L ECTURE 8 Monetary Policy at the Zero Lower Bound October 19, 2011 - PowerPoint PPT Presentation

Economics 210c/236a Christina Romer Fall 2011 David Romer L ECTURE 8 Monetary Policy at the Zero Lower Bound October 19, 2011

I. P AUL K RUGMAN , “I T ’ S B AAACK : J APAN ’ S S LUMP AND THE R ETURN OF THE L IQUIDITY T RAP ”

Krugman’s Baseline Model – Assumptions (I) • Discrete time. • Identical, infinitely-lived agents. • Representative agent has U = ∑ t D t ln c t , 0 < D < 1. • Each agent receives an endowment y of the consumption good each period. • Can sell endowment for money, and buy goods with money. • Economy is competitive and prices are perfectly flexible (!). • Perfect foresight.

Krugman’s Baseline Model – Assumptions (II) • Cash-in-advance constraint. Within period t: • Agents start with some holdings of money and bonds (from period t-1). • There’s then a market for trading money and bonds. • Call the representative agent’s holdings after these trades M t and B t . • The cash-in-advance constraint is c t ≤ M t /P t . • After the agent has bought and sold goods, it receives interest on its bond holdings, and any lump-sum taxes or transfers are implemented. • The cash-in-advance constraint and perfect foresight imply that c t = M t /P t or i t = 0 (or both).

Households’ First-Order Condition • Suppose the economy is in equilibrium, and consider an agent thinking of spending $1 less on c t and using the proceeds to increase c t+1 . = • MC ( 1 / P )( 1 / y ) t = + • MB [( 1 i ) / P ]( D / y ) + t t 1 = − • => … => (*) i ( 1 / D )( P / P ) 1 + t t 1 t • Note that this holds even if i t = 0.

The Steady State with Constant M • Suppose M is constant at some level (denoted M*). • If there is a steady state, P is constant. Call this P*. = − • Then equation (*), , simplifies to i ( 1 / D )( P / P ) 1 + t t 1 t = − = − for all t, or i t ( 1 / D ) 1 i * ( 1 D ) / D . • Note that i* > 0.

The Possibility of a “Liquidity Trap” • Assume that starting in Period 2, the economy is in steady state. • So P 2 = P*, i 2 = i* > 0. = − • So (*) becomes i ( 1 / D )( P * / P ) 1 1 1 i 1 CC 0 P 1

The Possibility of a “Liquidity Trap” (cont.) • Households’ allocation of wealth between money and bonds in period 1: • If i 1 > 0: M 1 /P 1 = y => P 1 = M 1 /y. • If i 1 = 0: M 1 /P 1 ≥ y => P 1 ≤ M 1 /y. MM i 1 0 P 1

The Effects of an Increase in M 1 when i 1 > 0 i 1 CC MM MM’ 0 P 1

The Effects of an Increase in M 1 when i 1 = 0 i 1 CC MM MM’ 0 P 1

The Effects of an Increase in M* when i 1 = 0 Recall CC equation: i 1 = (1/D)(P*/P 1 ) - 1 i 1 CC CC’ MM 0 P 1

Some More Experiments (I) • Suppose the economy is in a liquidity trap in periods 1 and 2, then in steady state with i = i* > 0. Raising M 1 or M 2 has no effect on aggregate demand in any period. But raising M* raises aggregate demand in period 2 and in period 1. • Continue to assume a liquidity trap in period 1 and steady state starting in period 3. Suppose initially i 2 > 0. Raising M 2 to the point where i 2 = 0 raises aggregate demand in period 1. That is, when the economy is in a liquidity trap, promising to stay in the trap longer rises aggregate demand.

Some More Experiments (II) • Consider raising M by the same proportion in all periods. Then P rises by the same proportion in all periods. • Suppose the economy is in steady state starting in period 2, and suppose the central bank targets a zero inflation rate from period 1 to period 2. Thus its choice of M* moves one- for-one with movements in P 1 . Then if something pushes the equilibrium real rate in period 1 below 0, there is no equilibrium: P 1 falls without limit. Inflation targeting eliminates any nominal anchor for the economy.

FOMC Statement, Aug. 12, 2003 “The Committee judges that, on balance, the risk of inflation becoming undesirably low is likely to be the predominant concern for the foreseeable future. In these circumstances, the Committee believes that policy accommodation can be maintained for a considerable period.”

II. B EN B ERNANKE , “J APANESE M ONETARY P OLICY : A C ASE OF S ELF -I NDUCED P ARALYSIS ?”

Channels of Monetary Policy Transmission • Nominal interest rates. • Expected inflation. • Asset prices. • The extent of credit-market imperfections. • The real exchange rate (and expectations about the real exchange rate). • Expectations abut future output. • The price level (and expectations about the price level).

