Download Discussion of “International Coordination” by

Discussion of
“International Coordination”
by
Jeffrey Frankel
Discussion by Charles Engel
University of Wisconsin
Policy Challenges in a Diverging Global Economy,
Federal Reserve Bank of San Francisco, 19-20 November, 2015
1
The Aim of the Paper
• There have been increasing calls for international monetary and
fiscal policy coordination.
• The paper makes many important observations but the chief point is
that there is disagreement about models, which leads to disagreement
about the nature of spillovers of policy.
Fiscal policy
1. Some feel that non-cooperative policy is too contractionary. Positive
spillovers are not as great as they could be because countries are
concerned about trade deficits.
2. Some feel that non-cooperative policy is too expansionary. Countries
run deficits that are too large, meaning somebody else will have to bail
them out.
2
The Aim of the Paper
My comments will focus on the monetary policy part:
Monetary policy
1. Some people believe non-cooperative monetary policy is too
expansionary. Currency war – nobody depreciates, but perhaps monetary
policy is too loose?
2. Some feel monetary policy is excessively contractionary – too
concerned about inflation, but ignore spillovers.
3
Two Propositions
• We can all agree that examining the benefits of cooperation is very
difficult. The lessons we learn are very much model dependent.
Proposition 1
No analysis of the gains from cooperation in a particular model is
general enough to be useful.
Proposition 2
Nothing that can be said about cooperation that is general is useful.
• My comments:
1. A series of general observations. I think we all agree on them.
But by Proposition 2, they are useless.
2. A new general observation.
4
General Comment #1: Sufficient Number of Instruments
• Suppose Home and Foreign policymakers target output, y and y * .
Their targets are y and y *.
• Suppose they each have an effective instrument, m and m * .
• There are spillovers so output depends on both policies:
y = F ( m, m *)
y* = G ( m*, m )
• Then there should be a set of policies m and m * that achieve y and
y *.
• There is no need for cooperation. Each policymaker adjusts her
instrument until she achieves her target.
• There is no need for agreement on the model!
5
General Comment #2: Insufficient Number of Instruments
The problems come when there are more targets than instruments.
• For example, a country may have both an inflation and an output
target, and they cannot simultaneously hit both.
• Or they may have a target for capital flows, the current account,
financial stability, etc.
Spillovers from another country may change the tradeoff.
The problem comes when spillovers worsen the tradeoff.
Are there gains from cooperation?
1. They may be small. (Obstfeld and Rogoff, QJE, 2002)
2. Policymakers may disagree on the model. Especially if posterior
beliefs are not much influenced by data, there may be little point in
cooperating. (Frankel, 2015)
6
General Comment #3: Currency War?
• Notice that nothing I have said so far presumes the exchange rate is a
target of the non-cooperative policymaker.
• Perhaps it is, but policymakers usually deny it.
• Instead, the exchange rate is an endogenous variable that changes
when the policymaker changes its instrument in order to hit its
targets.
• It may be the vehicle of spillovers, but as Frankel notes, there are
other channels of spillovers. Notably import demand and capital
flows.
1. Are spillovers negative?
2. Can domestic policy be adjusted to correct for the spillovers?
3. Can better outcomes be achieved for both through cooperation?
7
General Comment #4: Is the Fed non-Cooperative?
• Let me characterize some comments by Fed policymakers this way
(though this may be a caricaturization rather than a characterization):
• “Our legal mandate is to achieve low inflation and high employment.
We use our policy instruments to achieve those goals. We don’t pay
attention to the rest of the world – that is not our mandate.
So we are not engaged in non-cooperative policy.”
• But that is exactly what non-cooperative policy is: maximizing your
own welfare taking the other guy’s policy instrument as given.
• Non-cooperation doesn’t mean that you are obstinate or deliberately
evil.
• Why cooperate? Because cooperation may allow you to achieve the
Humphrey-Hawkins objectives better than non-cooperation.
8
General Comment #5: What is Different under the ZLB?
• Does the ZLB mean that there is no scope for cooperation? Is the
exchange rate the only mechanism for expansion?
• The ZLB does change the mechanism, but as Frankel points out,
there are still mechanisms for monetary policy to affect the economy
(not to mention fiscal policy!)
