Mundell-Fleming Model

A theory built in the basement of the IMF

In the early 1960s the post-war Bretton Woods system was beginning to creak, the first wave of European financial integration was under way, and two economists working independently — both, as it happened, sitting in offices at the International Monetary Fund — wrote down essentially the same model. The Canadian Robert Mundell published a sequence of papers in 1960–63, most importantly "Capital Mobility and Stabilization Policy under Fixed and Flexible Exchange Rates" (1963). The Scotsman J. Marcus Fleming published "Domestic Financial Policies under Fixed and under Floating Exchange Rates" in 1962. Neither initially knew of the other's work; the model now bears both names by convention. Mundell received the 1999 Nobel Memorial Prize in Economic Sciences "for his analysis of monetary and fiscal policy under different exchange rate regimes and for his analysis of optimum currency areas." Fleming had died in 1976, before the Nobel committee got around to recognising the contribution.

The model is the open-economy generalisation of John Hicks's 1937 IS-LM diagram — itself a teaching apparatus distilled from Keynes's General Theory. IS-LM lives in the closed-economy world: there is no exchange rate, no capital account, and the central bank's money supply is exogenous. Mundell-Fleming relaxes those assumptions one at a time and arrives at a strikingly clean result: the exchange-rate regime determines which policy lever works. Under fixed rates, fiscal policy is dominant and monetary policy is impotent. Under floating rates, the assignment reverses. Behind both flips lies the same deeper proposition — the impossible trinity, that no country can simultaneously have free capital, fixed FX, and an autonomous central bank.

Three curves in (Y, r) space

Mundell-Fleming plots three relations on the same axes as IS-LM — output Y on the horizontal axis, the domestic interest rate r on the vertical axis.

• IS (goods market). All combinations of Y and r at which planned spending equals output. Slopes downward: higher r reduces investment and (in an open economy) net exports, lowering equilibrium output. Shifted right by fiscal stimulus (higher G, lower T) and by a real depreciation that boosts NX.
• LM (money market). All combinations at which money demand equals supply. Slopes upward: higher Y raises money demand, so r must rise. Shifted right (down) by an increase in real money supply M/P.
• BP (balance of payments). All combinations at which the current account plus capital account sum to zero. Its slope depends entirely on the degree of capital mobility — and that is the key new feature of the model.

At full equilibrium, all three curves intersect at the same (Y, r) point. In the closed economy you only need IS and LM, and the third equation is the labour market. In the open economy the third equation is the BP condition, and how it is configured changes everything.

The BP curve and the capital-mobility spectrum

The BP curve traces the (Y, r) pairs that balance the external accounts. Higher Y lifts imports, worsening the current account, so to keep BP = 0 the capital account must compensate — which requires a higher r to pull in foreign capital. BP therefore slopes upward, and the steepness depends on how much capital flow a given interest-rate gap can summon.

Capital-mobility regime	BP slope	Implied condition	Real-world example
No capital mobility	Vertical	K-account = 0; current account alone must balance	Capital controls (China today, Bretton Woods era)
Low / imperfect mobility	Steeper than LM	K responds weakly to r-differentials	Many EMs in 1980s
High but imperfect mobility	Flatter than LM	K responds strongly to r-differentials	Most OECD, late 20th century
Perfect mobility	Horizontal at r*	r = r* always; infinite K-flow at any gap	Modern integrated capital markets (textbook benchmark)

The textbook Mundell-Fleming takes the limit of perfect capital mobility: BP is flat at the world interest rate r*. The slightest deviation r ≠ r* triggers unbounded capital flows that arbitrage the rate back to r*. This is an idealisation, but a useful one: it sharpens the comparative-statics intuition and isolates the role of the exchange-rate regime. Everything that follows assumes perfect capital mobility unless noted.

Fixed exchange rates: fiscal works, monetary fails

Suppose the central bank has fixed the exchange rate at e̅ and stands ready to buy or sell foreign reserves to defend that peg. Now consider the two stabilisation experiments.

Monetary expansion under fixed rates

The central bank prints money. LM shifts right, putting downward pressure on r. With perfect capital mobility, the moment r dips below r* capital floods out, the demand for the domestic currency collapses, and the exchange rate wants to depreciate. To defend the peg the central bank must sell foreign reserves and buy back its own currency, which automatically contracts the money supply. LM shifts back to where it started.

