Greenspan vs. Taylor

Frank McCormick sent me the following WSJ article:

He said he is baffled by all the blame that has been piled on him. Since the recession, critics have said the increased money supply and low interest rates during his tenure at the Fed from 1987 to 2006 led to bubble investments. Mr. Greenspan first heard that theory, he says, in 2007, when John Taylor, a professor of economics at Stanford University who has advised Republicans, made the connection between easy money and the housing bubble. “It had absolutely nothing to do with the housing bubble,” he says. “That’s ridiculous.”

Instead, he says Prof. Taylor’s statement “served a lot of political purposes of people who have been picking on the Fed from both sides of the aisle.” Mr. Greenspan wrote a rebuttal in a paper for the Brookings Institution, going through Prof. Taylor’s points one by one. “I thought, that’ll kill it,” he remembers. “It didn’t, because nobody read the paper.”

Mr. Greenspan said he didn’t press the issue because Prof. Taylor is a friend, but he had no idea how far Prof. Taylor’s idea would go. “The trouble, unfortunately, with the argument is there’s no evidence that happened, but he’s won the battle, and his view is conventional wisdom,” he says.

Prof. Taylor stands by the paper in which he presented the idea. “The paper provided empirical evidence…that unusually low interest rates set by the Fed in 2003-2005 compared with policy decisions in the prior two decades exacerbated the housing boom,” he wrote in an email. Other economists have corroborated the findings, he added, and “the results are quite robust.”

Let’s investigate this like forensic scientists.  Common sense suggests that the housing “bubble” might have had at least three causes:

1.  Low interest rates.

2.  Bad regulation.

3.  Irrational exuberance among bankers and homebuyers.

By “bad regulation” I don’t mean any one thing, but a whole range of interventions that encourage excessive mortgage debt:  interest deductibility, the GSEs, too big to fail, the CRA, the moral hazard created by FDIC, etc, etc.

Of course low interest rates are not really an explanation, as one should never reason from a price change. If the low interest rates were caused by weak business investment, as in 2002, then they will tend to boost housing. But in that case it makes more sense to talk about low business investment boosting housing. After all, if the low rates are caused by the housing bust (as in 2008-09) then they obviously wouldn’t boost housing.

So which is it, were the low rates of the bubble years caused by easy money, or by weak business investment?  Let’s start with some basic principles.  If the Fed successfully targets inflation or NGDP then interest rates are 100% endogenous. In that case any interest rate –> housing causality would not involve monetary policy at all, at least if one assumes the macro targets were appropriate. If inflation and NGDP were on target during the housing bubble, then Taylor would be 100% wrong and Greenspan would be 100% right.

Reasonable people can read the evidence differently, but here’s what seems plausible to me:

1.  During 2002-04 Greenspan was 100% right and Taylor was 100% wrong.  The macro variables did not indicate an excessively expansionary monetary policy (from a dual mandate perspective), and hence the low rates were due to the business investment drought post-2000.  And perhaps partly the other “X-factors” driving real rates steadily lower for decades.

2.  During 2005-06 Fed policy might have been a bit too expansionary.  Marcus Nunes says it wasn’t and David Beckworth says it was.  Both have reasonable arguments.  NGDP growth and inflation were clearly a bit above target, but Nunes would reply that NGDP was close to target using a level targeting framework.

My view is that this isn’t an important argument.  I really don’t care whether money was about right during 2005-06, or slightly too easy.  Either way it wasn’t at all unusual compared to earlier periods of our history. Indeed during most of my life policy was far more expansionary during cyclical expansions than 2002-06.

To conclude, three factors seem to have contributed to the housing bubble.  The Fed may or may not have played a small role in one of the three factors.  Probably not at all during 2002-04, maybe a little bit in 2005-06.  Overall I’d say Greenspan is around 90% right and Taylor is perhaps 10% right, but only for the years 2005-06.  And that’s being generous to Taylor.  If Nunes is correct then Greenspan is 100% right.

PS.  Greenspan is right that Taylor has won the battle.  Just as those who thought easy money in the 1920s led to the Great Depression had won the battle by 1930. But of course they eventually lost the war, after Friedman and Schwartz showed that the real problem was tight money after 1929, not easy money before 1929.

