IS-LM: A Teaching Device and Its Discontents

Why a diagram became famous

In the 1930s, John Maynard Keynes did something unusual: he tried to make deep depression and chronic underemployment analytically respectable without pretending they were a temporary deviation from a frictionless, self-correcting system. The General Theory of Employment, Interest and Money (1936) is a strange book: parts read like a manifesto for reconstruction, other parts are careful accounting of expectations, the psychology of liquidity preference, and the way investment collapses for reasons that a simple “supply and demand” story cannot capture.

A few years later, a younger theorist, John R. Hicks, offered a two-curve summary. In a short paper often remembered by its “Mr. Keynes and the Classics” framing, he drew IS (investment = saving) and LM (liquidity = money) and showed, at least on a blackboard, how fiscal policy might shift the whole picture. The diagram spread because it is portable: it can be sketched in chalk, it fits into textbooks, and it helps students see how something—government spending, the money supply, animal spirits in investment—might change income and the interest rate at once.

This essay is not a victory lap for the IS-LM model. It is a reader’s map: what the device gets right as a teaching step, what it silences from Keynes’s more radical claims, and why generations of post-Keynesian and financial-instability readers still complain that a diagram can look like a map while hiding the territory.

Jargon, explained first: IS stands for a condition where intended investment equals intended saving; LM is where the demand to hold real money balances matches the available quantity (however defined). General equilibrium in this mini-model is the crossing point of the two.

The kindness of a toy model

A toy model is not an insult. Good pedagogy needs compression. A first-year class often starts from two intuitions: (1) if spending rises, income and output tend to follow (the multiplier intuition that Reckonomics covers in the essay on the Keynesian multiplier in perspective); (2) if the interest rate is low enough, people may hold more money, but if the central bank is constraining a monetary aggregate, something has to give.

Stitch those together, and the IS and LM curves offer a two-equation picture of a closed economy, at least in the most stripped-down form. A fiscal expansion shifts IS right; under conventional slopes, you get higher income, and the interest rate may rise, “crowding out” some private investment. A monetary expansion shifts LM, lowering the interest rate, stimulating real activity—again, in the most textbook telling.

That is not a complete policy manual, but it is a useful warning label for sophomores: policy instruments interact, and a story that is only fiscal or only monetary may miss the joint determination of income and the rate of interest. For readers of Keynes’s General Theory in plain terms (see our plain-English map), the Hicks diagram can feel like a friendly translation.

What IS-LM leaves out, even when it is “right”

The classic complaint—made memorably in spirit by the Cambridge Joan Robinson in her frustration with a “Keynesian” school that, in her view, tamed the master—is that IS-LM is not “what was in the book General Theory,” but a stabilized version of it that imports equilibrium habits Keynes wanted to question.

A short list of things that a chalkboard two-curve world often underdescribes, even when the algebra is “correct” given its assumptions:

Radical uncertainty. In Keynes, expectations are not a parameter you refine with “better forecasts.” The future is in many ways unknowable; investors’ confidence can swing for reasons the model does not internalize as structural shocks. A mechanical IS shift driven by a random “expectations shock” in modern macro is not the same as Keynes’s long passages on the beauty contest, conventions, and the precariousness of long-run calcuations.

Money and finance are not a neutral veil. A neat LM curve with exogenous M and a well-behaved demand for money is a first pass. A world with liquidity preference as a defense against uncertainty is a world where the entire financial architecture—banks, inside money, credit creation, and leverage—shapes the transmission mechanism. The essay on Minsky and financial fragility is a reminder: balance-sheet postures, not a single r in a diagram, often drive the cycle.

Price and wage adjustment is not a footnote. IS-LM in its textbook version often delays questions about the labor market and the structure of markups, sometimes pushing them to an AS-AD add-on. Post-Keynesian pricing stories—markup rules, effective demand first—fit awkwardly in a system where “the” interest rate and “the” output level are doing most of the work.

The open economy and international finance. A closed-economy IS-LM is a classroom convenience. A country with a floating rate, a heavy foreign debt, or a banking system with dollar liabilities, needs different machinery—sometimes three-curve stories (Mundell-Fleming) or, more simply, a frank admission that a trade shock is not a neat parallel shift.

None of that means the Hicks diagram is useless. It means it is a ladder that should be climbed and then, when needed, left without pretending you are still seeing the same landscape.

How to read IS-LM without re-importing a myth of self-correction

If you treat IS-LM as “Keynes in full,” you risk two opposite mistakes.

The pessimist mistake is to act as if policy never matters because “complexity is infinite,” which is a counsel of paralysis. The overoptimist mistake is to treat a crossing point of curves as proof that a benevolent technocrat can always steer to full employment, because the model has an interior solution. Keynes’s political point was that wage cuts are not a reliable cure in a monetary production economy, that aggregate demand can fail, and that expectations can validate bad equilibria.