ΔM (open-market op)   →   r ↓   →   K outflow   →   FX wants to depreciate
   ↓ (peg defence)
Central bank sells reserves, buys home currency
   ↓
M shrinks back   →   LM returns to original position   →   Y unchanged

The conclusion is sharp: under fixed rates with capital mobility, the central bank does not control the money supply. M is endogenous, pinned by whatever amount is consistent with r = r* at the prevailing output. Monetary policy as a stabilisation tool is impotent.

Fiscal expansion under fixed rates

Now government spending rises. IS shifts right, putting upward pressure on r. Capital flows in, the exchange rate wants to appreciate, and the central bank — defending the peg — must buy foreign reserves and sell domestic currency, expanding M. LM shifts right until r is back at r*. Output ends up much higher than the closed-economy IS-LM analysis would predict, because the monetary accommodation prevents any interest-rate-driven crowding-out.

Fiscal policy is therefore fully potent under fixed exchange rates with capital mobility. The implied policy multiplier in this regime is the simple Keynesian multiplier without monetary leakage — exactly the case in which fiscal expansion is most effective.

Floating exchange rates: monetary works, fiscal fails

Now release the peg. The central bank no longer defends an FX target; it sets M and lets the exchange rate float to clear the BoP.

Monetary expansion under floating rates

The central bank prints money. LM shifts right, r dips. Capital flows out, the currency depreciates. Depreciation makes exports cheaper and imports dearer, lifting net exports. This shifts IS to the right. The new equilibrium has Y much higher and r back at r*. Monetary policy works through two reinforcing channels — the interest-rate channel (boosting investment) and the exchange-rate channel (boosting net exports).

Fiscal expansion under floating rates

Government spending rises, IS shifts right, r presses upward, capital flows in, the currency appreciates. Appreciation makes exports more expensive and imports cheaper, subtracting from NX. IS shifts back to the left until r returns to r*. The fiscal stimulus is completely crowded out — not by higher rates squeezing private investment (as in the closed-economy story), but by the appreciation eating away at net exports. The exchange rate has done all the work of restoring external balance.

ΔG   →   IS right   →   r ↑   →   K inflow   →   FX appreciates
                                                       ↓
                                            X ↓, M ↑  →  NX falls
                                                       ↓
                                            IS shifts back to original
                                                       ↓
                                            Y unchanged, NX crowds out G

The four cases, side by side

Regime	Monetary expansion	Fiscal expansion	Key mechanism
Fixed FX, capital mobile	Impotent (M endogenous)	Fully potent	Peg defence sterilises monetary policy and accommodates fiscal policy
Floating FX, capital mobile	Fully potent (amplified by FX)	Impotent (crowded out by FX)	The exchange rate absorbs all external pressure
Fixed FX, no capital mobility	Potent (closed-economy-like)	Potent	Capital controls insulate monetary policy
Floating FX, no capital mobility	Potent	Potent (no FX feedback)	FX still adjusts to balance current account only

The "diagonal" of the matrix — fixed FX + capital mobile (where monetary policy fails) and floating FX + capital mobile (where fiscal policy fails) — is what most textbooks call the Mundell-Fleming result. The two off-diagonal cases with capital controls are the world before financial liberalisation.

The impossible trinity: the deeper proposition

Step back from the comparative statics and the punchline becomes a constraint on policy itself. There are three things a national central bank might want simultaneously:

1. Free capital mobility. Citizens and firms can move money across borders without controls.
2. A fixed exchange rate. The currency is pegged to another (or to a basket, or to gold).
3. An independent monetary policy. The central bank sets its own interest rate to stabilise domestic output and inflation.

Mundell-Fleming says you cannot have all three at once. With capital mobile and the rate fixed, the domestic rate must equal r* and the central bank cannot deviate. With capital mobile and policy autonomous, the rate must float (because the policy rate diverges from r* and arbitrage flows force currency adjustment). With the rate fixed and policy autonomous, capital must be controlled (otherwise arbitrage would crush either the peg or the rate). The trilemma is sometimes drawn as a triangle whose three sides are the three policies — you pick one side and forfeit the opposite vertex.

