went up and down in the 19th century.”

    I agree, when did I say otherwise?

    And no Ray, I am not using the IS-LM model. You aren’t even on the same planet in terms of this discussion.

    Britonomist, You said:

    “However with ceteris paribus you’re supposed to assume that none of these positive shocks happen, and to only analyze the response to a rate rise exogenously without any demand or productivity shock to force a raise in the first place.”

    That’s exactly what I am doing. I am making no assumption about why i changed, just looking at the impact, ceteris paribus. Now I know that i always changes for a reason, so I decided to specifically assume the reason was not a change in M, and look at the impact if due to some other factor. The vast majority of changes in interest rates are not due to M.

    John, You said:

    “Any real shock is caused by either a shift in preferences (e.g. a sudden desire to hold more money)”

    You are entitled to your definitions, but they are not standard. The desire to hold more money is usually termed a demand shock. But that’s not the problem, this is:

    “The lower interest rate, because of the money demand function, serves to also impart some damage through nominal rigidities. That, to me, is very different from lower nominal interest rates being contractionary. They don’t cause the contraction in the first place.”

    When I say it’s contractionary, I mean it causes a lower V. I do not mean that the lower i directly reduces MV (total spending), I mean it reduces MV indirectly by directly reducing V.

    You said:

    “I suppose you’ve even helped to point out the liquidity trap argument; it turns out that no money supply would have satiated money demand in 2008, so a recession was bound to happen and so were low rates.”

    I strongly disagree here. Interest rates were not even at zero in 2008, and even if they had been at zero a sufficiently expansionary monetary policy would have prevented a recession.

    Some of the middle part of your comment I can agree with, but I’d caution you that it’s at least theoretically possible that a expansionary monetary policy leads to higher real interest rates, if it leads to a sufficiently powerful increase in real GDP growth expectations. That’s been shown in ratex models (Robert King I recall made this point.) Having said that, I agree that easier money usually depresses real rates. But tight money often depresses real rates with a modest lag, by creating recessions.

    I also have a reply on the previous post, and I’m doing some posts at Econlog on this as well—I’ll have a new one soon.

    Carl, No, while it is usually true that falling interest rates are associated with a weakening economy, recall that I assumed that the monetary base was held constant. If the Fed reduced rates with a bigger monetary base, or a lower IOR, it would be expansionary.

    Britonomist, You said:

    “While everyone else is assuming the base is endogenous and it’s the interest rate which is exogenous?”

    Not Paul Krugman! He insists that the Fed is not actually doing anything unless they are changing the monetary base. (In his critique of Friedman on the 1930s.)

    Kevin, I agree with most of that, and may have misinterpreted your comment.

    I will say that I’m an investor in rental property, and very much regret it. I’d much rather earn 4-5% in the stock market than the housing market, even with the greater risk. The hassles in managing housing are just not worth it for me. Perhaps the greater inequality in America is making people who used to think being a landlord was a good deal, now think it’s not worth the effort.

    But you know much more about this than me. I’d want to investigate related questions, like what do REITs earn? What do corporate bonds with similar risk to REITs earn? What has happened to the gap between Treasuries and corporate bonds? In other words, not just houses and T-bonds, but everything in between. Where is the breakpoint? Where exactly are returns inexplicably high (or low), on that continuum?

    Gary, You said:

    “is beyond my comprehension.”

    Yup.

  40. Gravatar of Kevin Erdmann Kevin Erdmann
    23. December 2015 at 14:33

    I’m with you, Scott. Even though I have spent the past year basically developing the argument that the single most lucrative investment available now is leveraged rental housing, it just wouldn’t be my cup of tea. On the other hand, in the aggregate, most of us enter this market as owner-occupiers, where those problems don’t exist. The institutional obstacles to management of rentals is one of the reasons the market can remain out of sync, because REITs and institutional landlords can’t easily double or triple their footprint in the single family home business without some growing pains.

    On the other asset classes, except for home ownership they seem to be relatively close to long-term ranges of spreads between related classes. This even includes mortgages. The disconnect is strong. Aggregate BEA stats which would imply historical returns to home ownership of about 2 1/2% to 4% show them currently at the top of that range – around 4%. This is a real return. 30 year Mortgages have a nominal rate that is also about 4%. Normally, real mortgage rates would be similar to real implied home returns. Now, mortgage rates imply no inflation, or deflation, and this is while national aggregate rent inflation (the inflation home owners care about) is moving above 3%. It’s a free lunch. That’s because there isn’t limited access to mortgage funding, so mortgages, like the other asset classes, don’t provide an excess return.

  41. Gravatar of Kevin Erdmann Kevin Erdmann
    23. December 2015 at 15:00

    BTW, real estate investors would probably tell you I’m full of nonsense. These relationships are entirely emergent, and I have only inferred them from the data. Local factors and depreciation are such a large part of real estate cash flows and profits that it appears to me that the real estate sector simply uses their own set of measures. Since long term real rates used to be very stable, and local issues dominated, things like capitalization rates became rules of thumb.

    The funny thing is, I have seen real estate finance people pushing REIT funds because returns on other asset classes have gone down, but their cap rates are still high. And they only look at it as an income issue – that you’ll earn higher income on those REITs. As far as I know, it’s not even normal to compare real estate returns to TIPS bonds, even though this is an obvious comparison. They use nominal bonds as a comparison, as far as I can tell. So, they’ll talk about how their internal analysts are talking about moving their benchmark cap rate target from, say 5% to 4%, which will open up new investment possibilities for the fund while still providing a good return relative to other classes. As a more generalized speculator, my response is that moving the required returns on 50 year + duration assets should imply a huge capital gain from today’s values. But, they don’t seem to look at it that way. It’s all about income, and cap rates are just a sort of number that the guys in the back move around as market conditions require it.

    So, just like the bankers that are convinced that lending was crazy in the 2000s just because they were suddenly making loans that the FHA/Ginnie Mae used to handle, real estate insiders don’t seem to be that consciously aware of the excess returns coming from real estate, because they just aren’t in the habit of pegging their required returns to real risk free market rates. They just see their funds as a good source of income, and frankly I think they tend to move their required returns around as an after-thought.

    It’s kind of like CEO’s who mistakenly think they are moving production to low-wage locations when they are really moving to rising-wage locations. The insiders don’t necessarily need to understand what’s happening in order to emergent equilibriums to develop. In fact, the problem with the financial crisis was that we collectively had a conscious disagreement with the emergent prices we were creating, and we consciously imposed the wrong solution on those markets. It’s the political power of our consciousness that screws things up…or it could be that I’m just full of it.

