Winner of the New Statesman SPERI Prize in Political Economy 2016

Friday 29 March 2013

Reactions to Textbook Mundell Fleming

Some of the reactions to my earlier post, like Paul Krugman’s, have been that we no longer teach Textbook Mundell Fleming (TMF), but instead teach something along the lines I was suggesting. Which is great. [1] Of course I never meant to suggest everyone is taught TMF, but enough are to make the post warranted.

Another line, which I attempted (unsuccessfully) to head off in the original post, was UIP does not fit the data, so why build a core model that uses it? There are three replies to this:

1) Although UIP does terribly in the short run, its deficiencies in the longer term may not be so bad. Menzie Chinn sent me a summary of his own work that suggests this, which is here. In my two period model, where the first period is a Keynesian short run, then its the longer term predictive power that is relevant. Having said this, I also in my lectures stress that UIP is not well supported by the data, and that some people believe you can make money by betting against it (e.g. the ‘carry trade’).

2) However I do not think there is a more empirically based alternative that is satisfactory. Some people suggested a random walk for the exchange rate, and indeed - as Nick Rowe points out - we can recover TMF from UIP this way if expectations are formed on that basis. But a random walk model is inconsistent with almost any macroeconomic theory involving exchange rates that we might care to teach.

I know many people react negatively to rational expectations, but really all I’m arguing for here is common sense. Take again the example of a temporary (first period) fiscal expansion in a two period model without any backward dynamics. Unless the model contains hysteresis, we know the second period exchange rate will not change relative to the no fiscal expansion case. So to assume that agents expect any change in the first period exchange rate to carry through into the second period is the height of irrationality. I really do believe we should not be teaching models to undergraduates where people are that dumb.[2]

3) Its what macroeconomists use. I know many people do not like this, but I have always felt strongly that the core of first or second year undergraduate macro should be consistent with the macro that, say, central banks use. It is bound to be simpler, but it should be consistent. Of course we should emphasise its deficiencies, and mention alternatives, and also generally say something about the history of how we got here. The problem with too much undergraduate macro is that it is history in the wrong sort of way - its just out of date. [3]

In my earlier post I tried to emphasise the inconsistency between TMF and UIP, and not go on about the fixed money supply assumption. However, just as teaching up to date macro involves using the IS part of IS-LM along with, say, a Taylor rule, so it should be possible to adapt TMF to a world where central banks set interest rates if the model was any good. Yet if we dropped the LM curve from TMF, and assumed instead that interest rates were the policy instrument, we would conclude that an independent monetary policy in an open economy with flexible exchange rates was impossible. According to TMF, central banks cannot set interest rates to anything other than world interest rates. Why do we want to start students off with a model that suggests this?

And so finally to the LM curve itself. I have yet to hear a remotely convincing justification as to why this remains in the first macro model that we teach students. What great insight do students get when we pretend that the central bank fixes the money supply? I’ve talked about the muddles teaching the LM creates before - David Romer presents seven advantages of doing something more realistic here. [4] Is it really the conservatism of textbook writers that means that we are condemned to carry on confusing students, and cannot we do anything about this?  

[1] I always recommend students read Krugman, Obstfeld and Melitz when they get confused about open economy macro, but I had my sights trained on other macro texts.

[2] Yet I do teach a backward looking (static expectations) Phillips curve alongside a New Keynesian Phillips curve (NKPC). Why the double standard? I can suggest three reasons: (1) using static inflation expectations is not so obviously silly if the monetary authorities give us little clue about what they are doing, which describes the world a few decades ago, (2) there are empirical features (e.g. Phillips curve loops) which are difficult to rationalise with the NKPC (3) firms and workers do not spend large amounts of money trying to predict future inflation.

[3] So, for example, I’m happy mentioning TMF as a lead up to UIP. Saying something like “originally economists modelled capital flows as a function of only interest rate differentials, but this either ignored expected capital gains, or made a very naive assumption about exchange rate expectations. We now know better.” For much the same reason you might mention that the original Phillips curve did not have an expected inflation term in it, but we now know better.

[4] Oddly, although Romer carries through his arguments to an open economy, he does not embrace UIP: in fact I do not think he even mentions it.


  1. Simon,

    "And so finally to the LM curve itself. I have yet to hear a remotely convincing justification as to why this remains in the first macro model that we teach students."