Worked example: a peg under monetary stress

Consider a small open economy pegged to the dollar at e̅ = 1 unit of domestic currency per dollar. Suppose r* = 4% and the country's central bank holds $50 billion of foreign reserves. Domestic output is below potential and policymakers want to cut the policy rate to 2%. They expand the money supply by $10 billion in a typical open-market operation.

Within hours, money-market traders observe that the domestic rate has dipped below 4%. With perfect capital mobility the implied carry trade is overwhelming — short the domestic currency, long the dollar. Capital outflows immediately exceed the daily turnover of reserve operations. To defend the peg, the central bank must sell reserves at the rate at which the private sector wants to convert. Suppose $10 billion flows out before the original monetary impulse is fully unwound. Reserves fall from $50 B to $40 B; the central bank has bought back $10 B of domestic currency; the money supply is back where it started.

Initial:   M = 100 B,  reserves = 50 B,  r = 4%, e = 1
+ OMO:    M = 110 B,  reserves = 50 B,  r = 2% (instantaneous, before flows)
+ Defend: M = 100 B,  reserves = 40 B,  r = 4%
Net effect:   M unchanged,  10 B of reserves spent for zero gain.

If the central bank persists, it will simply spend down its reserves. Once reserves run out, the peg must break. This is the mechanism behind every classic currency crisis from the UK 1992 to Thailand 1997 to Argentina 2001-02: an attempt to use monetary policy against a peg that ends in either capitulation on the peg or capital controls.

Case study: ERM 1992 and Black Wednesday

The European Exchange Rate Mechanism, established in 1979, was a soft-peg system in which member currencies floated within narrow bands against the deutschmark. Through the 1980s the ERM operated reasonably well in a low-inflation environment. After German reunification in 1990, the Bundesbank tightened monetary policy aggressively to fight the inflationary impulse of absorbing East Germany, pushing German short rates above 9%.

Other ERM members faced the trilemma directly. Sterling, the lira, the franc, and the peseta all came under intense market pressure: with capital fully mobile inside Europe, the only way to defend the peg against German rates was to follow the Bundesbank up — into recessionary domestic territory. The UK in particular had joined the ERM at what proved an overvalued rate (DM 2.95 to the pound), and the recessionary cost of defending it was politically intolerable.

On 16 September 1992 — "Black Wednesday" — speculative pressure on the pound became overwhelming. The Bank of England raised its base rate from 10% to 12% to 15% in a single day, deployed roughly £27 billion of reserves, and still could not hold the rate. By 7:30 p.m. the UK announced it was leaving the ERM. Sterling devalued by about 15% within weeks. Italy left at the same time; Spain and Portugal devalued within the ERM bands. George Soros's Quantum Fund had taken a $10 billion short position against the pound and reportedly made over $1 billion of profit. The episode is the canonical empirical illustration of the trilemma: the UK had been trying to maintain free capital, fixed FX, and at least some monetary autonomy, and the third corner had to give.

Modern applications: where every country sits on the triangle

• United States. Free capital + independent Fed + floating dollar. The US gave up the fixed exchange rate when Bretton Woods collapsed in August 1971. Today the dollar floats against everything that floats; the Fed sets its own rate based on the US dual mandate.
• Eurozone. Members give up individual monetary policy. The ECB sets a single rate for the bloc and the euro floats against the rest of the world. Free capital movement is enshrined in the EU treaties. Each individual country has effectively swapped its national monetary policy for a vote on the ECB's governing council.
• Hong Kong. Free capital + peg to the US dollar (since 1983 at HK$7.80 ± 0.05) → Hong Kong has no independent monetary policy. The Hong Kong Monetary Authority's policy rate tracks the Fed's almost mechanically. This is the cleanest real-world instance of the "fixed FX + capital mobility → no own monetary policy" corner.
• China. Independent monetary policy + managed float against a currency basket + extensive capital controls. The renminbi is not fully convertible on the capital account: cross-border portfolio flows are tightly regulated, FDI requires approvals, and capital outflows by households are limited. This is the "give up free capital" corner of the trilemma; it is what allows the PBOC to run interest-rate and reserve-requirement policy targeted at domestic conditions.
• Bretton Woods (1944–1971). The post-war system fixed all major currencies to the US dollar (and the dollar to gold at $35/oz) while permitting independent national monetary policy. This worked because capital controls were universal in the immediate post-war decades — the IMF Articles of Agreement explicitly allowed (and most members imposed) restrictions on capital-account transactions. As capital markets re-integrated in the 1960s, the system became progressively unsustainable; Nixon closed the gold window in 1971.

