Ch. 12: Output, Exchange Rates and Macroeconomic Policies

Based on Feenstra and Taylor, Chapter 18 (KOM chapter 17). There are no lecture slides for this topic on the site yet; check against the lecture when it is given.

In Chapter 11 output Y was taken as given and we studied how money and expectations move the interest rate and the exchange rate. This chapter makes output endogenous. The result is the IS-LM-FX model, the standard Keynesian model of a small open economy in the short run. It answers the questions policy makers care about most: can monetary and fiscal policy raise output in a recession, how do they affect the exchange rate and the trade balance, and how does the answer change when the exchange rate is fixed? The chapter has five parts: (1) demand in the open economy; (2) goods market equilibrium (Keynesian cross); (3) the IS curve; (4) the LM curve; (5) the full IS-LM-FX model with policy under floating and fixed exchange rates, and the limits of stabilization policy, including the J curve.

Assumptions

• Prices are sticky in the short run: home P and foreign P* are fixed. So any change in the nominal exchange rate E is also a change in the real exchange rate q = EP*/P.
• Output is demand-determined: firms supply whatever is demanded at the given prices (there is spare capacity).
• The foreign economy is given: foreign output Y*, foreign taxes T* and the foreign interest rate i* are exogenous.
• Policies are temporary unless stated otherwise, so the expected future exchange rate E e is fixed.
• Government spending G and taxes T are set by the government; investment depends on the interest rate.

Part 1: Demand in the open economy

Total demand for home output is:

Consumption. C rises with disposable income. The marginal propensity to consume (MPC) is the fraction of an extra euro of disposable income that is spent, between 0 and 1.

Investment. Firms borrow to invest. A higher real interest rate raises the cost of borrowing, so I falls when i rises. With sticky prices and stable expected inflation, the nominal and real interest rates move together.

Trade balance. Three determinants:

1. Real exchange rate q = EP*/P. A rise in q (real depreciation) makes home goods cheaper relative to foreign goods. Foreigners buy more home exports, home residents switch from imports to home goods, so TB rises. This is expenditure switching. (Strictly, the value of imports in home goods also rises with q, so the TB improves only if quantities respond enough: the Marshall-Lerner condition, that the sum of the export and import demand elasticities exceeds 1. We assume it holds.)
2. Home disposable income. Higher Y minus T raises spending on everything, including imports, so TB falls. The marginal propensity to import (MPM) is the part of each extra euro of disposable income spent on imports.
3. Foreign disposable income. Higher Y* minus T* raises foreign demand for home exports, so TB rises.

Evidence: real exchange rates and the trade balance. FT show that large real depreciations go with trade balance improvements, for example the US trade deficit narrowed after the dollar's fall in the late 1980s, and in many emerging market crises (Mexico 1994, Asia 1997, Argentina 2002) a sharp real depreciation was followed by a large swing toward trade surplus, helped by falling home demand for imports.

Part 2: Goods market equilibrium: the Keynesian cross

The goods market is in equilibrium when supply (output Y) equals demand D. Draw demand D on the vertical axis and output Y on the horizontal axis.

• The 45-degree line shows all points where D = Y (supply equals demand).
• The demand line D slopes upward: as Y rises, consumption rises by MPC, but part of that goes to imports, so demand for home goods rises by MPC minus MPM. The slope is less than 1, so the demand line is flatter than the 45-degree line.
• Equilibrium is where the two cross. If output is below equilibrium, demand exceeds output, inventories fall and firms raise production.

What shifts the demand line up (and raises equilibrium Y):

Changes in Y itself move us along the demand line, not shift it.

Part 3: The IS curve (goods and forex market equilibrium)

The IS curve shows all combinations of output Y and interest rate i at which the goods market and the forex market are both in equilibrium. Axes: i vertical, Y horizontal. It slopes downward. Take a fall in the home interest rate:

1. Investment channel: lower i raises I, so D rises.
2. Exchange rate channel (the open-economy addition): lower i shifts the domestic return line left in the FX market, so the home currency depreciates (E rises). With sticky prices this is a real depreciation, so TB rises and D rises further.
3. Higher demand raises equilibrium output through the multiplier. So lower i goes with higher Y: a downward-sloping IS.

The exchange rate channel makes the open-economy IS curve flatter than the closed-economy one: a given cut in i raises output more because it works through both I and TB.