    That's a good place to start. Then expunge the money "supply" from all subsequently taught macro models. Even highly trained macro economists can't seem to escape the idea that somehow there is some economic
    relevancy of the supply, even contingent on knowing the short rate. This is not benign, as it seems to have caused a whole generation of
    economists, including many who work at central banks, to be totally confused about the potential benefits of QE, which has become a disastrous distraction.

    Bottom line is, the supply of non-interest bearing money is *only* relevant to the economy via its effect on the short rate. Once we know the short rate we know longer have to care about the existence of money (except to know that the willingness of the CB to supply it makes negative rates impossible). So yes, IS/MP is a good place to start. But we need to follow through with the rest of the curriculum from there.

    For those who claim to know of statistically significant deviations from UIP (and the EMH in general), it should be recommended that they they start an FX hedge fund. It requires very little capital and I'm sure they will become very wealthy.

  2. While I often agree with much of what you say, in the comments I'll focus on points of disagreement.

    First, a point of agreement: The LM curve is indeed a 'barbarous relic'. Personally I find the 3eq model (IS-PC-TR) in the early part of Carlin and Soskice a good starting point.

    "I know many people react negatively to rational expectations, ..."

    Including such Chicago heavyweights as James Heckman, and Gregory Chow.

    Speaking for myself, it is the overall context within which ratex is used that is problematic. This includes the representative agent and various no-arbitrage, no-ponzi conditions. Does a model exist in the air, independent of reality, or is it meant to analyze some aspect of the real world? Obviously the latter (though Ed Prescott would strenuously disagree). So what does the representative agent in the model represent of the real world? Does the RA represent n identical agents behaving like a Borgian collective or is it a simplification for tractability meant to represent people who are heterogeneous along multiple dimensions and also interacting among themselves in ways the model ignores? In the latter case the appearance of any "free lunch" in the model could be just an illusion, since no such exploitable free lunch is likely to exist in reality. cf the Peter Diamond quote in a comment to your previous post.

    (This is not about the formal properties of RE or aggregation theorems, but a response to your oft-repeated claim that allowing a "free lunch" in a RE + RA model is "obviously irrational".)

  3. Simon

    I agree that the empirical relevance of UIP is not the key question here. Still, I think invoking the UIP to defend the relevance of fiscal policy in open economy macro is wholly unjustified.

    All that the UIP tells us is the slope of the term structure of future exchange rates. What happens to the spot rate in response to a particular policy decision is completely beyond its remit. As Rowe says, your model is contingent on exchange rates adjusting to their LR equilibrium in next period, which is assumed to be what they were in this period before policy decision.

    Using UIP-like rational expectations arbitrage logic, hard to see why the exchange rate will then move at all, and all that we will get is risk-less arbitrage profits for Fx traders for some time before the soevereign reverses its decision. UIP is rational expectations arbitrage-driven partial equilibrium,. Extend it to general equilibrium and all manners of Keynesian policy recommendations fly away out of the window.

    The other way you could defend policy action is through arguing for sticky goods prices. Then, you'd have to combine Mundell Fleming with UIP, allow the derivative of price to be closer to zero but the log-derivative to be perfectly flexible, and basically get Dornbusch's currency overshoot model. That's a good model. But it's fundamentally an MF model, with the nominal rigidity assumption driving the conclusion, just as in mainstream IS/LM. UIP is an added constraint on it.

    Your attempt to defend fiscal policy using UIP alone is still shaky.

    1. I do not understand this at all. First, my post was not about justifying fiscal policy - that was just a useful experiment to show what was wrong with TMF. I thought it was obvious that the first period in my model involved sticky prices: that is why aggregate demand determines the real exchange rate in the second period, but output in the first period. Dornbusch overshooting is not fundamentally an MF model - it contradicts it, because it uses UIP.

  4. Romer briefly mentions UIP on page 235 (and largely in a footnote) of his 3rd edition Advanced Macroeconomics. I only know this because in one of the way-too-spooky coincidences I happened to have been reading re-reading chapter 5 the night before you posted your Mundell-Fleming piece. And in yet another way-too-spooky coincidence I had exactly the same fleeting thought about the inconsistency with Mundell-Fleming.

