Krugman Doesn’t Understand IS-LM, Part 1

by Steve Keen, Debt Watch

Read Part 2.

Krugman describes himself as a “sorta-kinda New Keynesian” (Krugman 2012, p. 104), and explains in End This Depression NOW! (Krugman 2012) that New Keynesian macroeconomics evolved in reaction to the failure of the New Classical approach to “explain the basic facts of recessions” (Krugman 2012, p. 103). His “sorta-kinda” qualification is because both New Keynesian and New Classical models are “micro-founded”-derived from applying assumptions about the behaviour of individuals and markets at the level of the macroeconomy-and he has a healthy scepticism about these assumptions:

I don’t really buy the assumptions about rationality and markets that are embedded in many modern theoretical models, my own included, and I often turn to Old Keynesian ideas, but I see the usefulness of such models as a way to think through some issues carefully-an attitude that is actually widely shared on the saltwater side of the great divide. (Krugman 2012, p. 104)

This is one aspect of Krugman that I genuinely applaud: the awareness that models aren’t reality. At best they are representations of reality, but some Neoclassicals show an amazing capacity to believe that their models are reality-as in this piece by Gilles Saint-Paul which Unlearning Economics deservedly flogged recently-in a way that leads to truly delusional thinking about the real world.

Of course, we need models to think about the economy in the first instance, because it is such a complex entity. Even those who deride modelling in economics are using a model when they talk about the economy-it’s just a verbal (or even unarticulated) one, as opposed to the mathematical models that academic economists develop.

One might hope that experience and experimentation over time would weed out unrealistic models in economics-but that hasn’t happened. In many ways, the models that dominate economics today are less realistic than those which prevailed as much as seventy years ago.

Krugman alludes to this by his reference to “Old Keynesian ideas” above. In particular, he champions John Hicks’s IS-LM model, first published in 1937 (Hicks 1937), as an explanation of our economic crisis today.

In that paper, Hicks proposed that the Great Depression was caused by what was later termed a “liquidity trap“. Hicks argued that it was possible for the economy to be in an equilibrium (a word I’ll be labouring in this post) in which there was involuntary unemployment.

(Why the word “involuntary”? Because the “Freshwater”, New Classical economists that Krugman is accustomed to fighting argue that all unemployment is voluntary: people look at the current wage, consider the loss of leisure they’d incur to have to work for it, and decide that not working gives them higher utility.)

Hicks’ model had two intersecting curves: the IS curve and the LL curve (which was later relabeled LM) drawn on a diagram with the rate of interest (i) on the vertical axis and the level of income (I) on the horizontal (Income was later labelled “Y” to get away from the sheer bloody confusion of multiple “I”s). I’ll delay explaining how the curves are derived until I get to Krugman’s use of the model; for now the important thing is the shape of the curves.

Figure 1: Hicks’s original IS-LL model, from page 153 of his 1937 paper “Mr Keynes & the Classics”

Hicks argued that the LM curve would be flat at low levels of the rate of interest, and steep at high levels of income (as shown in his Figure 2), because:

there is (1) some minimum below which the rate of interest is unlikely to go, and (2) (though Mr. Keynes does not stress this) there is a maximum to the level of income which can possibly be financed with a given amount of money. (Hicks 1937, p. 154)

Hicks then used this to divide the diagram into two regions, depending on where the IS curve intersected the LM.

If the intersection was in the section where LM was rising steeply, then an increase in demand-caused by, for example, a budget deficit-would mainly drive up the rate of interest, with very little impact on the level of income. This was the “Classical” region where what is today called “crowding out” would occur and where deficits only cause bad things like higher interest rates (and, in more elaborate models, inflation).

However if the IS curve intersected with LM in its flat region, then an increase in demand via a government deficit would drive the equilibrium income level higher, while having very little impact on the rate of interest. This was the region where Keynes’s arguments applied, said Hicks: that when the economy was depressed and income was very low, the government should stimulate demand by expansionary policy.

It was also where monetary policy was ineffective-as Hicks illustrated in his Figure 2. Because there was already a minimum level for the rate of interest-which later economists christened “the zero lower bound”-then this bit of the curve couldn’t be shifted by increasing the amount of money. Only if the economy were in the “Classical” region, would increasing the supply of money move the LM curve further out-as shown by the dotted line in Hicks’s Figure 2. Therefore monetary policy was effective when income was high, but ineffective when income was low (and unemployment was high).

On the other hand, in the region where the LM curve was flat, monetary policy was ineffective but fiscal policy-which could shift the IS curve-worked. Since this was what Keynes was advocating in The General Theory (Keynes 1936), Hicks concluded:

So the General Theory of Employment is the Economics of Depression. (Hicks 1937, p. 155).