    Sorry, I’m rambling on…

  42. Gravatar of Britonomist Britonomist
    23. December 2015 at 15:22

    “so I decided to specifically assume the reason was not a change in M, and look at the impact if due to some other factor.”

    But I think this is still not properly ceteris paribus unless your ‘other factor’ is not something already in the model. I think your ‘other factor’ is some shock to IS, which lowers interest rates and output – and then you’re saying this means a drop in i is contractionary ceteris paribus, but it’s not the change in i which is contractionary it’s the shock to IS, where lower i is just a symptom.

  43. Gravatar of Dan W. Dan W.
    23. December 2015 at 15:59

    Kevin,

    Did mortgage standards loosen or tighten between 1992 and 2002?

    Did mortgage standards loosen or tighten between 2002 and 2006?

    When you decry the tightening of mortgage standards after 2006 to what year / era did they tighten?

    This 2006 Wall Street Journal article discusses the increased frequency of “stated income loans” (ie liars loans) to get people into mortgages: http://www.wsj.com/articles/SB115620791723241743

    Is it your contention that the frequency of fraudulent mortgage applications did not increase between 2000 and 2006? The industry observers at the time disagree.

  44. Gravatar of Gary Anderson Gary Anderson
    23. December 2015 at 17:28

    Scott, that some formulas are beyond my comprehension I readily admit. Your insults don’t add to my previous admissions that I am not an economist. But apparently measuring demand for long bonds is beyond your comprehension as well. Even Larry Summers said there is a shortage of long bonds. That affects yield. So, can insult me all you want but you aren’t dealing with a subject that could distort all your other measurements. It is your blog and your decision to insult and not learn something from someone who thinks outside the narrow box you find yourself in.

    Real economists measure the impact of supply and demand in all sorts of ways. You do that. But when it comes to supply and demand for treasury bonds, which could ultimately affect lending and bank behavior, you are silent.

  45. Gravatar of Gary Anderson Gary Anderson
    23. December 2015 at 17:31

    Dan, there isn’t a mortgage Kevin doesn’t love. There is nothing loose enough. There is no mortgage that can’t be paid back. Like Rick Santelli said, the public just bet incorrectly. Well, fook Santelli.

    Of course, that lend recklessly view ignores the FIRREA Act of 1989. But what do economists care about the consumer? They work for the producer.

  46. Gravatar of John Handley John Handley
    23. December 2015 at 17:39

    Scott,

    “You are entitled to your definitions, but they are not standard. The desire to hold more money is usually termed a demand shock.”

    Fair enough. I’ll stick to calling them ‘real’ and ‘nominal’ shocks from now on.

    “When I say it’s contractionary, I mean it causes a lower V. I do not mean that the lower i directly reduces MV (total spending), I mean it reduces MV indirectly by directly reducing V.”

    Yes. I was abstracting from that when I wrote the comment; I guess I should have made it clear that that’s what I meant. I still don’t like the word ‘contractionary,’ but I guess that’s mostly semantics and has nothing to do with the economics.

    “I strongly disagree here. Interest rates were not even at zero in 2008, and even if they had been at zero a sufficiently expansionary monetary policy would have prevented a recession.”

    So, in your mind, velocity doesn’t simply fall to accommodate monetary expansion at the zero lower bound. I’m still not sure what your theoretical (and empirical) basis for this is, but that’s not eminently an issue most of the time (except for when the zero lower bound is binding, in which case I’d be inclined to say that the central bank’s balance sheet is completely irrelevant).

    “I’d caution you that it’s at least theoretically possible that a expansionary monetary policy leads to higher real interest rates, if it leads to a sufficiently powerful increase in real GDP growth expectations. That’s been shown in ratex models (Robert King I recall made this point.)”

    I do agree that it’s possible, I just don’t think it’s likely, unless the preceding slump really was terribly large and prolonged (which I guess would be the case right now). If the economy were at equilibrium before the shock, then I don’t think monetary expansion would really every be consistent with significantly higher real interest rates. Unless, perhaps, monetary policy is expected to continue being expansionary in the future and agents don’t react correspondingly with the adequate price/wage increases.

  47. Gravatar of Ray Lopez Ray Lopez
    23. December 2015 at 17:45

    Sumner in his post: “A few perceptive observers, such as Nick Rowe and Jonathan, noticed that it had the same implication as the IS-LM model.”

    Sumner to me: “And no Ray, I am not using the IS-LM model. You aren’t even on the same planet in terms of this discussion.”

    Sumner = the Harry Houdini of economics (and the Rush Limbaugh). Just when you thought you figured him out, he throws a curveball at you. Nobody understands Sumner and Sumner understands nobody, unless you happen to agree with him or sponsor him, like Ken Duda does.

    Economist to employer who asks “2 + 2 equals what?”: ‘what do you want it to equal, boss?’.

  48. Gravatar of Gary Anderson Gary Anderson
    23. December 2015 at 18:07

    As far as rental property is concerned, Scott should have bought in Palm Springs as my son did. The condo market there allows for massive nightly rents, since Palm Springs is super popular, without a hotel infrastructure needed for the influx of guests.

  49. Gravatar of Major.Freedom Major.Freedom
    23. December 2015 at 18:11

    Nick,

    “But their individual attempts to do something that is impossible in aggregate cause something (P, or Y, or r) to adjust”.

    Of course, but the key is which variable or variables do adjust, and what effect that variable or variables have on the other variables.

    Sumner’s and your contention is that changes in r cause P*Y changes, given unchanged M, whereas my contention is that it is the reverse. The “Gibson Paradox” talks about a correlation between r and P. I say this is due to changes in P causing changes in r, whereas you and Sumner say it is due to changes in r causing changes in P.

    What are the determinants of nominal interest rates? It is no good to imagine a model where the determinant is an arbitrary whim by some random group of individuals who impose a monopoly on interest rates and money. We have to ask why interest rates have been around since the dawn of seashells as currency, between private lenders and borrowers, including implied interest rates during the medieval times when there were “usury” laws.

    The ultimate reason is time preference, but on loans specifically the proximate reason is that there is a persistent trend in the market for aggregate expenditures to exceed productive expenditures, even with a completely fixed money supply. That is what makes interest on loans even possible for business firms. Without profits, business firms that invest using borrowed funds, and business firms that lend to other business firms, at interest, would go bankrupt.