  5. I checked on the Carlin/Soskice treatment, and yes, it does incorporate UIP in a MF model. But it uses adaptive expectations of the exchange rate (with some side-trips to Dornbusch overshooting). So your objection is to adaptive expectations?

    Well, if we know that there is really no theory of e that works empirically, what's so bad about adaptive expectations? If e is really more like a state variable, its probably rational to use adaptive expectations.

    So I'm not convinced. I don't spend much time on open economy macro, and I need a teachable, succinct model. Some kind of MF still fits the bill.

    1. Sorry to reply to my own post, but forgot to mention this. The reason to teach the LM curve is that it lets you explain the debate about the self-adjusting nature of a market economy. I.e., will the Keynes effect guarantee full employment, even with a laissez-faire policy? Answer: not in a liquidity trap, especially if Fisher debt-deflation is operating. That's an important conversation to have.

    2. If agents assumed tomorrows exchange rate always equals todays, and UIP holds, then countries cannot run independent interest rate policies. Countries do run independent interest rate policies. You would think an open economy macro model would try and explain what such policies might do, not deny they can exist. QED?

      On the LM curve, why does fixing the money supply tell you anything about the self adjusting nature of the economy. Why not fix the real interest rate instead, which tells you why an appropriate monetary policy is essential for the economy to successfully adjust.

    3. "If agents assumed tomorrows exchange rate always equals todays, and UIP holds, then countries cannot run independent interest rate policies"

      Not quite. Countries can still run independent interest rate policies, but they create risk-less arbitrage profits for traders, that do not vanish through the usual process that arbitrage profits are supposed to vanish.

      Instead, they vanish all at one go, when independent central banks abandon the pursuit one rate regime for a different one. Then, the accumulated risk-less profits vanish in one go, causing severe one-time capital losses.

      Or, UIP arbitraging is the classic pennies in front of a steamroller trade. USD/yen carry trade, for example.

    4. Further, what UIP+differential interest rate policies does imply is that exchange rates will be best approximaed by a random walk, because the spot rate has no LR equilibrium as anchor.

      Or, Meese-Rogoff.

  6. The usual argument against the textbook treatment of the quantity of money and the LM curve is that Central Banks set an interest rate and supply whatever quantity of high-powered money is demanded. As a consequence the quantity of money is essentially indeterminate and the LM curve is hoizontal. Is that your objection to the textbook case?

    But the interest rate is surely not surely not an end in itself; rather it is set with some deeper objective in mind. For the UK - and many other countries - this deeper objective is presumably a target inflation rate. Is it really not "remotely plausible" to suggest that from this perspective the Central Bank sets the interest rate with the aim of achieving a target level (or rate of change) for the quantity of money (and hence ultimately the price level or its rate of change). And if it does, and does so accurately, then in effect the Central Bank sets the quantity of money; the interest rate is merely the mechanism it uses to do so.

    From this perspective, whether the Central Bank sets the level of high powered money directly or whether it sets the interest rate with the objective of achieving a target value for the quantity of money doesn't matter: the textbook case is "remotely plausible".

    I think Nick Rowe has made a similar argument to the one I'm making here.

    1. There is a leap of logic here. Yes, countries have inflation targets. So lets assume they have money supply targets (!?) The two things are not the same. Inflation targets are simpler, because they do not involve this variable called money. So why not teach the simpler, and more realistic, thing?

    2. Is it really a huge logical leap to suggest that one of the things a CB with an inflation target would look at is what's happening to the quantity of money? If, having set an interest rate, it noticed that the quantity of money was expanding at a rate of 50% would it really not think about raising the interest rate because of concerns about inflation? OK, it doesn't have an explicit target for the quantity of money but is it really not "remotely plausible" that implicitly it does?

      As for the teaching, the general point I'd want to get across to students is that the fact that the CB operates monetary policy by setting the interest rate doesn't by itself mean, as it might superficially appear, that it is not concerned about the quantity of money, or indeed that it is not in fact setting the quantity of money as the textbook model suggests. It depends upon what causes it to set the interest rate it does.

    3. In some countries central banks are told to target inflation, and in the US they have a dual mandate. So I'm supposed to tell my students to ignore all that - we will instead assume that really they have a target for the money supply. Remotely plausible? Pretty remote.

