Comments on Matt Rognlie

Matt Rognlie has a number of new posts that are quite good.  Here he emphasizes the importance of monetary stimulus:

In failing to understand this core logic, most commentary about “deleveraging” is rather bizarre. At some level, it’s the same cluelessness that we once saw from central planners: they’d trip over themselves in the complexity of fixing a shortage in one market or a glut in another, never quite realizing that the price mechanism would do their work for them.  Right now, historically low inflation expectations and below-potential output are prima facie evidence that real interest rates are too high. That’s what every macro model tells us is associated with contractionary policy by the Fed. Yet we see pundits lost in all kinds of complicated, small-bore proposals to stimulate the economy””when the fundamental, overriding dilemma is getting the price (in this case, the interest rate) right.

. . .

 The Fed’s failure to use all the tools at its disposal””in particular, its failure to make a conditional commitment like the one proposed by Chicago Fed President Charles Evans“”is by far the most serious failure of economic policy today.

And here he explains why the argument for monetary stimulus doesn’t go away just because we face a “balance sheet recession.”

Even households drowning in debt tend to have some assets: a house, and maybe a 401(k) or IRA. All else equal, low interest rates place upward pressure on home prices (since they bring down the cost of financing for those who can obtain it) and make both equities and long-term bonds much more valuable (since lower rates increase the discounted value of an asset’s payout). This can actually help fix household balance sheets: it brings them out from underwater on their mortgages (or, at least, makes them less underwater than they otherwise would be) and increases the value of their other financial assets. In this light, it’s entirely conceivable that crippled balance sheets make monetary policy more effective, not less. Although the “wealth effect” from higher equity and bond prices matters most for the richest Americans, it’s useful for a much broader group.

In this post his criticizes the IS-LM and AS/AD models:

This is nothing, of course, compared to the abomination that is the AD/AS model, also included in undergraduate textbooks. AD slopes down for the same outdated reason that LM slopes up: given a constant money supply rule, lower prices imply higher real money balances and therefore lower real interest rates, which lead to higher demand. (It can also be justified using real balance effects, which are quantitatively irrelevant, or fixed exchange rates, which only exist in a few cases.) This has absolutely nothing to do with monetary policy as it’s currently implemented. Yet the simple AS and AD curves, made appealing by the apparent (but false) analogy to ordinary supply and demand, lurk somewhere in the minds of countless former economics students. This leads to all kinds of bad intuition””like the notion that sticky prices are problematic because they prevent the adjustment to equilibrium on the AD/AS diagram. (Wrong. Under current Fed policy, the price decrease -> lower interest rate -> improvement in demand mechanism is no longer operative, unless deflation combines with the Taylor rule to force a policy that the Fed should have chosen anyway. Certainly this is no use at the zero lower bound, where price flexibility is actually harmful, because it leads to more deflation and higher real interest rates.)

I entirely agree with Matt’s argument that the AS/AD model he describes (the standard Keynesian one) is an abomination, and that AS/AD has nothing to do with supply and demand (and hence is misnamed.)

But I actually like a different AS/AD model; indeed it is the model I use to take a first pass at macro issues.  The version I like uses an AD curve that represents a given level of aggregate expenditure, or NGDP.  Thus it’s drawn as a rectangular hyperbola.  Last time I posted on AS/AD some commenters assumed the rectangular shape had something to do with the Quantity Theory.  It doesn’t, so please don’t make that assumption.  It’s an arbitrary definition, and like all definitions it’s not a theory.

Is it a useful definition?  I think so.  I view monetary policy as the force that (actively or passively) determines the path of NGDP.  Then we use the SRAS to describe how shifts in nominal spending are partitioned between prices and output in the short run, and the LRAS to explain how spending shifts are partitioned in the long run.

I don’t agree with Matt’s assumption that increased price flexibility is destabilizing, under any conditions.  Under a gold standard the future expected real value of gold puts a floor under deflation.  Under fiat money regimes the floor comes from the fact that most debts are nominal, and it’s politically impossible to deflate very fast.  In that case the faster wages and prices fall to the appropriate equilibrium, the quicker the economy can recover.  We recovered very fast from the severe 1921 deflation, precisely because wages and prices adjusted very quickly in the relatively laissez-faire policy environment of the Harding years.  That’s not to say 1921 wasn’t a severe recession.  It was severe, but if wages had been artificially prevented from falling fast (as in 1929-30) the damage would probably have been even greater.

I think it’s also a mistake to assume that increased price flexibility makes the real interest rate rise more sharply during an adverse demand shock.  Prices do fall much faster, but expected inflation doesn’t necessarily fall, at least over the sort of time frame that is relevant for business cycles.  Again, because there is a floor on how much prices can fall, a quick downward adjustment in commodity and asset prices (which are the prices most relevant for the Fisher equation), means real rates can more quickly regain their natural rate.  Even during 1929-32 there is evidence that the deflation was unexpected, and hence real rates weren’t particularly high.