    The height of interest rates is determined by the height of profits. The height of profits is determined by the extent to which aggregate expenditures exceed productive expenditures. The less investment that takes place, and the more consumption that takes place, the higher will profits and interest become. This is why we see emerging economies having higher nominal profits. It is because productive expenditures represent a smaller portion of aggregate expenditures. As economies mature, productive expenditures increase relative to aggregate expenditures, and that makes nominal profits fall…but not continually so, as no matter how much investment there is, the investors themselves will always need to consume, and that makes aggregate expenditures exceed productive expenditures all else equal. If aggregate spending falls below investment, avoidance of losses will put a reversing pressure on this trend until there is again profitability.

    These grand movements in aggregate revenues and aggregate costs are what force market interest rates up and down.

  50. Gravatar of Kevin Erdmann Kevin Erdmann
    23. December 2015 at 18:32

    Dan W., the character of the aggregate set of home buyers did not change significantly, according to the survey of consumer finance. Even if it had, everyone was so convinced about it that nobody considered the unlikely number of new marginal homeowners that the anecdotes implied there would be and the lack of a mechanism for it to translate into higher prices. Almost all the new mortgages were to the top 40% of households by income.

  51. Gravatar of Dan W. Dan W.
    24. December 2015 at 04:20

    Kevin,

    The character of home ownership, mortgage debt and loan type changed considerably between the early 1990s and the mid 2000s, with the greatest changes occurring in the 2000s. Here are some facts: http://www.stat.unc.edu/faculty/cji/fys/2012/Subprime%20mortgage%20crisis.pdf

    (1) The USA home ownership rate increased from 64% in 1994 (about where it had been since 1980) to an all-time high of 69.2% in 2004

    (2) During the two decades ending in 2001, the national median home price ranged from 2.9 to 3.1 times median household income. This ratio rose to 4.0 in 2004, and 4.6 in 2006.

    (3) USA household debt as a percentage of annual disposable personal income was 127% at the end of 2007, versus 77% in 1990.

    (4) Between 2004–2006, the share of subprime mortgages relative to total originations ranged from 18%–21%, versus less than 10% in 2001–2003 and during 2007.

    (5) Free cash used by consumers from home equity extraction doubled from $627 billion in 2001 to $1,428 billion in 2005

    (6) From 2001 to 2007, U.S. mortgage debt almost doubled, and the amount of mortgage debt per household rose more than 63%, from $91,500 to $149,500

    (7) Subprime mortgages amounted to $35 billion (5% of total originations) in 1994, (9%) in 1996, $160 billion (13%) in 1999, and $600 billion (20%) in 2006.

    In sum, the housing finance activity in the 2000s was an anomaly and it proved to be unsustainable. The essence of the unsustainability is the financial model required home prices to perpetually increase as, for most subprime and even many prime loans, incomes alone were insufficient to cover the cost of home ownership*. When home prices stopped increasing the financial model collapsed. And there is little NGDP or monetary policy could do to alter expectations of home prices.

    *The cost of a home includes much more than just mortgage principle and interest. There are property taxes, HOA fees, utilities, maintenance and repair. Increases in these costs, as well as rising interest rates in 2006-2007, can and did quickly erode the ability of marginal buyers to pay their loans. The promotion of home ownership served the agenda of politicians and industry but it did not serve the interest of the marginal home buyer.

  52. Gravatar of Saturos Saturos
    24. December 2015 at 08:23

    The IS-LM model is the classic way of combining those three relations between interest rates and monetary policy, and they still teach it. But people graduating economics degrees nonetheless seem to come away with the impression that low rates = easy money. And even Mankiw in his standard text that everyone reads suggests that it’s ok to think of interest rates varying in proportion to the tightness of money, as long as you read once his brief explanation that the Fed moves these rates by changing the money supply. Result: all economics commentators make no sense. (And of course, it’s likely that thinking in terms of fixed IS curves isn’t a good thing in the first place.)

  53. Gravatar of Gary Anderson Gary Anderson
    24. December 2015 at 08:58

    Kevin, the idea that in order to prosper we have to have 69 percent of the people owning homes is absurd. In Greece and Italy the percentage is higher and in Germany it is 42 percent and they have growth!! Of course, the Greeks and Italians look at the Germans as being poor because they rent. Go figure. I wrote an article on Business Insider a long time ago about this strange perplexity. http://www.businessinsider.com/maybe-this-is-why-merkel-does-not-want-ecb-as-bailout-machine-2011

  54. Gravatar of Ray Lopez Ray Lopez
    24. December 2015 at 11:09

    @Gary Anderson – German real estate does not appreciate much, compared to the UK, USA and even Greece in its heyday, so renting rather than owning makes more sense in Germany.

  55. Gravatar of Kevin Erdmann Kevin Erdmann
    24. December 2015 at 11:45

    Dan W., there is too much here to address in a comment. At my blog I have a long series addressing all of those points. Just for a teaser, I will note that even within your list, there are some problems. For instance the time frames of your point number 1 (rising ownership) and your point number 4 (rising subprime) are mutually exclusive. There is much less of a relationship between rising prices, rising ownership, and rising subprime than it appears at first blush.

    Gary, I think the optimal system would probably have low home ownership rates. I am explicitly opposed to homeowner tax benefits. I have not done any work with the goal of pushing homeownership higher for its own sake.

  56. Gravatar of collin collin
    24. December 2015 at 13:11

    What if low rates (given the minor reaction to the Fed rate increase) is only correlated to low rates? What if it were another variable driving the economy and rates? What if the demographic impact is much larger than assumed in terms of AD and AS?

    The more I witness the low rate economy today the more I accept the 1970s inflation/high rates were a function of the huge demographic expansion of labor supply? And now we are on the other side of that coin?

  57. Gravatar of collin collin
    24. December 2015 at 13:12

    I meant:

    What if low rates (given the minor reaction to the Fed rate increase) is only correlated to low growth?

  58. Gravatar of TravisV TravisV
    24. December 2015 at 14:40

    Outrage. Outrage!! :)

    http://www.vox.com/2015/12/23/10657762/women-products-cost-more

    “The hidden tax women pay on just about everything”

  59. Gravatar of Gary Anderson Gary Anderson
    24. December 2015 at 15:42

    You know why it does not appreciate, Ray? German banks are not permitted to offer German citizens easy money loans. They only are allowed to offer the PIIGS nations easy money loans. That means that Germany protects its citizens, admitting that these loans are evil. But it takes advantage of the citizens of other nations, even the USA. And our banks do as well!