The IS-LM model took over the academic profession in part because it eliminated the apparent existential threat to Neoclassical economics that appeared to exist elsewhere in Keynes’s work-as for example in these wonderfully confrontational lines from Keynes’s own summary of his message in the paper “The General Theory of Employment” in 1937:

I accuse the classical economic theory of being itself one of these pretty, polite techniques which tries to deal with the present by abstracting from the fact that we know very little about the future. (Keynes 1937, p. 215)

Instead, there could be peaceful co-existence: Neoclassicals owned the boom times, while Keynesians got the bad times. But just as with peaceful coexistence in the geopolitical sphere, at least one side didn’t really believe in it. Neoclassicals hoped for total domination, and in the battle to achieve it they were aided both by economic circumstances and by the nature of that other sphere of economics, microeconomics.

After WWII, the times appeared to be good all the time-certainly in comparison to The Great Depression, when unemployment hit 26% (see Figure 2)-so that Neoclassicals got most of the airplay and Keynesians very little. Neoclassical theory also ruled the microeconomic roost, and from this fortress emerged a campaign to destroy Keynesian thought entirely.

Figure 2: US Unemployment rate. The U-6 measure today is comparable to the official measure in the 1930s

In what became known as the Microfoundations debate, Neoclassicals attacked the Keynesian part of the profession with the charge that Keynes “did not have good microfoundations”-that Keynesian results like an equilibrium with unemployment contradicted microeconomic theory. As Lucas emphasized in his 2003 speech to the History of Political Economy conference when he was President of the American Economic Association, a key target in this assault was Hicks’s IS-LM model:

Nobody was satisfied with IS-LM as the end of macroeconomic theorizing. The idea was we were going to tie it together with microeconomics and that was the job of our generation. (Lucas 2004, p. 20)

Now the bad times are back, and Krugman is trying resuscitate IS-LM. I argue that he should leave it dead-not for the reasons that the New Classicals killed it off (being inconsistent with Neoclassical microeconomics is a plus in my books) but because it’s a lousy model for what we’re experiencing right now. There’s no better way to show this than to outline how Krugman is trying to use it, and show that he gets it wrong.

Krugman’s Derivation

Krugman describes his derivation of IS-LM in two posts that feature as “essential reads” on his blog: “IS-LMentary” and “Liquidity preference, loanable funds, and Niall Ferguson (wonkish)” (Krugman 2009). He portrays IS-LM as “a way to reconcile two seemingly incompatible views about what determines interest rates”:

One view says that the interest rate is determined by the supply of and demand for savings – the “loanable funds” approach. The other says that the interest rate is determined by the tradeoff between bonds, which pay interest, and money, which doesn’t, but which you can use for transactions and therefore has special value due to its liquidity – the “liquidity preference” approach. (Krugman 2011)

In his attack on Niall Ferguson, Krugman guesses that Ferguson has a simplistic view of the Loanable Funds model in isolation as the basis of his opposition to fiscal stimulus. That model shows the supply of saved money as increasing as the rate of interest rises, while the demand for borrowed money falls as the rate of interest rises. Where the two lines intersect, the demand for savings from firms equals the supply of savings from households (see Figure 3, from Krugman’s takedown of Niall Ferguson).

Figure 3: Step 1 in Krugman’s derivation of an IS curve

Krugman speculated that this vision, and this alone, explained why Ferguson thought “that fiscal expansion will actually be contractionary, because it will drive up interest rates”. But Krugman points out that this picture alone ignores the fact that both the investment demand for money (I) and the household supply of money (S) depend on the level of GDP: a higher level of GDP will enable a higher level of savings, and it will also be associated with a higher level of investment. So you need to know GDP as well to work out the interest rate in the market for Loanable Funds.

Both the S and the I lines will shift out as GDP rises-which one will shift more. Here is the first bit of “shifty” logic: to ensure that the IS curve slopes downwards, Krugman assumes that savings will rise more than investment does for a given increase in GDP:

Suppose GDP rises; some of this increase in income will be saved, pushing the savings schedule to the right. There may also be a rise in investment demand, but ordinarily we’d expect the savings rise to be larger, so that the interest rate falls.

Krugman concludes his derivation of the IS curve with a dynamic observation: that a fall in the interest rate (for some other reason-say an attempt by the Fed to stimulate the economy) can cause both savings supply and investment demand to expand:

Suppose that desired savings and desired investment spending are currently equal, and that something causes the interest rate to fall. Must it rise back to its original level? Not necessarily. An excess of desired investment over desired savings can lead to economic expansion, which drives up income. And since some of the rise in income will be saved – and assuming that investment demand doesn’t rise by as much – a sufficiently large rise in GDP can restore equality between desired savings and desired investment at the new interest rate.