PPS.  The most interesting part of the article is the opening:

Alan Greenspan, the former chairman of the Federal Reserve, goes to a lot of parties. He and his wife, the TV journalist Andrea Mitchell, “sort of get invited everywhere,” he says, sitting in front of the long bay window in his office on Connecticut Avenue in Washington, D.C. Lately, though, cocktails and dinners seem to have guest lists drawn almost exclusively from one political party or the other. “It used to be a ritualistic 50-50 at parties””the doyennes of culture and partying were very strict about bipartisanship,” he adds. “That doesn’t exist anymore.”

That’s right, the current members of Congress and other top officials are so emotionally immature that they are incapable of socializing with people of different ideologies.  What a bunch of babies!!

Update:  Here’s Marcus Nunes’ post on this topic.


Tags:

 
 
 

22 Responses to “Greenspan vs. Taylor”

  1. Gravatar of Saturos Saturos
    20. October 2013 at 05:55

    The low rates could also have been caused by a global savings glut, as many suppose.

  2. Gravatar of ssumner ssumner
    20. October 2013 at 06:42

    Saturos, Yes, that was the “X factor” I referred to, or at least one of them.

  3. Gravatar of Jason Jason
    20. October 2013 at 09:37

    “That’s right, the current members of Congress and other top officials are so emotionally immature that they are incapable of socializing with people of different ideologies.”

    Only the ones that invite Alan Greenspan to their parties.

  4. Gravatar of Tom Tom
    20. October 2013 at 10:02

    See Beckworth’s blog post on inflation targeting:
    http://macromarketmusings.blogspot.sk/2013/09/at-least-fed-has-inflation-target-right.html

    2004-2007 inflation over 2% (tho max 2.4 is far less than 3).
    This “too loose” is totally correct — but by a very small amount.

    What does “tight money” mean in practice? That those who want loans don’t get them.

    Interest rates are usually, but not always, a good single number proxy for that. Not since 2008 tho.

    Like you say, we have tight money. But the way to show that is to see how many loan applications are being approved and funded, vs how many are being rejected. Or perhaps other loan-based metrics. But those are not the numbers we get from most macro-economists.

    Why no, prof Sumner? Why don’t monetarists focus on how much loan money is asked for and given out by banks?

    The unprecedented CDO crash and loans based on MBS has to be included in the total, somehow — and that’s why house investment money was super easy thru 2006.

    Just like cooking stuff burns at the bottom, avg temperatures don’t capture all the important stuff in cooking; nor does even NGDP capture all the important stuff in the real economy.

  5. Gravatar of Tommy Dorsett Tommy Dorsett
    20. October 2013 at 10:19

    Scott – Base growth slowed into the low single digits by mid-2005; moreover, the Fed inverted the yield curve in 2006, leading base growth to a virtual halt in 2007. Hard to square these phenomena with a too-easy Fed, especially because NGDP began to slow sharply after the curve pancaked:

    http://research.stlouisfed.org/fred2/graph/?g=ny8

    http://research.stlouisfed.org/fred2/graph/?g=nyb

    And the housing bubble, which inflated between 1998-2006, doesn’t really fit the timeline very well.

    Also, Taylor’s Rule suggested the Fed should have hiked Fed funds to 6% in the summer and fall of 2008. I wonder how that would have worked out.

  6. Gravatar of ssumner ssumner
    20. October 2013 at 10:20

    Tom, Yes, NGDP doesn’t capture all that’s important in the real economy, but it does capture AD. Interest rates are almost never a good indicator of policy.

  7. Gravatar of benjamin cole benjamin cole
    20. October 2013 at 10:42

    Everyone forgets, there was a mirror bust in commercial real estate too—including the big stuff, the trophy office buildings etc. Very sophisticated buyers, financiers etc. Low rates made such buyers and financiers go crazy?
    Buyers expected rising rents…and might have been right if the Fed had not tightened at precisely the wrong moment…

  8. Gravatar of Geoff Geoff
    20. October 2013 at 10:43

    “During 2002-04 Greenspan was 100% right and Taylor was 100% wrong. The macro variables did not indicate an excessively expansionary monetary policy (from a dual mandate perspective)”

    No, Greenspan was 100% wrong. The Fed did indeed inflate in order to reverse the fallout from the 2001 slump.