  60. Gravatar of ssumner ssumner
    24. December 2015 at 15:57

    Kevin, Interesting, as usual.

    Britonomist, I have carefully explained the causal mechanism that makes lower rates contractionary. If you have a problem with the causal mechanism I outlined, what is it? Just pointing to IS-LM doesn’t persuade me, as I find the model to be almost useless.

    Gary, You said:

    “Even Larry Summers said there is a shortage of long bonds.”

    What does the term ‘shortage’ mean? In EC101, it means quantity demanded exceeds quantity supplied at the current price. Is that your claim?

    John, This blog is full of empirical evidence that monetary stimulus is effective at the zero bound. The best recent example is probably Japan, where monetary stimulus drove the yen sharply lower.

    I agree with your last point about real rates. But a highly contractionary policy at full employment might reduce real rates quicker than you’d expect.

    Ray, Hmm, I noted that Nick saw a connection with IS-LM, and then said I was not using IS-LM. Fascinating.

    Saturos, As you know I hate the IS-LM model, but I think you are right, the actual problem is how it’s used. Guns don’t kill people, . . .

    IS-LM models don’t kill economies, . . .

    Travis, Those naughty, naughty corporations.

  61. Gravatar of Gary Anderson Gary Anderson
    24. December 2015 at 16:52

    Scott, the relentless decline in yield over time since 1984 shows a pickup in demand over time for long bonds. There has been some manipulation of long bond prices downward and that may explain the difference between our long rates and those of Europe. The Globe and Mail reported a shortage of bonds, and Washington’s blog reported a lawsuit over manipulation of the price of bonds. But downward yield over time indicates increase in demand on the long end, which is controlled by supply and demand most times.

  62. Gravatar of Gary Anderson Gary Anderson
    24. December 2015 at 16:56

    Sorry, the Globe and Mail reported a shortage of bond buyers, which corresponds to the lawsuit we saw at Washington’s blog. I am unable to post the links I guess. You can search google for them. In reality, based on yield, there appears to be an increasing demand for bonds, relative to supply, over time, and that in spite of manipulation to the contrary.

  63. Gravatar of TravisV TravisV
    24. December 2015 at 17:14

    Merry Christmas, Prof. Sumner!

  64. Gravatar of TravisV TravisV
    25. December 2015 at 07:19

    “More from Robin Hanson on sexist pricing, which might just be sampling bias”

    http://marginalrevolution.com/marginalrevolution/2015/12/more-from-robin-hanson-on-sexist-pricing.html

  65. Gravatar of TravisV TravisV
    25. December 2015 at 07:22

    Awesome comment section here:

    http://www.interfluidity.com/v2/6287.html

    “Home is where the cartel is”

  66. Gravatar of Gary Anderson Gary Anderson
    25. December 2015 at 09:12

    Rent controls are the answer for a city with severe earthquake risk, like SF.

  67. Gravatar of Daniel Harris Daniel Harris
    25. December 2015 at 22:49

    “I’m trying to convince people that market interest rates don’t reflect central bank policy, they reflect the economy.”

    I had read over the previous two posts several times, and couldn’t get my head around them.

    Then I read this sentence and it all clicked into place. Thank you. I had been so, so confused.

  68. Gravatar of Postkey Postkey
    26. December 2015 at 01:38

    ” . . . you should note – the central bank controls the short-term interest and all the rest of the rates largely follow suit.
    In effect, the central bank sets the interest rates at the short-end of the yield curve and the term structure follows expectations of inflation risk rather than default risk.
    The “balance sheet” operations of the central bank can then control any other interest rate at any maturity that the bank desires. How?
    The central bank could announce, for example, a ceiling on a longer-term yield, which was below the current market rate if it wanted the longer-term – investment rates – lower.
    It could enforce that aim by being prepared to purchase the targeted assets at the desired yield.
    Mainstream economists repeatedly claim that this strategy would have to be ratified by the market for it to work. That is, the only way the central bank could cap yields on longer-term assets is it the target yield was consistent with the market expectations of the yield.
    Which, of course, is pure nonsense. These economists mostly use so-called ‘frictionless financial market’ models where there is no time or transaction costs and everyone has perfect information and equal access. That is, they count angels on the top of pin heads.
    Clearly that sort of model has nothing to say about the real world we live in.
    The only consequence of a discrepancy between the targeted yields and the market expectations of future yields would be that the central bank would end up holding all of the targeted asset and the bond traders would not buy at all because they thought yields would have to rise and bond prices fall.
    My assessment of that outcome – excellent.
    The bond traders might boycott the issues and ‘force’ the central bank to take up all the volume on offer. So what? This doesn’t negate the effectiveness of the strategy it just means that the private buyers are missing out on a risk-free asset and have to put their funds elsewhere. Their loss!
    Eventually, if the government bond was the preferred asset the bond traders would learn that the central bank was committed to the strategy and would realise that if they didn’t take up the issue the bank would. End of story – the rats would come marching into town piped in by the central bank resolve.
    The BIS paper concurs that the central bank holds all the cards in this regard.’ ”

    http://bilbo.economicoutlook.net/blog/?p=32029#more-32029

  69. Gravatar of ssumner ssumner
    26. December 2015 at 07:50

    Travis, Merry Christmas to you as well.

    Thanks Daniel.

  70. Gravatar of Britonomist Britonomist
    26. December 2015 at 09:00

    Merry (late) Christmas

    “Britonomist, I have carefully explained the causal mechanism that makes lower rates contractionary. If you have a problem with the causal mechanism I outlined, what is it? Just pointing to IS-LM doesn’t persuade me, as I find the model to be almost useless.”

    I have multiple objections, but to avoid a huge quote mountain I’ll start with just one. In the previous post you said:

    “When you don’t spend you save, and saving goes into investment, which is also part of GDP.”

    In a modern, non gold-standard economy this is just incredibly nonsensical. The problem must be that we’re using stupidly overly simplified models which don’t distinguish between the nominal rate the central bank sets, and the marginal product of capital. So, if the interest rate (i.e. marginal product of capital) rises, of course that will encourage more investment because the return on investment becomes greater.

