Tuesday, May 30, 2006

The IS-LM Model

A reader emails me the following question:

Dear professor Mankiw:

I like your blog a lot. I daily go to it in order to read good economics. Keep up the excellent work!

May I ask you why economists authors of textbooks on intermediate macroeconomics like you keep using the IS-LM model even though we already know that the Central Bank does not set the monetary supply. Instead, it does set the interest rate. Shouldn´t you do like Wendy Carlin and David Soskice in their recent and fantastic book "Macroeconomics: Imperfections, Institutions and Policies" where they replace the LM curve by a monetary rule (for example, a Taylor rule). Wouldn´t that be more representative of what occurs in reality rather than supposing that the institution gets the control of the quantity of money?

Thanks for your attention in advance.

[name withheld]

To answer this question, let me start with an excerpt from Chapter 11 of my intermediate macro text. This passage shows how I handle these issues when teaching this course:

What Is the Fed's Policy Instrument--The Money Supply or the Interest Rate?

Our analysis of monetary policy has been based on the assumption that the Fed influences the economy by controlling the money supply. By contrast, when the media report on changes in Fed policy, they often just say that the Fed has raised or lowered interest rates. Which is right? Even though these two views may seem different, both are correct, and it is important to understand why.

In recent years, the Fed has used the federal funds rate--the interest rate that banks charge one another for overnight loans--as its short-term policy instrument. When the Federal Open Market Committee meets every six weeks to set monetary policy, it votes on a target for this interest rate that will apply until the next meeting. After the meeting is over, the Fed's bond traders in New York are told to conduct the open-market operations necessary to hit that target. These open-market operations change the money supply and shift the LM curve so that the equilibrium interest rate (determined by the intersection of the IS and LM curves) equals the target interest rate that the Federal Open Market Committee has chosen.

As a result of this operating procedure, Fed policy is often discussed in terms of changing interest rates. Keep in mind, however, that behind these changes in interest rates are the necessary changes in the money supply. A newspaper might report, for instance, that "the Fed has lowered interest rates." To be more precise, we can translate this statement as meaning "the Federal Open Market Committee has instructed the Fed bond traders to buy bonds in open-market operations so as to increase the money supply, shift the LM curve, and reduce the equilibrium interest rate to hit a new lower target."

Why has the Fed chosen to use an interest rate, rather than the money supply, as its short-term policy instrument? One possible answer is that shocks to the LM curve are more prevalent than shocks to the IS curve. When the Fed targets interest rates, it automatically offsets LM shocks by adjusting the money supply, although this policy exacerbates IS shocks. If LM shocks are the more prevalent type, then a policy of targeting the interest rate leads to greater economic stability than a policy of targeting the money supply. (Problem 7 at the end of this chapter asks you to analyze this issue more fully.)

Another possible reason for using the interest rate as the short-term policy instrument is that interest rates are easier to measure than the money supply. As we saw in Chapter 4, the Fed has several different measures of money--M1, M2, and so on--which sometimes move in different directions. Rather than deciding which measure is best, the Fed avoids the question by using the federal funds rate as its policy instrument.

----[end of excerpt]

My email correspondent wonders whether it would be better just to jettison the traditional IS-LM model in favor of an alternative framework that ignores the money supply altogether and simply takes an interest-rate rule as given. This approach has been advocated by my old friend David Romer. (Economics trivia fact: I was the best man at David Romer's wedding, and he at mine.) You can find David's approach here (figures here). David calls his alternative presentation the IS-MP model, because it combines an IS curve with a monetary policy reaction function.

The first thing to understand about the choice between IS-LM and IS-MP is that it is not about determining which is the better model of short-run fluctuations. There is no truly substantive debate here. These two models are alternative presentations of the same set of ideas. The key issue in deciding which approach to prefer is not theoretical or empirical but pedagogical.

The IS-LM approach has a long history behind it. That is one reason to stick with it, but it is not dispositive. If I were convinced that the IS-MP model was a clear and substantial step forward, I would switch. So far, however, I am not convinced that the new approach is easier to teach or more intuitive for students.

The key difference between the two approaches is what you hold constant when considering various hypothetical policy experiments. The IS-LM model takes the money supply as the exogenous variable, while the IS-MP model takes the monetary policy reaction function as exogenous. In practice, both the money supply and the monetary policy reaction function can and do change in response to events. Exogeneity here is meant to be more of a thought experiment than it is a claim about the world. The two approaches focus the student's attention on different sets of thought experiments.

I like the IS-LM model because it keeps the student focused on the important connections between the money supply, interest rates, and economic activity, whereas the IS-MP model leaves some of that in the background. The IS-MP model also has some quirky features: In this model, for instance, an increase in government purchases causes a permanent increase in the inflation rate. No one really believes that result as an empirical prediction, for the simple reason that the monetary policy reaction function would change if the natural interest rate (that is, the real interest rate consistent with full employment) changed. This observation highlights that neither model's exogeneity assumption should be taken too seriously.

In the end, I remain open-minded, but at this point I prefer the IS-LM model when teaching (at the intermediate level) about the short-run effects of monetary and fiscal policy. If one were to teach IS-MP to undergrads, I would prefer to do it as an supplement, rather than a substitute, for IS-LM.

Related link: Here (and here in published form) is Paul Krugman's cogent defense of teaching the IS-LM model. The article was written quite a while ago, before IS-MP hit the scene, so I don't know what he would say about this alternative framework. But the Krugman piece is interesting, if only vaguely on point, so I wanted to give it some free advertising.