Optimal Monetary Policy Theory and Lessons V. V. Chari University of Minnesota and Federal Reserve Bank of Minneapolis Big Picture Lessons Commitment best way to do policy – Promise to do right thing is not commitment – Gold standard is commitment Rules only way to think of policy – Inflation targeting is not a rule – Taylor Rule is a rule Optimality: Typically different responses to different shocks – Taylor Rule does not have property – Optimal policy depends on model details Smaller Picture Lessons Friedman Rule: Zero nominal interest rates + deflation Standard Model: Friedman Rule optimal Distorting Taxes: Friedman Rule still optimal Sticky Prices: Get slight deflation Punchline: In most models optimal inflation in [–3,0] Friedman Rule Private cost of money = nominal interest rate (R) Social cost = 0 Hot potato problem Policy: Set nominal interest rate to zero Distorting Taxes R = 0 implies deflation Deflation needs declining M Tax revenues needed to shrink M Phelps: Trade off tax distortions versus hot potato problem Surprisingly still get R = 0 with distorting taxes Sticky Prices Falling prices implies sectoral distortions Want to keep price level constant Want to get R to zero for hot potato reasons Optimal policy is compromise (inflation in [–3,0]) Optimal policy responds differently to different shocks ○ Expand M in response to technology shocks ○ Contract M in response to fall in money demand The Problems of Large Deflations Households have endowment e Cash-in-advance constraint Prices stuck one period at a time ct1 1 t Preferences Keep future P, output fixed Pc M Inefficiency Possible Equilibrium summarized by ○ Py M , y e P1 e , i0 P y ○ (1 i) As M , may hit i = 0 before y = e Not much of a problem if P higher than –3 percent P Lessons for Policy Design Commitment important No consensus on detailed policy rule How about ○ Pick policy rule for, say, 3 years ○ Require supermajority on FOMC to deviate from rule