A careful reader of our article on the stagflation and the Keynesian consensus will already sense why the 1970s matter here: a supply shock plus wage-price processes broke the “simple” demand-management picture in ways that a classroom IS-LM, without supply side and expectations, was ill-equipped to explain. That is not a claim that Hicks caused the confusion, but a reminder that one teaching device can become a stand-in for reality in public debate.

IS-LM and the post-Keynesian alternatives

Post-Keynesian macroeconomics (see endogenous money and banking views) often replaces or heavily qualifies LM-style money stories with a focus on credit and the behavior of banks. The short version: a bank makes loans and creates deposits; the money supply, in a broad sense, is not a single lever a central bank sets with mechanical precision, especially in normal times, though institutions and regulation matter enormously. If LM treats money as a stock moving along a static demand function, a banking view treats money as a liability of the financial system that expands and contracts with creditworthiness and collateral values.

A second post-Keynesian flavor focuses on distribution and class conflict without pretending that income shares are a marginal productivity residual—something Sraffian and neo-Ricardian lines emphasize in a different, price-of-production vocabulary. A third, associated with Minsky, foregrounds the stability–instability dialectic: the financial system, during calm, encourages positions that make sense only if others keep funding them.

None of this requires you to burn the Hicks diagram, but it does require you to relegate it to a limited domain: a partial illustration of a narrow question, not the canonical photograph of a monetary economy.

Pedagogical use: a staged curriculum for honest readers

If you are building your own map of ideas, a staged approach helps:

Stage 1: Master the “closed economy, fixed prices, exogenous money” IS-LM story as algebra practice. It teaches comparative statics, the logic of simultaneous determination, and the vocabulary of crowding out in a first pass.

Stage 2: Immediately pair it with a one-page “what we assumed,” especially about prices, the labor market, the exchange rate, and the financial system.

**Stage 3: Read a chapter of Keynes on long-term expectations and compare the tone to a textbook. Ask whether the mechanism of recovery you imagine is a movement along curves, or a change in the entire political–economic state of confidence.

Stage 4: Read Hayek-Keynes splits on money and order not to pick a sports team, but to see the debates about knowledge and coordination that IS-LM cannot display in two lines.

Stage 5: Return to the multiplier essay and connect the fiscal debate to a balance-sheet world, where a surge in public debt during a slump may be a lifeboat and not, by itself, a moral parable about thrift.

Three classroom exercises that build intuition without lying

Exercise 1: Trace the logic of crowding out. In a closed economy with standard slopes, move IS right with a pure government-purchase increase. Watch the interest rate rise. Ask: which part of the story is a telescoping of many markets into one interest rate, and which part would change if the central bank held the short rate fixed by supplying reserves? The point is to learn the tension between fiscal impulse and the monetary policy reaction function—a tension that a static diagram can hint at, but that real data resolves only with institutions.

Exercise 2: Add an open-economy footnote in words. A fixed exchange rate with free capital movement can invert a domestic story. Fiscal expansion might drain reserves; monetary expansion might blow up a peg. A floating rate changes the transmission again. A serious reader of Bretton Woods and its end and of modern exchange-rate regimes will see why the same IS and LM labels are not enough.

Exercise 3: Read a paragraph of Joan Robinson alongside your textbook and list three claims she makes that no two-line intersection can contain. If you cannot list three, you have not read carefully enough. That is not elitism; it is a check on the humility of the modeler.

IS-LM and the Lucas critique, briefly

A later generation asked a devastating question, documented in the essay on what a model is and why models break under policy change: if people’s expectations and institutions shift when policy rules shift, a reduced-form curve you estimated yesterday may misbehave tomorrow. The Lucas critique is not, by itself, a one-line refutation of IS-LM; it is a warning that deeper structure must connect policy to behavior. A purely descriptive IS–LM system can still teach the algebra of simultaneity while you treat expectations as a next layer.

In practice, the New Keynesian triad of intertemporal choice, sticky prices, and a policy rule replaced much undergraduate IS-LM pedagogy in some programs—yet the ghost of an IS and an interest–output tradeoff often remains in narrative form. Students deserve to know: we changed the toy but did not always improve the honesty about finance.

A fair verdict

IS-LM is a tremendous teaching tool and an incomplete research program for serious macro. It is not “wrong” the way a botched data series is wrong; it is thin—a sketch that can, if you are not careful, be mistaken for the whole of Keynes when it was meant as a first translation of part of a larger, messier book.

In that way, the history of the IS-LM model is really a parable about pedagogy and power: a society will reach for a figure it can teach quickly, and then treat that figure as a skeleton key, because skeleton keys are easy to hand to central bankers, editorial boards, and students who need a starting point.

If you read one more thing after this, read Keynes on uncertainty, then Minsky on finance, and only then return to the curves to see if they still feel like a faithful portrait—or like a first draft that served its day.