• Caballero, Farhi, and Gourinchas (NBER, 2015)
• Engel (AER, 2016) – the model (based on Nagel, 2015) gives a
simple way to think about this. There are near-money assets such as
T-bills that have a liquidity return. Liquidity comes from these assets
usefulness as collateral or to meet balance sheet requirements.
• In a sense, even when the pecuniary return is zero, the total return is
not at the ZLB. Quantitative easing lowers the liquidity return on
near-money assets because it increases the supply of reserves.
9
Main Observation (page 1): The Global Cooperative Objective
• The objective of the global policymaker may not simply be the sum
of the objectives of the individual policymakers.
• We might think that the Home policymaker wants to minimize some
weighted sum of squared inflation and squared output gap, and so
does the Foreign policymaker. Possibly other objectives as well.
• But from a global standpoint, there may be more to the objective
function.
• In particular, I argue that minimizing currency misalignment is an
objective (Engel, AER 2011) for the global policymaker.
• How is it possible that the global objective function is not a weighted
sum of each country’s objective function?
10
Main Observation (page 2): Quick Primer on New Keynesian economics
• What is the difference in the New Keynesian economics and regular
old Keynesian economics?
• I think one of the key differences is the public finance approach to
policy.
• Rather than thinking of some ad hoc goal of policy (output,
inflation), the objective is to reduce distortions caused by price
stickiness, monopoly power, wage stickiness, credit constraints, etc.
• We want to think about how a particular policy may reduce the
distortions, and how we trade off the losses from different
distortions.
11
Main Observation (page 3): Utility Functions and Loss Functions
Example from a simple model:
• The policymaker wants to maximize the utility of households, which
depends on consumption and leisure
Et ∑ j =0 U ( Ct + j , Lt + j )
∞
• But, cleverly, the literature has shown how in some cases we can
transform the objective into minimizing a loss function:
Et ∑ j =0 V ( yt + j − yt + j , π t + j − π t + j ) ,
∞
where yt + j and π t + j are output and inflation levels in an efficient
(“Wicksellian”) economy. This is appealing both intuitively and as
a pedagogical device.
12
Main Observation (page 4): We Can Sum Utility Fns. but Not Loss Fns.
• Certainly it is reasonable to state the objective of the global
policymaker as a weighted sum of Home and Foreign utility:
ω Et ∑ j 0=
U ( Ct + j , Lt + j ) + (1 − ω ) Et ∑ j 0 U * ( Ct*+ j , L*t + j )
∞
∞
• But this does not translate to a weighted sum of loss functions:
ω Et ∑ j 0=
V ( yt + j − yt + j , π t + j − π t + j ) + (1− ω ) Et ∑ j 0 V * ( yt*+ j − yt*+ j , π t*+ j − π t*+ j )
∞
∞
The global policymaker may have different concerns.
• Intuitively, the individual country loss functions do not incorporate
the loss imposed on the other economy under non-cooperative
policy.
13
Main Observation (page 5): Currency Misalignments
• An important example of what might matter from the global
perspective is currency misalignment.
• Suppose both countries were producing at full employment and had
zero inflation? Why would we care about currency misalignment?
• With local-currency consumer price stickiness, consumers in
different countries could be paying very different prices for identical
or near-substitute goods. (The purchasing power of US consumers
has dropped 20 percent relative to European and Japanese consumers
in past year.)
• With nominal wage and price stickiness, production patterns may be
misaligned. For example, with a weak euro, German exporters may
be advantaged relative to US firms. The US economy may then tilt
too much toward non-traded goods and services.
14
Conclusions: Global Distortions
• To reiterate, in the end, policymakers care about the welfare of
individuals.
• From the New Keynesian perspective, loss functions are a
convenient, intuitive way to summarize utility.
• While global welfare is a weighted sum of each country’s welfare,
global losses are not a weighted average of each country’s losses.
• There are global distortions, in other words. These are things that
matter for global welfare, but no country finds it in their individual
interest to target.
• This is the key point missing from Bernanke’s (IMF, 2015) recent
discussion, and is implicitly missing from Frankel, 2015.
15