Discussion of “Caution or Activism? Monetary Policy Strategies in an Open Economy” by Martin Ellison, Lucio Sarno and Jouko Vilmunen Ulf S¨oderstr¨om IGIER, Bocconi University June 2006

Background ˆ Optimal policy with unknown parameters: Trade-off control/estimation ˆ Common result:

– Passive learning: optimal policy more cautious – Active learning: optimal policy more activist ˆ Bertocchi and Spagat (1993), Balvers and Cosimano (1994), Wieland (2000) ˆ Intuition: Activist policy makes policymaker learn faster

Trade off worse control today for better control in future

1

This paper ˆ Passive vs Active learning in open economy ˆ Coordinated vs Non-coordinated policy ˆ Results:

– Passive learning: optimal policy less activist – Active learning: * Optimal policy more activist if coordination * Optimal policy less activist if no coordination ˆ Intuition: Activist policy also makes foreign central bank learn faster, which is

bad for domestic central bank ˆ More general: Non-coordination reduces incentives for experimentation

2

What do I think? ˆ Nice polished paper ˆ Important message: Optimal policy in open economy may be very different from

a closed economy ˆ Main comment: Are the results relevant in practice?

– Experimentation – Policy coordination

3

Experimentation ˆ Should central banks experiment to better understand the economy? ˆ Blinder (1998):

“While there are some fairly sophisticated techniques for dealing with parameter uncertainty in optimal control models with learning, those methods have not attracted the attention of either macroeconomists or policymakers. There are good reasons for this inattention, I think: You don’t conduct experiments on a real economy solely to sharpen your econometric estimates.” ˆ So what’s wrong with experimenting?

– Managing private expectations – Model misspecification

4

Experimentation and private expectations ˆ What is the effect of experimentation on private expectations? ˆ Ellison and Valla (2001):

“The problem with the activist policy is that it induces additional volatility into the inflation expectations of private agents.. . . This additional cost associated with the extra volatility in inflation expectations counteracts the call for greater activism. . . ” ˆ ESV ignore expectations ˆ Taking expectations seriously ⇒ more cautious policy also if coordinated?

5

Experimentation and model misspecification ˆ What if the model is misspecified? ˆ What will the central bank learn from experimentation? ˆ Beck and Wieland (2002): weaker incentives for experimentation if unknown

parameters vary over time ˆ And what if central bank and private agents have different models in mind? ˆ Transparency vs Experimentation

6

Policy coordination ˆ In practice, do central banks coordinate?

Sometimes: Plaza agreement 1985, EMS/EMU ˆ Do central banks try to exploit each other?

Sometimes: Competitive devaluations pre-EMU ˆ How large are the gains from coordination?

– Mundell (1968): negative spillovers, beggar-thy-neighbor policy – Obstfeld and Rogoff (1995): positive spillovers, coordination reduce welfare – Betts and Devereux (2000): coordination increases welfare if very high degree of pricing-to-market ˆ In ESV model, beggar-thy neighbor policy is optimal

Gains from coordination are large, much larger than gains from experimentation In practice gains may be small 7

Other issues ˆ Microfoundations?

– Expectations taken into account – Coordination increases or reduces welfare? – Proper welfare analysis ˆ Sources of uncertainty?

– No deviations from UIP: ρt = rt∗ − rt = i∗t − it ˆ Empirical implication: Central banks in large open economies less active than in

small open economies and closed economies ˆ The role of the discount factor

– More discounting ⇒ smaller incentives to experiment? – What discount factor is reasonable for the central bank? 8

Conclusions ˆ Nice well-written paper ˆ Good example of how open economies are different from closed economies ˆ Practical relevance?

– Experimentation may not be very attractive in practice – Monetary policy coordination may not be very important ˆ In any case: Non-coordination reduces incentives for experimentation ˆ Another argument against experimentation

9

References Balvers, Ronald J. and Thomas F. Cosimano (1994), “Inflation variability and gradualist monetary policy,” Review of Economic Studies, 61 (4), 721–738. Beck, G¨ unter and Volker Wieland (2002), “Learning and control in a changing economic environment,” Journal of Economic Dynamics and Control , 26 (9–10), 1359–1377. Bertocchi, Graziella and Michael Spagat (1993), “Learning, experimentation, and monetary policy,” Journal of Monetary Economics, 32 (1), 169–183. Betts, Caroline and Michael B. Devereux (2000), “International monetary policy coordination and competitive depreciation: A reevaluation,” Journal of Money, Credit, and Banking, 32 (4 (Part 1)), 722–745. Blinder, Alan S. (1998), Central Banking in Theory and Practice, The MIT Press. Ellison, Martin and Natacha Valla (2001), “Learning, uncertainty and central bank activism in an economy with strategic interactions,” Journal of Monetary Economics, 48 (1), 153–171. Mundell, Robert A. (1968), International Economics, MacMillan, New York. Obstfeld, Maurice and Kenneth Rogoff (1995), “Exchange rate dynamics redux,” Journal of Political Economy, 103 (3), 624–660. Wieland, Volker (2000), “Learning by doing and the value of optimal experimentation,” Journal of Economic Dynamics and Control , 24 (4), 501–534.

10