The preceding should not be misconstrued as implying I “favor deflation.”  I don’t, I favor monetary policies that allow for steady NGDP growth.  But if the monetary authority screws up and lets NGDP fall by X%, it’s better that wages and prices quickly adjust to reflect the X% lower NGDP.  That’s the implication of AS/AD, and it’s a correct implication.



10 Responses to “Comments on Matt Rognlie”

  1. Gravatar of Nick Rowe Nick Rowe
    8. October 2011 at 20:06

    Suppose the central bank was targeting the price level, rather than NGDP, and doing so successfully. Then I would argue that the AD curve, under that monetary regime, should usefully be drawn as horizontal.

    I can’t help thinking that you find the rectangular hyerbola AD curve useful because you think that targeting NGDP is what central banks *ought* to be doing. (Which is not necessarily a bad argument, I suppose. But yours is a normative AD curve?)

  2. Gravatar of Morgan Warstler Morgan Warstler
    8. October 2011 at 20:38

    You folks can circle jerk this all you want…

    but the fact is…

    “Right now, historically low inflation expectations and below-potential output are prima facie evidence that real interest rates are too high.”


    it is prima facie evidence tha prices are too HIGH.

    WHO CARES if people lose houses they have no equity in?

    WHO CARES if people lose houses they have no equity in?

    WHO CARES if people lose houses they have no equity in?

    WHO CARES if people lose houses they have no equity in?

    How many times do I have to say it?

    We are down to worrying about CREDIT SCORES of the losers.

    They have no equity.

    NO ONE with equity is losing a house.

    Admit this!

    Once you admit this, what all of you are really just trying to save the banks.

    Keeping prices up, and we ARE just talking about housing prices, if the ONLY thing you are all rationalizing.

    Labor prices are falling. See UAW.


    Psst, watch this…

    When the CPI threatens to go negative, the Fed WILL gve you QE3.

    IF the banks have to eat the losses and put 6M shadow inventory houses on the market, the CPI WILL GO NEGATIVE.

    If home prices fall, YOU GET QE3!

    Why don’t you want home prices to fall?

  3. Gravatar of John John
    8. October 2011 at 22:57


    Thank you for admitting that laissez-faire policies allow prices and wages to fall fast enough to prevent a sustained depression. That is the truth. If the bush-Obama administrations had allowed prices to adjust, there would’ve been no sustained depression right now.

  4. Gravatar of Mark A. Sadowski Mark A. Sadowski
    8. October 2011 at 22:58

    When I teach AS/AD I go out of my way to explain why that model should never be confused with the micro models. I trust that others do as well. If not then that should be condemned.

    Oddly, giving my specialty (macro) I have far more experience teaching micro. I suspect that it has something to do with my outlook. (God help me, but I actually believe monetary policy is still effective at the zero lower bound.) Evidently I’m such a crank that despite the fact that macro is my specialty I’m far too nuts to be allowed to teach it at the principles level.

    By the way, I’m having a great time teaching this fall. I’m completely hypomanic and I don’t care what my psychologist says to the contrary. I think my students benefit from seeing me truly happy and energized.

  5. Gravatar of anon anon
    9. October 2011 at 01:02

    Mark, I could be wrong, but I think Scott’s variant of AS/AD is a bona-fide supply and demand diagram. A given level of nominal expenditure (“AD”) can be identified with the demand for “everything but money balances”, interacting with the supply of “everything but money” (goods and services) to determine the price level and real output. This would clarify the connection between AS/AD and monetary phenomena, which is arguably obscured (although it’s still there) in the standard Keynesian model.

  6. Gravatar of Bill Woolsey Bill Woolsey
    9. October 2011 at 04:27

    I have my doubts about nominal debts making deflation

    The Harding scenario occured with a gold standard. That ties down the expected future price level somewhat.

    I also think it helps alot if the “deflation” is returning to the long run expected price level. Second best is if the deflation is temporary and the price level is expected to rise back up to normal. Third best is that the price level is moving to a new, permanently lower level that is unknown.

    But the scary scenario is where no one really knows where the bottom might be. A central bank that is so focused on interest rate targets that it doesn’t even know that money exists (or maybe it just insists on scientific models, the existing ones don’t include money) could get into this situation.

  7. Gravatar of Ram Ram
    9. October 2011 at 05:05

    What kinds of evidence would convince you that the monetary authority does not always have the capacity to accelerate nominal GDP growth? I’m not taking a position on the issue, but based on the way you think about how monetary policy works, it is hard to see what could change your mind on this particular subject. If I had to guess, I’d say that you would change your mind if there was evidence of faster growth in the current and expected future supply of base money that did not lead to faster growth in the current and expected future path of nominal GDP. But without the appropriate futures markets to reveal the appropriate expectations, I’m not sure how such evidence could be identified. Moreover, the supposition that we can infer these expectations from certain asset prices itself is premised upon, say, the EMH or something like it, and there too it’s hard to see what would make you change your mind on that. Don’t get me wrong, I’m not asserting that you’re incorrect about these things, I’m just wondering how empirical evidence could enter into the picture in a way that would cause you to revisit these seemingly axiomatic commitments.