Given that assumption, the intersection of I2 and S2-which represent investment demand and savings supply at a higher level of GDP than I1 and S1-is lower than the intersection for I1 and S1. If you join these equilibrium points up, you get Krugman’s downward-sloping IS curve (which I’ve added in to Figure 4 as a red dotted line)

Figure 4: Step 2 in Krugman’s derivation of an IS curve

Then there’s the LM curve. Krugman explains this as follows:

Meanwhile, people deciding how to allocate their wealth are making tradeoffs between money and bonds. There’s a downward-sloping demand for money – the higher the interest rate, the more people will skimp on liquidity in favor of higher returns. Suppose temporarily that the Fed holds the money supply fixed; in that case the interest rate must be such as to match that demand to the quantity of money. And the Fed can move the interest rate by changing the money supply: increase the supply of money and the interest rate must fall to induce people to hold a larger quantity.

Krugman doesn’t provide a diagrammatic derivation of the LM curve, so I’ll provide mine from Debunking Economics (you can download the supplement with all the figures in the book from here; this Figure 5 below is Figure 61 on page 25 of the supplement). A key part step is the proposition that the Federal Reserve controls the supply of money, and that it can move it at will-so the money supply is an “exogenous” factor in the model since it is not controlled by the market. Therefore the money supply is shown as a fixed vertical line in the model: it’s impervious to both the rate of interest, and the level of income.

The demand for money however depends on both those things: a lower interest rate will increase the demand for money, since there’s less benefit in foregoing ready access to your money and buying bonds instead; a higher income will increase the demand for money, since there are more transactions taking place and you need more money on hand for them. The first factor is shown by having a downward-sloping demand for money curve for any given level of income; the second is shown by moving the demand curve out to the right as income rises.

Figure 5: Deriving the LM curve–Figure 61 in Debunking Economics

Equilibrium in the LM market thus depends on both the rate of interest and the level of GDP. When you plot these equilibrium points on a diagram with GDP on the horizontal axis and the rate of interest on the vertical, you get an upward-sloping LM curve-the second half of the overall IS-LM model.

Put the two curves together and the equilibrium of both-the point where the two curves cross-gives you the equilibrium interest rate and GDP for the economy. As Krugman puts it:

The point where the curves cross determines both GDP and the interest rate, and at that point both loanable funds and liquidity preference are valid.

Figure 6: The IS-LM model (from Krugman’s IS-LMentary)

Well Yada Yada. After all that, we’re staring at the economist’s favourite abstraction, a pair of intersecting lines. How does Krugman use them to explain the current crisis-and why is he wrong?

I’ll cover those topics in Part 2.


Hicks, J. R. (1937). “Mr. Keynes and the “Classics”; A Suggested Interpretation.” Econometrica 5(2): 147-159.

Keynes, J. M. (1936). The general theory of employment, interest and money. London, Macmillan.

Keynes, J. M. (1937). “The General Theory of Employment.” The Quarterly Journal of Economics 51(2): 209-223.

Krugman, P. (2009). “Liquidity preference, loanable funds, and Niall Ferguson (wonkish).” The Conscience of a Liberal

Krugman, P. (2011). “IS-LMentary.” The Conscience of a Liberal

Krugman, P. (2012). End this Depression Now! New York, W.W. Norton.

Lucas, R. E., Jr. (2004). “Keynote Address to the 2003 HOPE Conference: My Keynesian Education.” History of Political Economy 36: 12-24.

Read Part 2.

Special notice: Global Economic Intersection recommends you consider supporting the Kickstarter campaign of Steve Keen to keep his efforts to develop the “best macro modeling tool ever” moving forward in the absenced of university affiliation.  GEI Managing Editor John Lounsbury has made a contribution and explained whyGEI News has been following the progress of the Kickstarter campaign.  To get an in depth review by Steve himself (and to make a pledge) go to the Minsky Kickstarter page.

Share this Econintersect Article:
  • Print
  • Digg
  • Facebook
  • Yahoo! Buzz
  • Twitter
  • Google Bookmarks
  • LinkedIn
  • Wikio
  • email
  • RSS
This entry was posted in macroeconomics and tagged , , , , , , . Bookmark the permalink.

Make a Comment

Econintersect wants your comments, data and opinion on the articles posted. You can also comment using Facebook directly using he comment block below.