      What is odd here is motivation. Sometimes economists assume implausible things because those things are in some sense optimal. So, it is argued, although agents do not appear to do x, they are better off if they do x, so perhaps they in effect do x even though they appear not to. But that argument does not apply to central banks and the money supply - it is not optimal to target the money supply rather than inflation, if what you care about is inflation. So where does this idea that they really target the money supply come from?

    4. No, I'm not telling you to tell your students to ignore all that. But I am suggesting that you point out to them that the interest rate isn't an end in itself. So you have to ask why a CB sets the interest rate it does rather than stop the analysis there. And I'd go on to ask them what they think the BoE would do if it found that at the current interest rate the quantity of money was expanding at 50%, but the inflation was still only 2%. I'd suggest to them that finding itself in that situation the BoE might get alarmed and raise the interest rate since such a rise in the quantity of money is highly likely to be a precursor of future inflation.

      And then I'd point out to them that such a response suggests that interest rates are set with at least some eye on what is happening to the quantity of money. I'd use that to illustrate that appearances can be deceptive and that the setting of the interest rate doesn't necessarily indicate that the quantity of money is endogenously determined; it may merely be the means by which control of the quantity of money is achieved. And I would use that as one "remotely plausible" justification for the textbook treatment of monetary policy and the LM curve.

      Of course I would then go on to say that if interest rates are set with no account taken of the effect on the quantity of money then a better approach would be to treat the quantity of money as entirely endogenously determined and draw the LM curve as horizontal.

    5. Of course you need to say what the goals of the central bank are. So you tell students its to hit an inflation target, or its to hit an imaginary money supply target. I am still mystified why you want to tell them the latter.

      If you asked the Bank of England about your hypothetical involving 50% growth in money, they would probably reply that they would only be concerned to the extent that it influenced their ability to achieve their mandate - the inflation target. And you could have the same hypothetical for lots of other variables, like house prices, but we do not teach students that really the central bank is targeting house prices.

      Imagine, in particular, that we substitute the exchange rate for the money supply in your example. The central bank would probably be concerned by a 50% depreciation. Does this mean that it is remotely plausible that the central bank is really targeting the exchange rate? Should we as a result be telling students that really flexible exchange rates are fixed exchange rates?

  7. But the big question is: what's your purpose in teaching macroeconomics to undergraduates?

    It could be that it is to ensure that your students adopt the neoliberal neoclassical mindset, and in that case you just get a popular USA textbook or equivalent and follow that, with the implication that the economy is like that except for governmental distortion.

    It could be that it is to equip your students with the terminology and concepts to discuss with their peers, and then you can teach a stylized introduction to currently popular approaches, making clear that is required for your students' professional development, and in a more subdued way that you make no claims as to whether those terminology and concept relate to the economy, as opposed to Economics.

    Or you could attempt to teach about the economy rather than Economics, and then you have a really difficult task, because currently popular macro approaches don't seem to me suitable to that.

    Because attempts then to teach how to understand the economy becomes mostly introducing an contrasting different approaches and delivering your subjective impression on how helpful they are in different periods of economic evolution. But that puts you and perhaps even your students outside the "mainstream".

  8. «but we do not teach students that really the central bank is targeting house prices.»

    Despite 25 years of Greenspan Put and Bernanke Swap and endless talk by central bankers of the "wealth effect" of house (and share) price increases on demand, and that wage rises are inflationary but capital gains are not, and despite a balance sheet expansion of trillions by central banks like the Fed Board and the ECB a large part of that into deeply impaired mortgage based securities to boost their prices?

    Uhm :-)

  9. I think that the MF model implicitly assumes that the value of the domestic bonds is equal to those of other countries, and obviously this assumption is unrealistic (we can see the risk premium). Also, it seems that the MF model presupposes the domestic country is in inflation so that the yields of the bonds can rise.

    Also, it is dubious that any stimulus via fiscal expansion can be cancelled out by the capital inflow. To what extent is the stimulus cancelled out by the increase of the current account deficit? The answer should be done quantitatively, not qualitatively.

    James Tobin's paper says that in some cases the MF model is not valid:

    Anyway the MF model is not applicable to the large open economy like the US. And if the Marshall-Lerner condition is not satisfied, the MF model becomes invalid.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.