Comments on Arnold Kling on Larry Summers

As usual, I’m in a sort of triangle, where I don’t agree with either side.  But FWIW, here’s how I’d respond to Arnold Kling’s new post:

Cowen’s stagnation story is that the pace of innovation has slowed, resulting in declining growth in aggregate supply. In contrast, Summers’ story is one of a permanent shortfall of aggregate demand, due to an excess of desired saving over desired investment, which can only be eliminated at a negative real interest rate.

The stories complement each other.  Summers claims that some sort of exogenous shock has reduced the long run real interest rate on safe assets.  Slower growth in RGDP (due to supply-side factors) is one likely explanation.

Here are some criticisms that come to mind.

1. If “the” full-employment real interest rate is negative, then why do we need quantitative easing? Why does not the excess of saving over investment not by itself drive long-term rates to zero?

Because nominal yields are expected to rise above zero in the long run.

2. Summers wants to claim that full employment has been achieved in recent years because of asset bubbles. However, in a world of negative real interest rates, there is no such thing as an asset bubble. Real assets have infinite value in such a world.

Most assets are riskier than T-bills, and long term rates are not negative.

3. As Tyler points out, it is hard to reconcile positive economic growth with negative real interest rates. We have had positive economic growth, even since 2008.

And real T-bill yields have been negative since 2009.  It’s not a question of whether you can “reconcile” the negative rates with growth; they are a fact.  The question is what’s causing them?

4. As Tyler also points out, we observe higher interest rates for risky assets. In fact, if you want to understand the low interest rates that Summers and Krugman are talking about, then my suggestion is to “follow the guarantees.” In one way or another, the U.S. government has provided a guarantee on many investments. Government bonds are one example. Mortgages are another.

I think everyone agrees that one of the reasons why T-bill yields are so low is that they are risk-free. But they were risk free before 2008, and had positive yields.  The fact that the rates on various safe assets are low only because of a government guarantee has no bearing on Summers’ argument.

5. The prime rate at banks averaged 5 percent from 2001-2004, almost 7 percent from 2005-2008, and 3.25 percent from 2009-2012. Inflation over these periods averaged 2.3 percent, 3.4 percent, and 1.5 percent respectively, so that the real rate of interest has been positive throughout.

Yes, but these are not risk free assets, so their positive yield has no bearing on Summers’ argument.  He’s making a claim about risk free assets.

6. Summers’ revival of the secular stagnation hypothesis has not been broadly peer reviewed. Before people jump on the bandwagon, I would wait until it has been evaluated by a broader range of economists.

The best peer review is the markets.  The long term bond markets suggest that Summers is partly right, but overstates his case.  The 30 year T-bond yields 3.81%, which is low, but well above the 1.64% in Japan.  The market probably see rates hitting zero in future recessions and initial recoveries, but not otherwise.

PS.  The Keynesian liquidity trap model suggests that monetary stimulus is mostly ineffective if T-bill yields are zero, regardless of the level of other interest rates.  I don’t buy this model (especially with the ad hoc “bubble” addendum), but the arguments used by Kling are not going to convince any Keynesian, as their model allows for other assets to have positive yields.



6 Responses to “Comments on Arnold Kling on Larry Summers”

  1. Gravatar of benjamin cole benjamin cole
    20. November 2013 at 07:40

    Re Summers: surely he would agree that printing money will boost AD.
    The chronic global capital glut: call me crazy, but does perennial debt monetization make sense?

  2. Gravatar of MG MG
    20. November 2013 at 08:35

    You are technically correct that Summers was specifically referring to negative real rates on T-bills. But I think you should not dismiss Arnold’s (implicit) extrapolation: if you start out with a seriously negative factor-one risk premium (zero duration, zero credit, zero liquidity risk) (a) markets will compress all add-on risk premia and (b) the sum of all of premia many yet leave other assets trading at nominal rates indistinguishable from zero, and ill capable of cushioning risk.

  3. Gravatar of Morgan Warstler Morgan Warstler
    20. November 2013 at 12:47

    Let’s say this out loud:

    There is no unemployment problem in America, we simply have not yet adopted GI / CYB.

    And Larry Summers, and all econos should be required to say that everytime they speak:

    “Of course we should first use GI / CYB to employ everyone and raise consumption, before we do anything as stupid as borrow and spend more money…. ” and then make whatever stupid observation that have to make.


    What we ought to be noodling is the lack of places to invest money “safely”

    We have a TON of productivity gains being made, but they are all FREE, they subtract, not add to GDP – see GD Freebie.

    It looks like we need a 1% NGDPLT where we eat 2%+ deflation and 3%+ RGDP, so that we must SEEK OUT AND DESTROY sticky prices and wages in any policy / law we can find.

    Free Harvard Online, free movies, music, govt run with 5M public employees 24/7 via apps, driverless shared cars, there’s nothing left to park money in that isn’t space exploration.

    Look, the guys on my side of the world are modeling out everything atomic costing $1 per pound and everything digital being free (Jurvetson) and Greenspan, god love him, agrees…

    Someone ought to be thinking about what happens if literally, every year there is 2% LESS STUFF to lever debt on.

  4. Gravatar of ssumner ssumner
    20. November 2013 at 13:20

    MG, And what are the implications of that?

  5. Gravatar of Morgan Warstler Morgan Warstler
    20. November 2013 at 13:21

    Here’s a good article to help kinda crystalize how fast free is driving thru our world kicking things over:

  6. Gravatar of Scott re-interprets Larry | askblog Scott re-interprets Larry | askblog
    21. November 2013 at 07:05

    [...] Scott Sumner writes, [...]

Leave a Reply