Shifts of the IS curve. Anything that raises demand at a given i shifts IS to the right:

• Fiscal expansion: G up or T down.
• Higher foreign income Y* (or lower T*).
• Higher foreign prices P* or lower home prices P (real depreciation at a given E).
• Changes that depreciate the currency at a given i: a rise in i* or a rise in E e. Both shift the foreign return curve up in the FX market, so E is higher at every i, TB is larger and IS shifts right.
• Autonomous shifts in consumption, investment or trade preferences (for example foreigners switching toward home goods).

Part 4: The LM curve (money market equilibrium)

The LM curve shows all combinations of Y and i at which the money market is in equilibrium: M/P = L(i)Y. It slopes upward: higher output raises money demand; with M/P fixed, the interest rate must rise to keep demand equal to supply.

Shifts of the LM curve. A rise in M (with P sticky) raises real money supply, so the interest rate must be lower at every Y: LM shifts down (to the right). A fall in M or a rise in P shifts LM up (left). An autonomous rise in money demand (L rises at every i) shifts LM up.

Part 5: The IS-LM-FX model

FT draw two panels side by side. Panel (a): the IS-LM diagram, i vertical and Y horizontal. Panel (b): the FX market, E vertical and returns horizontal. The equilibrium interest rate from panel (a) is the domestic return line in panel (b). The intersection of IS and LM gives Y and i; at that i, the FX diagram gives E. All three markets (goods, money, forex) are in equilibrium at once.

Macroeconomic policy under floating exchange rates

Temporary monetary expansion (floating)

1. M rises, LM shifts right (down).
2. i falls; in the FX market the domestic return line shifts left, so E rises (home depreciation).
3. Lower i raises I; depreciation raises TB. Output Y rises.

Result: Y up, i down, E up (depreciation), C up, I up, TB up. Monetary policy is very effective under floating because the exchange rate channel reinforces the interest rate channel.

Temporary fiscal expansion (floating)

1. G rises (or T falls), IS shifts right.
2. Higher Y raises money demand, so with M fixed i rises (move along LM).
3. Higher i shifts the domestic return line right: the home currency appreciates (E falls).
4. The appreciation lowers TB, and higher i lowers I: both partly offset the fiscal stimulus (crowding out).

Result: Y up (but less than the simple multiplier says), i up, E down (appreciation), C up, I down, TB down. Fiscal policy works under floating but is weakened by crowding out of investment and of the trade balance.

Macroeconomic policy under fixed exchange rates

Under a credible peg, E is fixed and E e = E, so UIP forces i = i*. In the IS-LM diagram this means the economy must always end up on the horizontal line i = i*. The central bank must adjust M to whatever level keeps i at i*: LM becomes endogenous (it shifts to wherever IS crosses i = i*).

Temporary monetary expansion (fixed)

If the central bank raises M, LM shifts right and i starts to fall below i*. In the FX market the home currency comes under pressure to depreciate. To defend the peg the central bank must sell foreign reserves and buy home currency, which reduces M again. LM shifts back. Result: no change in Y, i or E. Monetary policy is ineffective under a fixed exchange rate with capital mobility (the trilemma from Chapter 11).

Temporary fiscal expansion (fixed)

G rises, IS shifts right. At the old M, i would rise and the currency would appreciate. To prevent appreciation the central bank must buy foreign currency and sell home currency, raising M: LM shifts right until i is back at i*. Result: Y rises strongly, i and E unchanged, TB falls only through the income effect (higher Y raises imports), no crowding out of investment. Fiscal policy is very effective under a peg, because monetary policy is forced to accommodate it.

The rule to remember: under floating, monetary policy is strong and fiscal policy weakened; under a peg, monetary policy is powerless and fiscal policy strong. (This is the Mundell-Fleming result.)

Foreign shocks

The model also shows how shocks from abroad are transmitted. A rise in the foreign interest rate i* under floating shifts the foreign return curve up, depreciates the home currency, raises TB and shifts IS right: home output rises. Under a peg the home central bank must raise i to i*, contracting M: home output falls. A fall in foreign demand (lower Y*) shifts IS left in both regimes; under floating the resulting fall in i depreciates the currency and cushions the fall in output, while under a peg the full fall in demand hits output unless fiscal policy responds.

Stabilization policy and its limits

In principle, policy makers can use monetary or fiscal policy to offset shocks and keep output near its full-employment level. Example: a fall in foreign demand shifts IS left; under floating, a monetary expansion can shift LM right to restore output. This is stabilization policy.