    The problem is that if the central bank were to raise interest rates right now it WOULD NOT raise the marginal product of capital. Therefore this increased saving would be directed to government securities, NOT capital goods – you would not see a corresponding rise in investment. Any modern financial model that models returns from government securities as distinct from returns from risky assets will ALWAYS show a portfolio shift away from risky investments towards the ‘risk free’ rate (i.e. government bonds) – in other words rising the risk free rate ALWAYS REDUCES investment in capital. And if you read papers from central banks that go into detail about their transmission mechanism – rising rates reduces lending and credit creation, and increases holdings in government securities. Increasing the central bank interest rate reduces investment in capital ceteris paribus, PERIOD. ONLY when the marginal product of capital increases will investment increase, and raising the central bank rate will not raise the marginal product of capital.

  71. Gravatar of Scott Sumner Scott Sumner
    26. December 2015 at 12:50

    Britonomist, I think you’ve missed the whole point of the post. You are suggesting that if the central bank takes actions that cause interest rates to rise, the effect will be contractionary. Yes, that’s usually true, but I was looking at changes in interest rates caused by factors other than the central bank.

    And when the central bank causes higher interest rates, the increase in the rates themselves are contractionary, but the thing that cause the higher rates is expansionary (i.e. more money.) Or are you claiming that higher interest rates reduces V. Is that your claim?

    You said:

    “Therefore this increased saving would be directed to government securities, NOT capital goods – you would not see a corresponding rise in investment.”

    No, S=I is an identity. Check out any economics textbook. If ex post saving rises then investment must rise. Buying government bonds is not “saving” in the sense that economists define the term. That’s because the person that sold you the bond dissaves an identical amount.

    You need to think very carefully about how to model velocity. Right now you are trying to use a Keynesian approach, which almost inevitably leads to error, unless you are Paul Krugman, or one of the other very few people who actually knows how to use the model.

  72. Gravatar of Britonomist Britonomist
    26. December 2015 at 13:49

    “Or are you claiming that higher interest rates reduces V. Is that your claim?”

    Not quite, I’ve always found money velocity and money demand unclear. First of all it depends on which interest rates – this is the whole problem when using an overly generalized model. If by ‘interest rates’ we mean the nominal federal funds rate or government bond yield, then yes a ceteris paribus rise in those rate will decrease V, but remember this is holding base money constant, if base money also falls – since consumption is probably somewhat sticky, the fall in base money might be a larger proportion than the fall in V, causing *measured* V to rise in that case. On the other hand, if by ‘interest rates’ you actually mean the marginal product of capital or some kind of mean return on risky assets, then this may be expansionary if it encourages more spending on ‘capital goods’ and less demand for risk free savings. It’s just not useful to describe the economy in terms of one global interest rate anymore. It’s much easier if we talk about the ‘risk-free rate’.

    ‘”No, S=I is an identity.”‘

    When this last debate came up, it seemed very intelligent economists/commentators disagreed and debated the extent to which it was an identity, or simply an equilibrium condition (with even Krugman getting involved http://krugman.blogs.nytimes.com/2012/01/15/the-method-of-comparative-statics-very-very-wonkish/ ). In any case, it depends on how you define interest rates, as well as how you define saving.

    ” Buying government bonds is not “saving” in the sense that economists define the term.”

    In microeconomics, saving is defined as deferred consumption (with no implication for capital spending other than in Ramsey type models which are macro not micro), there is no standard definition of ‘savings’ for economists. If you’re not defining buying risk free government securities as ‘saving’, then fine – but it doesn’t matter what you call it. A higher risk free i WILL result in people buying government securities, people deleveraging, and even just people leaving more money in a savings accounts. NONE OF THIS necessarily leads to an increase in investment, if this isn’t an increase in saving by your definition it doesn’t matter – it will still happen if the risk free rate increases, reducing consumption and causing V to drop (again unless M drops with it, which usually happens).

  73. Gravatar of Britonomist Britonomist
    26. December 2015 at 14:31

    I flicked through my copy of David Romer’s ‘Advanced Macroeconomics’ which I was using during my masters. It’s interesting to note that velocity as a concept simply isn’t discussed, it doesn’t even have an entry in the index – this is why such discussion might be so confusing to modern macroeconomists because they just don’t use that kind of language any more.

    Anyway, I looked at sections covering interest rates in the NK models in the chapters before monetary policy is introduced – what I find is that in modern NK models a rise in interest rates is associated with a reduction in investment and an ambiguous effect on consumption (because of competing income and substitution effects, although this is lifetime consumption – first period consumption will certainly fall). I think it’s safe to say that a fall in investment and an ambiguous medium term effect on consumption (and fall in short term consumption) should be considered a ‘contractionary’ effect overall. So, unless you disagree with my assessment of NK models here, your claims do seem at least to be at odds with modern New Keynesian macro – of course you may not be a fan of New Keynesian models. I’d certainly regard NK models as more informative than old fashioned IS/LM however.

  74. Gravatar of Postkey Postkey
    27. December 2015 at 01:27

    In a ‘simple’ Keynesian model.

    I ≡ S is an ex post accounting identity. Say 2014.

    The equation
    I = S is ‘derived’ from the accounting identity, but the equality only holds when the when the economy is in equilibrium. That is, when planned aggregate demand {AD} is equal to planned income/output {Y}.

    If I is increased {to I1} when the economy is in equilibrium {Y = Yo, Io = So}, then the equality no longer holds.
    AD will increase {ceteris paribus} leading to an increase in Yo.
    Yo will increase until a new equilibrium is reached at, say, Y1.
    { Y1 > Yo}.
    The above equation will then hold, but with different {higher} value of S. {I1 = S1}.

  75. Gravatar of Ray Lopez Ray Lopez
    27. December 2015 at 07:55

    Britonomist is right; Sumner is wrong.

    Sumner: “No, S=I is an identity.” No it’s not an identity.
    From Blanchard’s Macroecon 101 textbook:

    S = I + G – T
    Or, if G-T = zero (no government activity, aka “public savings”), “ (private)Savings = Investment”

    When’s the last time G-T was zero? I don’t think since forever. Sumner wrong again.

  76. Gravatar of Ray Lopez Ray Lopez
    27. December 2015 at 08:05

    @myself: and let’s not ignore exports and imports

    If S = I
    then G – T = M – X

  77. Gravatar of Ken Simpson Ken Simpson
    27. December 2015 at 15:06

    Scott:

    You replied, “From August 2007 to May 2008 the Fed did nothing to lower interest rates, they fell in the free market, on their own accord. The Fed simply adjusted it’s target rate to market conditions.”