Extensions and refinements

• Imperfect capital mobility. The textbook flat BP curve at r* is an idealisation. With imperfect capital substitution and country risk premia, BP has a finite positive slope and the policy results soften — monetary policy under fixed rates is partially effective, fiscal under floating rates is partially effective. The qualitative ordering of the four cases survives.
• Sticky prices and the IS-LM closure. Mundell-Fleming assumes fixed prices in the short run (the Keynesian closure). With flexible prices, the long-run results converge toward the classical neutrality of monetary policy and the irrelevance of exchange-rate regimes for real output.
• Large open economy. Two-country versions allow a "large" country to move r* through its own policy. Empirically this matters for the US — Fed policy is partially transmitted globally as a shift in r* itself.
• Dornbusch overshooting (1976). Adds rational expectations and sticky prices in the long run, producing the famous result that monetary expansion causes the exchange rate to overshoot its long-run depreciation before reverting. A direct descendant of Mundell-Fleming with forward-looking FX markets.
• New Open Economy Macro. Microfounded versions (Obstfeld-Rogoff 1995 onwards) replace ad-hoc consumption and investment functions with optimising households and firms but preserve the trilemma core.
• The "dilemma not trilemma" critique (Rey 2013). Hélène Rey argued that the global financial cycle — driven by US monetary policy — propagates across borders so strongly that even floating-rate economies face a binary choice between capital mobility and monetary autonomy, regardless of the exchange-rate regime. The trilemma collapses toward a dilemma in tightly integrated markets.

Mundell's other Nobel-cited idea: optimum currency areas

The 1999 Nobel citation paired Mundell-Fleming with Mundell's 1961 paper "A Theory of Optimum Currency Areas." That paper asks a related question: when should two regions share a currency? Sharing a currency is the extreme of fixing the rate — irrevocably — and is therefore one corner of the trilemma made permanent. OCA theory provides the welfare framework for the trade-off: a currency union reduces transaction costs, increases trade, and disciplines policy, but it removes the exchange rate as a shock absorber. The benefits dominate when:

• The regions face symmetric shocks (so a single monetary policy suits both).
• Labour and capital can move freely between the regions (substituting for the lost exchange-rate adjustment).
• Fiscal transfers exist to absorb asymmetric shocks (federal redistribution).
• Wages and prices are flexible enough to adjust in place of the FX.

The Eurozone is the great natural experiment. OCA theorists were already arguing in the 1990s that Europe failed several of these tests — labour mobility is much lower than in the US, fiscal transfers are very limited, and the south-versus-north shock profile is asymmetric. The 2010–2012 sovereign-debt crisis exposed exactly the missing pieces. The OCA framework predicted these stress points; the eventual policy response (the ESM, the banking union, mutualised crisis instruments) is best understood as the bloc retrofitting the institutions the OCA literature said it needed all along.

Common pitfalls

• Conflating "fixed rates make fiscal policy work" with "any fiscal expansion is a free lunch." The result is comparative-static and assumes the economy is below capacity. Long-run sustainability of the peg requires policy and the price level to remain consistent with r*.
• Treating the BP curve as exogenous to expectations. Markets are forward-looking; the textbook flat BP at r* abstracts from time-varying risk premia and expected devaluation. In real crises, expected-devaluation premia open up enormous wedges between domestic and world rates.
• Reading the trilemma as static. Countries move along the trilemma over time. Bretton Woods used capital controls; the late 20th century unwound them; some emerging markets have selectively reimposed them after crises (Malaysia 1998, Iceland 2008). The choice is a policy variable, not a structural feature.
• Forgetting that capital mobility is a continuum. "Perfect" mobility is the textbook limit; real capital markets have transaction costs, information frictions, regulatory wedges, and country risk premia. The qualitative trilemma logic still applies, but the quantitative bite varies with the degree of integration.
• Applying it to large economies as if they were small. The "small open economy" assumption — r = r* exogenous — is unrealistic for the US, China, and the Eurozone. For these, the model is best read as a two-country or multi-country system in which r* itself is endogenous.