    For example,

    http://research.stlouisfed.org/fredgraph.png?g=nyl

    The Fed decreased the Fed Funds rate from a high of 6.5% in 2001 to just 1% by 2004. The Fed did this via massive expansion of bank reserves, i.e. inflation. Inflation of reserves is how the Fed lowers the Fed Funds rate from where it otherwise would have been.

    Just look at the rate of change in the monetary base (proxy for bank reserves) starting in 2001. The rate of change went from below zero starting in 2001, to a double digit increase by 2002, and then only gradually did the rate of increase start to decrease.

    It is a denial of reality to claim that the Fed was not expansionary during 2002-2004. The data is right there for everyone to see.

  9. Gravatar of Daniel Daniel
    20. October 2013 at 11:35

    Apparently, in Mises-land you can have “inflation” without an actual rise in prices.

    Fascinating 🙂

  10. Gravatar of kebko kebko
    20. October 2013 at 12:12

    Alright, Scott, I’ve got one for you to yell from the mountaintops!

    The housing boom was caused by TIGHT money!

    What I realized was that there were many parallels between the 1970’s and the 2000’s, and that both periods saw similar demographic trends and low or negative real interest rates. Relative home prices rose in both time periods, but the reason they didn’t rise so much in the 1970’s is because the Fed was actually loose in the 1970s. Since home finance is treated as consumption financing, high inflation in the 1970’s kept home prices down. (Banks approve mortgages based on current income compared to the monthly payment, not based on something like a long term cash flow analysis of real estate versus a risk free bond investment.) Since the Fed was tight in the 2000’s, real and nominal interest rates were low. This meant that, unlike in the 1970’s, the size of the monthly payment was not a limiting factor. Homes were a reasonable investment in both periods of time. In the 1970’s, they were a killer investment if you could afford the monthly payment. The financial engineering used in the 2000’s to lower equity and down payments in exchange for higher monthly payments was simply a reasonable way to make this investment available to more people. Those methods were not relevant in the 1970’s because the monthly payment was already the limiting factor, because of high nominal rates.

    I go into the details a little more here:
    http://idiosyncraticwhisk.blogspot.com/2013/08/real-interest-rates-and-housing-boom.html

  11. Gravatar of Jon Jon
    20. October 2013 at 13:56

    Urban containment is an important regulatory issue to include in any explanation of the housing price boom–emphasize price boom, not housing boom; the quantity of new house construction was not out of step with rates of household formation, but I broadly agree with Shiller’s claim that house prices were out-of-step. (http://www.econ.yale.edu/~shiller/data/ie_data.xls)

    Urban containment correlates with the locations of the price boom: las vegas, florida, etc. In brief it was a policy of preventing the outward expansion of population centers conjoined to the earlier polices of prevent upward expansion by building height restrictions. This is included state programs to purchase the development rights to farmland and outright purchase open-space.

    See for instance “Distortions Resulting from Residential Land Use
    Controls in Metropolitan Areas” by Jansen and Mills (http://goo.gl/Q0ISKd)

    More importantly, Urban Containment was an international phenomena, part of broad regulatory changes intended to implement Kyoto and reduce transportation carbon emissions.

  12. Gravatar of Geoff Geoff
    20. October 2013 at 14:04

    kebko:

    Your thesis blows up as soon as it is realized that in a context of “tighter money”, a rising in spending and prices of houses, should have been accompanied by a fall in spending and prices of all other goods. But spending and prices rose on average.

    The only possibility for this is if money was loosened, not tightened.

  13. Gravatar of ChargerCarl ChargerCarl
    20. October 2013 at 14:49

    One of my favorite blog posts:

    http://blog.andyharless.com/2010/08/what-housing-bubble.html

  14. Gravatar of John John
    20. October 2013 at 21:53

    Scott,

    This argument is predicated on the idea that interest rates are something that only affects NGDP instead of communicating information about scarcity like every single other price in the world. It may be a mainstream position but it seems pretty ridiculous.

    You could just as well say that the price of gasoline is a governor on NGDP and the government should step into the futures market and start buying gas until they are expecting to hit their targets for inflation/NGDP.