    Does the same hold true today, that the Fed did not just raise rates, that they simply adjusted its target rate to market conditions? Janet Yellen tried to imply some reading of the natural rate, but floundered.

    In the credit market, the Fed Funds rate competes with lending rates. Each time the Fed adjusts the FF rate, banks FOLLOW, and adjust their Prime rates accordingly, usually the same day.

    I think the evidence is crystal clear that the Prime Rate, and thus other rates tied to Prime, follow the Fed Funds rate.

  78. Gravatar of ssumner ssumner
    27. December 2015 at 17:22

    Britonomist, I don’t follow your first paragraph. You said:

    “If you’re not defining buying risk free government securities as ‘saving’, then fine”

    Sure, for the individual, but it’s equally dis-saving for the seller of the bond. No aggregate saving. There’s a difference between micro and macro.

    You said:

    “what I find is that in modern NK models a rise in interest rates is associated with a reduction in investment”

    I find that hard to believe. That would be reasoning from a price change, a basic EC101 error.

    You said:

    “So, unless you disagree with my assessment of NK models here”

    I do.

    Postkey, You said:

    “I = S is ‘derived’ from the accounting identity, but the equality only holds when the when the economy is in equilibrium.”

    Nope, check any EC101 textbook. Ex post it’s an identity, not an equilibrium condition.

    Ray, Like a child playing with toys.

    Ken, No, I have a recent post on that, either here or Econlog. They raised IOR, which is a tight money policy that caused market rates to rise. It’s about a week ago post.

  79. Gravatar of Britonomist Britonomist
    27. December 2015 at 18:34

    “Sure, for the individual, but it’s equally dis-saving for the seller of the bond. No aggregate saving. There’s a difference between micro and macro.”

    In response to higher interest rates, if everyone deleverages/pays down their debts/borrows less, and others just keep more money in their bank/risk free savings account – this will reduce consumption but will not (necessarily) increase investment – or at least you’ve provided no obvious mechanism for this to occur. The reason the S = I identity might still hold is because this forces companies to keep more of their stock in inventory, and then you can call this ‘investment’, but that won’t increase output.

    “I do.”

    If you don’t believe me just read Romer’s ‘Advanced Macroeconomics’ for yourself (mine is 4th edition). Try page 422 for a start:

    ‘Our model implies that increases in expected future short term rates reduces investment’

    ‘A permanent fall in the interest rate (…) produces a temporary boom in investment as the industry moves to a permanently higher capital stock’

    There are so many mechanisms as well. Have you read much into financial economics? As I mentioned previously, a higher risk free rate will cause investors to re-balance their portfolios away from risky assets and into government bonds. A higher expected future risk free rate increases the discount rate of future earnings, making investments less profitable – reducing investment into risky assets. I just never see a mechanism in any economic models where a higher risk free rate INCREASES investment, it ALWAYS reduces investment.

  80. Gravatar of Postkey Postkey
    27. December 2015 at 23:56

    ‘Postkey, You said:

    “I = S is ‘derived’ from the accounting identity, but the equality only holds when the when the economy is in equilibrium.”

    Nope, check any EC101 textbook. Ex post it’s an identity, not an equilibrium condition.’

    I said.
    In a ‘simple’ Keynesian model.

    I ≡ S is an ex post accounting identity. Say 2014.

    ex ante I = S is an ‘equilibrium’ condition.

  81. Gravatar of Scott Sumner Scott Sumner
    28. December 2015 at 11:26

    Britonomist, Your comments on saving and investment simply are not useful. You cannot examine causal factors in macroeconomics by looking at differences between saving and investment. It’s a dead end. What matters is NGDP and nominal hourly wages. Consumption and investment are endogenous.

    I am traveling, so I have no access to the textbook you cite. I would need to see the context.

    All economists agree (I think) that investment tends to be higher in years with higher interest rates (like 2000 and 2006) than years with low rates (like 2001 and 2009) I’d have to see the context. Is he making any assumptions about what causes interest rates to change? I.e. what if the investment schedule shifts right (or the IS curve, if you prefer). It’s nonsensical to talk about the effects of a change in i without discussing what you are holding constant.

    Postkey, I never claimed ex ante is the same, so what is your point?

  82. Gravatar of Britonomist Britonomist
    28. December 2015 at 12:29

    For some reason I’m unable to send my comment, I think this blog automatically blocks comments with links to certain sites, so I’ll have to try replace the URL with a google query:

    “Britonomist, Your comments on saving and investment simply are not useful.”

    But it was you that stated that what is not spent is invested, implying that any drop in consumption would be offset by a rise in investment. That’s just not a good argument, as Wren-Lewis explains: (google ‘Simon Wren-Lewis Savings Equals Investment?’)

    “All economists agree (I think) that investment tends to be higher in years with higher interest rates (like 2000 and 2006) than years with low rates (like 2001 and 2009) I’d have to see the context. ”

    Sure, but I’m certain these economists would also agree that the causality is reversed – greater investment will lead to higher rates, not the other way around.

    “Is he making any assumptions about what causes interest rates to change? ”

    No, the variable just changes value, temporarily or permanently.

  83. Gravatar of Gary Anderson Gary Anderson
    29. December 2015 at 08:09

    “If ex post saving rises then investment must rise. Buying government bonds is not “saving” in the sense that economists define the term. That’s because the person that sold you the bond dissaves an identical amount.”

    Dealers make fees on selling bonds. And if the Fed sells bonds they shrink the balance sheet, but a central bank isn’t a saver is it? Its balance sheet is not counted as a savings vehicle is it, Scott?

  84. Gravatar of Postkey Postkey
    29. December 2015 at 10:33

    “If ex post saving rises then investment must rise. . . . ”

    Can ex post savings increase?

  85. Gravatar of Scott Sumner Scott Sumner
    29. December 2015 at 19:17

    Britonomist, You said:

    “Sure, but I’m certain these economists would also agree that the causality is reversed – greater investment will lead to higher rates, not the other way around.

    “Is he making any assumptions about what causes interest rates to change? ”

    No, the variable just changes value, temporarily or permanently.”

    I hope you see the contradiction here.