  8. Gravatar of Marcelo Marcelo
    9. October 2011 at 05:21


    Karl Smith at has an interesting take on IS-LM, where he replaces the curves with a Bank Lending-Monetary Policy set up. It seems to be pretty interesting, wanted to know your thoughts on it…

    It is primarily explained here :

    and a little more explanation here:

    I think it is very intuitive way of looking piecing together the story with a good account of the ZLB, while also acknowledging that banks are the ones that must lend for inflation/NGDP to rise.

  9. Gravatar of Scott Sumner Scott Sumner
    9. October 2011 at 05:56

    Nick, I see why you say that, but that’s actually not the reason. It’s because I see monetary shocks and velocity shocks as having very similar effects on the economy. On the other hand I see inflation shocks as being quite different.

    Suppose they weren’t targeting NGDP, but rather M moved around randomly, say a pure gold standard. Also assume V moves around randomly. I’d argue that what matters is the change in M*V, and it does much matter which of the two is actually changing.

    In contrast, I see inflation shocks quite differently. An increase in either M or V moves you up and to the right on the SRAS. But an increase in inflation doesn’t do that, unless caused by more M or more V. Suppose the increase in inflation is caused by an adverse supply shock. I can’t even imagine how you’d show that on a AS/AD diagram where the price level represented AD.

    i think about macro this way:

    1. There are things that change nominal expenditures. I call these nominal shocks:

    2. There are various factors that explain how nominal shocks are partitioned into real GDP and price level changes. That’s SRAS/Phillips curve.

    I don’t know whether inflation is a nominal or real shocks. It depends on whether it is from the supply or demand side.

    Morgan, I should have commented on that quotation by Rognlie, although I wouldn’t have used the phrase “circle jerk” in my comment. A very powerful expansionary monetary shock can actually lead to higher real interest rates, because it might lead to higher growth expectations. Those who view the monetary transmission mechanism exclusively through the interest rate channel, can’t understand this process. A powerful expansionary monetary shock can sharply raise the expected prices of goods and assets two years out (via the expected hot potato effect.) If nominal wages are sticky, that also increases the expected future level of output, and real interest rates.

    John, I’ve made this point about 1921 many times, but people don’t tend to pay attention. However, it isn’t just bad policy. The radically changed nature of our labor force has made wages stickier even in the absence of government intervention. We no longer have huge armies of agricultural, mining, and industrial workers whose wage is determined in a sort of spot market.

    Mark, I fear that some instructors do say AS/AD is like S&D for the entire economy.

    anon, I don’t think so. The long run supply in micro is perfectly horizontal. In macro it’s perfectly vertical. For AD there is no substitution effect as one moves along the curve.

    Bill, You said;

    “I have my doubts about nominal debts making deflation

    Actually, I agree, that wasn’t my claim. My claim was that nominal debts make price flexibility stabilizing. I believe nominal debts make deflation (actually falling NGDP) even more destabilizing than otherwise.

    What I tried to say is that nominal debts make rapid deflation a political impossibility under a fiat regime. The public would demand the government stop the deflation. Even Japan has not had rapid deflation.

    Hong Kong is a good example of an economy that has occasional deflation, but the price flexibility is stabilizing. But they’d be better off without the deflation in the first place.

    Ram, You said;

    “What kinds of evidence would convince you that the monetary authority does not always have the capacity to accelerate nominal GDP growth?”

    Good question. Here are some things:

    1. If I saw a fiat money central bank trying to raise the rate of inflation, and failing.

    2. If the Fed did the things Bernanke recommended the BOJ do, and it failed. That would mean level targeting to make up for the recent inflation shortfall. Even if the Fed does level targeting, it’s very unlikely they’d try to make up for the recent inflation shortfall, as Bernanke recommended the Japanese do. So I’m not expecting to see my theory disproved, as it’s very unlikely any central bank will try to significantly raise inflation. BTW, QE2 did slightly raise inflation, so that example was a success. The problem with QE2 was that they had the wrong policy goal–they should have aimed for higher NGDP.

    If countries like Switzerland and Sweden were unable to depreciate their exchange rate, that would also give me doubts. But the market thinks they can, and hence they can.

    The best evidence would be if we created NGDP futures contracts, and getting the future price up to 5% was impossible, even with the Fed buying all the safe debt in the world economy. It wouldn’t mean that going on to unsafe debt wouldn’t work, but it would show that it’s so hard that it’s beyond the practical ability of policymakers (of course this assumes no IOR.)

  10. Gravatar of Scott Sumner Scott Sumner
    9. October 2011 at 06:01

    Marcelo, I don’t see monetary policy working through interest rates and bank lending, but rather changes in expected NGDP and current asset prices. Monetary stimulus raises future expected NGDP, and current stock, commodity, and commercial real estate prices. This causes firms to produce more (assuming sticky wages.)

Leave a Reply