    For that matter, why does all the time spent in micro learning about what happens when the government sets price floors and ceilings suddenly get thrown out the window when it comes to interest rates? Here’s my theory, the government took control of that function and economists serve the government rather than the truth.

  15. Gravatar of John John
    20. October 2013 at 21:58

    I think my last comment really boils down to one question. Is there a difference between interest rates going down because the Fed created liquidity and interest rates going down because people decided to save more money?

    I would argue that in one situation, people decided to consume less now and to consume more later and entrepreneurs should focus on projects with a later payoff. In the other situation, entrepreneurs will focus on longer term projects because of the lower interest rate but consumer preferences haven’t changed introducing a mismatch between desired consumption and resource allocation.

  16. Gravatar of ssumner ssumner
    21. October 2013 at 05:07

    kebko, Great post. It is worth yelling from the mountaintops.

    Chargercarl, Excellent Harless post.

    John, You said;

    “This argument is predicated on the idea that interest rates are something that only affects NGDP instead of communicating information about scarcity like every single other price in the world. It may be a mainstream position but it seems pretty ridiculous.”

    No, you misunderstood me. I don’t think interest rates affect NGDP, at least not very much. You are right that interest rates reflect all sorts of factors, such as supply/demand factors in the credit markets.

    As far as I know the government does not put any price floors or ceilings on interest rates. People are free to lend and borrow at whatever rate they choose.

  17. Gravatar of dlr dlr
    21. October 2013 at 05:55

    If the Fed successfully targets inflation or NGDP then interest rates are 100% endogenous. In that case any interest rate -> housing causality would not involve monetary policy at all, at least if one assumes the macro targets were appropriate. If inflation and NGDP were on target during the housing bubble, then Taylor would be 100% wrong and Greenspan would be 100% right.

    If the Fed’s OMOs are thought to be temporary they can still keep inflation and NGDP on target. Yet they can nonetheless depress real interest rates over the shorter of the OMO’s expected lifespan and the duration of price stickiness. This is a version of the Jeremy Stein question: Can the central bank plus sticky prices distort longer-lived asset prices with artificially low expected real interest rates over, say, 1-4 years even without changing NGDP and inflation for a while?

    You need frictions to get this result, i.e. turn a few years of depressed rates into bubbles of long-duration assets that takes a long time or reaches some quasi-equilibrium before filtering into goods and services price and price expectations, but why aren’t those frictions at least plausible to you? If we have sticky prices and time-varying OMOs, it is possible for some short term real rates to be a lousy indicator of whether money is easy or tight but still be exogenously low or high due to a CB with a conventionally defined neutral reaction function.

  18. Gravatar of John John
    21. October 2013 at 08:02

    Scott,

    The point of what I said was that there is a difference between interest rates going down due to saving or Fed action. I’m well aware that the Fed doesn’t explicitly set price floors or ceilings however, by injecting money they can lower the interest rate without consumers having to save more and that encourages production to go out of line with consumer preferences.

  19. Gravatar of ssumner ssumner
    21. October 2013 at 17:34

    dlr, My first reaction is that this would be almost impossible. Asset prices are forward looking. So how could a easy money policy that is so temporary that NGDP does not rise, somehow make asset prices rise? Perhaps there’s a model, but I’m dubious.

    John, Yes, the Fed can make short term rates fall, but if it is truly an easy money policy than inflation should rise. If inflation doesn’t rise, then the Fed is actually just going along with less demand for credit, and those rates really do reflect market conditions.

  20. Gravatar of Greenspan and the Housing Bubble | askblog Greenspan and the Housing Bubble | askblog
    22. October 2013 at 03:06

    […] Scott Sumner writes, […]

  21. Gravatar of The Non-Falsifiable “Empirical” Frameworks of Paul Krugman and Scott Sumner The Non-Falsifiable “Empirical” Frameworks of Paul Krugman and Scott Sumner
    22. October 2013 at 07:34

    […] #1: In this post, Scott Sumner reports that John Taylor accused Alan Greenspan of fueling the housing bubble by […]

  22. Gravatar of Tom Brown Tom Brown
    23. October 2013 at 13:46

    Greenspan interview (final segment):

    http://www.thedailyshow.com/full-episodes/mon-october-21-2013-alan-greenspan

Leave a Reply