    And yes, I’ve debated Wren-Lewis on savings/investment, I’m quite aware of the Keynesian view. That has no bearing on what I’m saying here.

  86. Gravatar of Britonomist Britonomist
    30. December 2015 at 03:31

    “I hope you see the contradiction here.”

    There is none, these are different scenarios and different people. One is a result in a model but BEFORE monetary policy is introduced so no central bank reaction function – another is simply a statistical correlation which can be explained by a monetary policy reaction function. There is no contradiction – saying that investment booms will lead to the central bank raising interest rates does not in any way contradict the model result.

    “And yes, I’ve debated Wren-Lewis on savings/investment, I’m quite aware of the Keynesian view. That has no bearing on what I’m saying here.”

    This isn’t just the Keynesian view, it’s the EC101 view. That post was specifically aimed at helping undergraduate students understand their course.

  87. Gravatar of ssumner ssumner
    31. December 2015 at 10:16

    Britonomist, You keep talking about central banks, but suppose they do not exist. Any good macro model (that claims to be fundamental) should be equally applicable to a world with no central bank.

    If you don’t make any assumptions about why interest rates change, you can’t draw any conclusions about whether the shocks are expansionary or contractionary. For instance, if rates rise and the money supply is held constant, the effect is expansionary. If the rise is rates is CAUSED by a reduction in the money supply, the net effect is contractionary. Just talking about interest rates in a vacuum, with no assumptions about the money supply, are meaningless.

    This would be much easier to see if you stopped thinking about monetary policy in terms of changes in interest rates. Unless you want to claim policy was expansionary in 1929-30.

    On your other point, most EC101 books do a poor job of explaining the model, but at least they get right the S=I identity.

  88. Gravatar of Britonomist Britonomist
    31. December 2015 at 11:12

    Happy New Year

    “Just talking about interest rates in a vacuum, with no assumptions about the money supply, are meaningless.”

    Why is it meaningless? It’s certainly not meaningless in the context of this model. And I deliberately chose models in the book BEFORE monetary policy/central banks are introduced. You seem like an MMTer – overly focused on accounting identities but ignoring preferences and behaviour, since people are risk averse – higher interest rates leads to contractionary *behaviour*, it reduces the preference for investment into risky assets, reduces first period consumption – these are not expansionary effects.

    And as far as I can tell, taking an accountants approach: holding money supply constant and assuming savings identity, this must mean a drop in output – as it’s the only way for the I = S identity to hold. As SWL explains: a drop in consumption leads to an increase in inventory capital which maintains the I = S identity – this increase is actually *contractionary* as it causes incomes and profits to fall, as they cutback on their output people lose income causing their savings to fall, this happens until no more inventory is being accumulated and the economy reaches a new equilibrium point at a lower output (unless the interest income is so high this offsets the drop in C or S, but I have no idea how this could be possible while keeping the money supply constant).

    ” For instance, if rates rise and the money supply is held constant, the effect is expansionary.”

    More likely the *effect* is not expansionary, rather something else is causing expansion in the economy leading interest rates to rise. What you’re doing appears to basically be Neo-Fisherism even if you claim otherwise. It seems you’re only saying it must be expansionary because it’s the only way for some accounting identity to hold – even though there’s no behavioural reasoning, just like Neo-Fisherism.

  89. Gravatar of Scott Sumner Scott Sumner
    1. January 2016 at 12:55

    Britonomist, You said:

    “You seem like an MMTer – overly focused on accounting identities but ignoring preferences and behaviour, since people are risk averse – higher interest rates leads to contractionary *behaviour*, it reduces the preference for investment into risky assets, reduces first period consumption – these are not expansionary effects.”

    You have this exactly backward. I’m the one talking about behavioral relationships. I am the one explaining how i affects V. You seem to ignore money demand behavior.

    I have no idea what point you are making with your SWL reference. I’ve said many times that a shift in the saving schedule to the right is contractionary if M is fixed. So what is your point? After all, when S shifts to the right interest rates fall. That’s contractionary.

  90. Gravatar of Britonomist Britonomist
    1. January 2016 at 17:26

    “You have this exactly backward. I’m the one talking about behavioral relationships. ”

    I gave you behavioural examples from more recent New Keynesian model as well as examples from financial economics. The agents in those models exhibit contractionary behaviour.

    “I am the one explaining how i affects V. You seem to ignore money demand behavior.”

    As far as I can tell, the model you cite only applies to gold/commodity money. It assumes gold to have a ‘marginal dividend’ from its non monetary uses, you cannot make that assumption for fiat cash. Their model appears to constantly rely on the demand for ‘non monetary gold stock’, there’s no such thing ‘non monetary cash’, I don’t see how you can apply this model to a modern fiat money economy. It also seems to consolidate both the return on capital and the return on bonds.

    Going back to your OP:

    “causation went from falling interest rates to higher demand for base money to recession in 2007-08.”

    The higher demand for base money from banks was because of the huge financial crisis and the huge asset write-offs causing cash-flow problems. Imagine how much more contractionary this would have been if the banks were forced to borrow this money at higher interest rates – meanwhile contrary to popular opinion the general public did not start withdrawing their money as cash em masse. I just don’t see how you can draw parallels to the demand for gold holdings during the early 20th century or how that model applies. Do you have a non gold standard model?

  91. Gravatar of ssumner ssumner
    2. January 2016 at 18:24

    Britonomist, You said:

    “I gave you behavioural examples from more recent New Keynesian model as well as examples from financial economics. The agents in those models exhibit contractionary behaviour.”

    No, the only way to evaluate whether the behavior of an agents is expansionary or contractionary is to look at how their actions impact velocity. Decisions to spend more or less, or save more or less, have no impact on NGDP except to the extent they impact velocity.

    You said:

    “As far as I can tell, the model you cite only applies to gold/commodity money. It assumes gold to have a ‘marginal dividend’ from its non monetary uses, you cannot make that assumption for fiat cash. Their model appears to constantly rely on the demand for ‘non monetary gold stock’, there’s no such thing ‘non monetary cash’, I don’t see how you can apply this model to a modern fiat money economy. It also seems to consolidate both the return on capital and the return on bonds.”

    Not at all. The demand for fiat money is negatively related to the opportunity cost of holding fiat money, which is the nominal interest rate (or the interest rate minus IOR for bank reserves.)