Application: Poland is not Latvia (2008 to 2009)

In the global financial crisis both countries faced a collapse in demand. Poland had a floating zloty. As the crisis hit, the zloty depreciated sharply, interest rates were cut, and the depreciation boosted net exports. Poland was the only EU economy to avoid a recession in 2009. Latvia kept its peg to the euro. It could not cut interest rates independently (in fact rates rose as markets doubted the peg) and could not depreciate; it also had to cut government spending. Latvian output fell by well over 10% in 2009, one of the deepest recessions in the world. The contrast illustrates the model: with a peg, the country gives up the monetary and exchange rate tools that a floating country can use to absorb a demand shock. Latvia instead had to go through an internal devaluation: years of falling wages and prices to regain competitiveness.

Problems with stabilization policy

• Policy constraints: a fixed exchange rate rules out monetary policy; high public debt or fiscal rules can rule out fiscal policy.
• Incomplete information and lags: recessions are recognized late, and policy works with long and variable lags, so policy can end up stimulating a recovered economy.
• Long-horizon plans: if households and firms see a policy as temporary, they may not change consumption or investment much.
• Weak links: in a liquidity trap the interest rate is at zero and cannot fall further, so monetary policy loses its interest rate channel.
• Pass-through and the J curve: the trade balance may respond weakly or slowly to the exchange rate (below).
• Political economy: fiscal policy is often used for reasons other than stabilization, and is hard to reverse.

Pass-through

Exchange rate pass-through is the fraction of an exchange rate change that shows up in the prices of traded goods. If US imports are invoiced in dollars and foreign exporters keep their dollar prices fixed (pricing to market), a dollar depreciation does not raise US import prices, so there is no expenditure switching. With pass-through below 100%, the real exchange rate relevant for trade moves less than E, and the trade balance responds less. Pass-through is low for the United States (because so much trade is priced in dollars) and higher for small open economies.

The J curve

After a depreciation the trade balance often worsens first and improves only later. Plotted over time, the trade balance traces a J shape.

• Short run: trade contracts and orders were placed in advance, so import and export quantities barely change. But each unit of imports now costs more in home currency. The value of imports rises, so TB falls.
• Longer run (often six months to a year or more): quantities adjust, exports rise and imports fall, and TB improves, once the Marshall-Lerner condition holds.

Numerical example. The US imports 100 units at €1 each and exports 110 units at $1 each, with E $/€ = 1.00: TB = 110 minus 100 = +$10. The dollar depreciates to 1.10. In the first months quantities stay the same: imports now cost 100 × 1.10 = $110, so TB = 110 minus 110 = 0. The TB has worsened. A year later exports rise to 125 units and imports fall to 90 units: TB = 125 minus 90 × 1.10 = 125 minus 99 = +$26. The TB has improved beyond its start.

Application: US policy in the early 1980s

A classic illustration (also in KOM): the Federal Reserve under Volcker tightened money sharply to fight inflation while the Reagan administration ran large fiscal deficits. In the model, both the monetary contraction (LM left) and the fiscal expansion (IS right) raise the US interest rate and appreciate the dollar. The dollar rose strongly between 1980 and 1985 and the US trade balance moved into large deficit: the "twin deficits" (budget and trade). The model predicts exactly this combination; the output effects of the two policies partly offset each other.

KOM view: the DD-AA model and permanent policies

KOM present the same model with the exchange rate E on the vertical axis and output Y on the horizontal axis:

• DD schedule (goods market): slopes upward. A depreciation raises the trade balance and demand, so output rises.
• AA schedule (asset markets: money plus forex): slopes downward. Higher Y raises money demand and i, which appreciates the currency.

A monetary expansion shifts AA up (right): Y up and E up. A fiscal expansion shifts DD right: Y up and E down. These are the same results as IS-LM-FX. KOM add the case of permanent policies, where E e changes too. A permanent monetary expansion raises E e and so shifts AA further, a bigger effect on Y and E (with overshooting). A permanent fiscal expansion lowers E e (an expected long-run real appreciation), shifting AA down as well: the currency appreciates so much that the trade balance falls by about as much as the fiscal stimulus, and output may not rise at all (complete crowding out in KOM's benchmark case). KOM also discuss the J curve and pass-through in the same way as FT, and show that the Marshall-Lerner condition is what makes DD slope upward.