Tools of Monetary Policy at the Zero Lower Bound • Communication about objectives, or the formal adoption of new objectives. • Communication about future path of safe short-term interest rate (or of supply of high-powered money). • Communication about the channels of monetary policy (such as the exchange rate or future output). • Purchases of assets other than short-term government debt. • Conventional open-market operations? • Money-financed fiscal expansions (helicopter drops)?

Some Important Questions • Could some of the tools be counterproductive? • Could the mix of outcomes (especially, in terms of output and inflation) be different for these tools than for conventional open-market operations in normal times?

The Overnight Call Rate in Japan

The Monetary Base in Japan, 1994–2011 From: Bank of Japan

III. O VERVIEW

Nominal Interest Rate on 3- to 6-month Treasury Notes 6 5 4 3 2 1 0 Jan-29 Jul-29 Jan-30 Jul-30 Jan-31 Jul-31 Jan-32 Jul-32 Jan-33 Jul-33 Jan-34 Jul-34 Jan-35 Jul-35

IV. G AUTI E GGERTSSON , “G REAT E XPECTATIONS AND THE E ND OF THE D EPRESSION ”

Industrial Production (Logarithms) 1.2 1.4 1.6 1.8 2.2 2.4 1 2 1929-01 1929-05 1929-09 1930-01 1930-05 1930-09 Industrial Production 1931-01 1931-05 1931-09 1932-01 1932-05 1932-09 1933-01 1933-05 1933-09 1934-01 1934-05 1934-09 1935-01 1935-05 1935-09 1936-01 1936-05 1936-09 1937-01 1937-05 1937-09

Producer Price Index, Logarithms 2.3 2.4 2.5 2.6 2.7 2.8 2.9 Jan-29 Producer Price Index, All Commodities Jun-29 Nov-29 Apr-30 Sep-30 Feb-31 Jul-31 Dec-31 May-32 Oct-32 Mar-33 Aug-33 Jan-34 Jun-34 Nov-34 Apr-35 Sep-35 Feb-36 Jul-36 Dec-36 May-37 Oct-37

Logarithms 2.8 2.9 3.1 3.2 3.3 3.4 3.5 3 1932:01 1932:03 1932:05 1932:07 1932:09 1932:11 1933:01 1933:03 1933:05 1933:07 1933:09 1933:11 1934:01 M1 1934:03 1934:05 1934:07 1934:09 1934:11 1935:01 1935:03 1935:05 1935:07 1935:09 1935:11 1936:01 1936:03 1936:05 1936:07 1936:09 1936:11

What are the key elements of the regime? • Gold standard • Commitment to a balanced budget • Belief in small government

What is the mechanism by which the regime change affected inflationary expectations? • Fiscal expansion gives the government an incentive to inflate. • So, fiscal expansion leads to monetary expansion.

What is Eggertsson’s evidence of regime change? • Narrative: Roosevelt quotes. • Actions

From Temin and Wigmore, “The End of One Big Deflation”

Evaluation of Evidence • Timing of actions • What happened to spending?

Federal Receipts, Outlays, and Surplus 15,000 Outlays 10,000 Receipts 5,000 0 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 Surplus -5,000 -10,000

From Temin and Wigmore, “The End of One Big Deflation”

From: Temin and Wigmore, “The End of One Big Deflation”

V. C HRISTINA R OMER , “W HAT E NDED THE G REAT D EPRESSION ?”

Millions of Dollars 1000 1500 2000 2500 3000 3500 4000 4500 -500 500 0 1919 1920 1921 1922 Gold Inflows to the U.S. 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940

Mishkin Method of Estimating Ex Ante Real Rate Ex Post Real Rate: r ep t = i t – π t where i is the nominal rate and π is actual inflation. Ex Ante Real Rate: r ea t = i t – π e t Where π e is expected inflation.

The difference between r ep and r ea is unanticipated inflation ( ε t ): r ep t = (i t – π t )+ ( π e t – π e t ) r ep t = (i t – π e t ) – ( π t – π e t ) = r ea t – ε t • Under rational expectations, expectation of unanticipated inflation at a point in time is zero. • You can’t expect to be surprised.

Think of constructing estimate of π e : π e t = αi t + β’X t where X is a vector of information known at time t. r ep t = i t – ( αi t + β’X t ) + ε t r ep t = (1 – α)i t – β’X t + ε t Regress r ep on i, and other explanatory variables known at time t. Fitted values are estimates of r ea .

Behavior of Different Types of Consumer Spending 3.5 Nondurables 3 2.5 Services 2 1.5 Durables 1 0.5 0 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942