    On your final point, I see zero evidence that the financial crisis increased the demand for base money in late 2007 and early 2008. Indeed I’m surprised that the fall in interest rates did not cause an even bigger increase in real money demand, there had to be other factors reducing the demand for base money during that period. So it makes no sense to look for other factors that might have increased money demand.

    My “model” is very simple, and uncontroversial—velocity is positively related to nominal interest rates. Are you claiming that higher nominal interest rates would reduce velocity?

  92. Gravatar of Britonomist Britonomist
    2. January 2016 at 19:47

    “No, the only way to evaluate whether the behavior of an agents is expansionary or contractionary is to look at how their actions impact velocity. Decisions to spend more or less, or save more or less, have no impact on NGDP except to the extent they impact velocity.”

    Not in New Keynesian economics – velocity is basically not mentioned in the literature, something is expansionary in NK if it increases output. But even if we do focus on velocity, how would something that lowers investment and first period consumption increase velocity, where would the increase come from?

    “Not at all. The demand for fiat money is negatively related to the opportunity cost of holding fiat money, which is the nominal interest rate (or the interest rate minus IOR for bank reserves.)”

    Do you mean cash here? I think for almost everyone the demand for cash is extremely inelastic in a modern economy. Lowering interest rates won’t cause people to rush to their bank account to withdraw money. These days cash is used to facilitate small payments, unless you’re a drug dealer. Also why are you separating IOR from ‘the interest rate’ in general? That begs the question of what exactly ‘the’ interest rate is now – is it the return on short term government bonds, raising those rates will just lead banks to exchange their reserves for government bonds – ah, so are you saying that *because the money supply must remain constant* the central bank will have to re-pump the money it sucks in from bond sales back into the economy so that the base remains constant – is that in fact your mechanism? How could it possibly do that without buying back bonds and therefore lowering interest rates though? This whole discussion becomes incoherent then because it presents an impossible situation.

    ” Are you claiming that higher nominal interest rates would reduce velocity?”

    Firstly I maintain it would reduce investment and consumption, but this could still raise velocity if the drop in output is not as large as the drop in the money supply. However you’re also giving this “keeping the money supply constant” condition which makes things a lot more complicated – you’re right that velocity must increase via the hot potato effect in a base money economy (this analysis breaks down when you add broad money to the mix but I’ll ignore that here for now) because there is nothing to soak up the surplus money – but this would basically require the bizarre assumption that the central bank or government would be desperate also to get rid of this surplus money but would have to find a way to do so without buying bonds back from central banks (otherwise the interest rate would lower again) – does that make sense?

  93. Gravatar of Britonomist Britonomist
    2. January 2016 at 20:23

    Last sentence should read “without buying bonds back from banks” (not central banks).

  94. Gravatar of ssumner ssumner
    3. January 2016 at 10:25

    Britonomist, You said:

    “Not in New Keynesian economics – velocity is basically not mentioned in the literature, something is expansionary in NK if it increases output.”

    That could be a supply shock or a demand shock. It’s much more useful to distinguish between things that cause NGDP to change (demand shocks) and things that raise RGDP for any level of NGDP (supply shocks). Recall the mistake Keynesians made in arguing that austerity had slowed growth in Britain, when the slowdown was actually due to low productivity, which is unrelated to AD.

    You said:

    “Do you mean cash here? I think for almost everyone the demand for cash is extremely inelastic in a modern economy. Lowering interest rates won’t cause people to rush to their bank account to withdraw money.”

    Oh really? Then why did the demand for cash soar after 2008, when interest rates fell to zero?

    I don’t even understand the rest of your comment, it’s gibberish to me.

  95. Gravatar of Britonomist Britonomist
    3. January 2016 at 10:55

    “That could be a supply shock or a demand shock.”

    So what do you class an interest rate ‘shock’ as? In an NK DSGE model a shock is an unexpected exogenous change to one variable (usually interest rate), the impulse response is the result, if Y increases it’s expansionary.

    “Oh really? Then why did the demand for cash soar after 2008, when interest rates fell to zero?”

    I’d like to see your data if possible? Was this literal cash or was it banks demanding more base money? I think those are separate issues.

    Also the rest of my post was me trying to work out what your model is, let me try again more succinctly:

    Is your model basically a hot potato effect? Higher interest rates leads to people wanting to get rid of their cash more quickly because of the higher opportunity cost – this increases velocity, and is expansionary if the money supply is held constant because the central bank cannot soak up the surplus cash and lower the money supply in that scenario. Is that your model?

  96. Gravatar of Scott Sumner Scott Sumner
    4. January 2016 at 12:44

    Britonomist, By that logic a rise in oil prices would be regarded as a “shock”, regardless of whether demand doubled or supply fell in half. Is that really your claim? It doesn’t matter why interest rates changed? In my view a change in interest rates is the effect of a more basic shock, such as a shift in IS or LM (if you insist on Keynesian language).

    Your last paragraph is correct, move the rest of the discussion to my more recent post.

  97. Gravatar of Britonomist Britonomist
    5. January 2016 at 10:24

    “Britonomist, By that logic a rise in oil prices would be regarded as a “shock”, regardless of whether demand doubled or supply fell in half.”

    Economic models are not closed comprehensive universal models of everything – at some point you’re simply going to have to ‘shock’ a variable to see what the response is. Suppose instead we don’t shock the price of oil but the supply of oil – you could just apply the same problem: doesn’t it matter if the supply dropped due to higher cost of extraction or lower income of oil producers? You can go back infinitely asking ‘why’ this happened, but you’ll never be able to answer any economic question if you do. Ceteris paribus means that you change the value of one variable, while leaving the values of all other variables the same – in this case you change the value of i_t from say 2% to 5%, and see what happens to Y_t and pi_t in response. You don’t assume a bunch of other variables are changing first in order to force i_t up, you’re simply increasing i_t and leaving all other variables unchanged.

    “It doesn’t matter why interest rates changed?”

    In the DSGE models I’m familiar with it matters whether the change is anticipated or unanticipated – but *the* interest rate is assumed to either be exogenous or endogenous only by a central bank reaction function. In my earlier posts I avoided looking at models with a central bank reaction function.

  98. Gravatar of Sumner Arguing That Low Interest Rates Are Contractionary Sumner Arguing That Low Interest Rates Are Contractionary
    6. January 2016 at 09:03

    […] has been posting some interesting stuff on this topic (one, two, three). Many of you–and Scott himself, at the tail end of this post–are bracing for me […]

Leave a Reply