Discussion of Eggertsson and Mehrotra
Patrick J. Kehoe
June 8, 2015
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 8, 2015
1/9
Goal
Develop a model in which
(Natural) Real Interest Rate can be permanently negative
Nominal wage rigidity generate long – lived fall in employment
Use the model to think about slow recoveries
Japan Crisis
Great Depression
Great Recession
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 8, 2015
4/9
Real Rate Negative
In a deterministic model, if gross real rate R = (1 + r ) < 1
If borrow b0 at 0, and roll over,
Owe R t b0 at t. This goes to zero.
So, present value budget constraint not well defined.
Eggertsson and Mehrotra shows:
In OG model no such problem.
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 8, 2015
5/9
Backward-looking wages
Let w ∗ = f 0 (l ∗ ) be real wages in the frictionless allocation.
Nominal wage norm
Wt = γWt −1 + (1 − γ) (Pt w ∗ )
Can have real wage permanently higher than w ∗ ?
Yes, if permanent deflation π = Pt /Pt −1 < 1,
Wt
W t − 1 Pt − 1
=γ
+ (1 − γ ) w ∗ ,
Pt
Pt −1 Pt
so,
wt
1−γ
=
> 1 iff π < 1
∗
w
1 − γ/π
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 9, 2015
7 / 10
Author’s takeaways
Deflation per se not critical
rather even at zero nominal interest rates
inflation is not high enough
Want story for why nominal rate = 0 long time
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 8, 2015
7/9
Story why nominal rate is 0 for long time
Central Bank chooses to set it to 0
Why choose i = 0?
if economy below trend Central Bank should set i low
Why it stays at 0?
monetary policy alone not powerful enough to cure problems
Example: take aspirin when get sick
if mildly sick it helps
if have cancer: take aspirin for 20 years but still dies
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 8, 2015
8/9
EM story
Persistently high real wages can account for
Japanese crisis
Great Depression
Great Recession
Such distortions show up as labor wedge
ul
< Fl = w
uc
⇐⇒
ul /uc
= 1 − τl < 1
Fl
Not as a productivity shock
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 8, 2015
9/9
A Typical Labor Wedge Recession: US
From 2008–2014
Data
Wedges
110
105
Output
Hours
Investment
100
100
90
95
80
70
2008
A
1 l
1/(1+ x)
90
2009
2010
2011
2012
2013
Y and L fall about 8%
Brinca, Chari, Kehoe, McGrattan
2014
85
2008
2009
2010
2011
2012
2013
2014
A falls about 3%, τl rises about 8
percentage points
Wedge Accounting
Spring 2015
22 / 43
Equilibrium Responses: US
One wedge at a time
Equilibrium Responses and Data--Y
Equilibrium Responses and Data--L
105
105
A
1 l
1/(1+ x)
A
1 l
1/(1+ x)
100
100
95
95
Data
Data
90
2008
2009
2010
2011
2012
2013
2014
τl accounts for 6% of 8% fall in Y
Brinca, Chari, Kehoe, McGrattan
90
2008
2009
2010
2011
2012
2013
2014
τl more than accounts for fall in L
Wedge Accounting
Spring 2015
23 / 43
Equilibrium Responses: US
One wedge at a time
Equilibrium Responses and Data--I
Equilibrium Responses and Data--C
110
A
1 l
1/(1+ x)
100
A
1 l
1/(1+ x)
105
100
90
95
80
Data
Data
70
2008
2009
2010
2011
2012
2013
2014
A and τl account for most of fall in I
Brinca, Chari, Kehoe, McGrattan
90
2008
2009
2010
2011
2012
2013
2014
A and τl account for most of fall in C
Wedge Accounting
Spring 2015
24 / 43
A Typical Efficiency Wedge Recession: Italy
From 2008–2014
Data
Wedges
110
105
Output
Hours
Investment
100
A
1 l
1/(1+ x)
100
90
95
80
90
70
2008
2009
2010
2011
2012
2013
Y falls 10% and L falls 5%
Brinca, Chari, Kehoe, McGrattan
2014
85
2008
2009
2010
2011
2012
2013
2014
A falls about 8%, τl rises modestly
Wedge Accounting
Spring 2015
25 / 43
Equilibrium Responses: Italy
One wedge at a time
Equilibrium Responses and Data--Y
Equilibrium Responses and Data--L
110
105
A
1 l
1/(1+ x)
A
1 l
1/(1+ x)
100
100
90
95
Data
Data
80
2008
2009
2010
2011
2012
2013
2014
A accounts for 8% of 10% fall in Y
Brinca, Chari, Kehoe, McGrattan
90
2008
2009
2010
2011
2012
2013
2014
τl accounts for 2% of 4% fall in L
Wedge Accounting
Spring 2015
26 / 43
Equilibrium Responses: Italy
One wedge at a time
Equilibrium Responses and Data--I
110
Equilibrium Responses and Data--C
A
1 l
1/(1+ x)
100
A
1 l
1/(1+ x)
110
100
90
80
90
Data
Data
70
2008
2009
2010
2011
2012
2013
2014
A accounts for most of fall in I
Brinca, Chari, Kehoe, McGrattan
80
2008
2009
2010
2011
2012
2013
2014
A accounts for most of fall in C
Wedge Accounting
Spring 2015
27 / 43
A Typical Investment Wedge Recession: Iceland
From 2008–2014
Data
Wedges
120
120
A
1 l
1/(1+ x)
Output
Hours
Investment
100
100
80
80
60
40
2008
2009
2010
2011
2012
2013
2014
Y falls 12% and L falls 10%
Brinca, Chari, Kehoe, McGrattan
60
2008
2009
2010
2011
2012
2013
2014
τx rises by about 25 percentage points
Wedge Accounting
Spring 2015
28 / 43
Equilibrium Responses: Iceland
One wedge at a time
Equilibrium Responses and Data--Y
Equilibrium Responses and Data--L
120
120
A
1 l
1/(1+ x)
110
110
100
100
90
90
Data
80
2008
2009
2010
2011
2012
Data
80
2013
2014
τx accounts for 6% of 12% fall in Y
Brinca, Chari, Kehoe, McGrattan
A
1 l
1/(1+ x)
2008
2009
2010
2011
2012
2013
2014
τx more than accounts for fall in L
Wedge Accounting
Spring 2015
29 / 43
Equilibrium Responses: Iceland
One wedge at a time
Equilibrium Responses and Data--I
140
Equilibrium Responses and Data--C
A
1 l
1/(1+ x)
120
100
120
A
1 l
1/(1+ x)
110
100
80
90
60
80
40
Data
Data
70
20
2008
2009
2010
2011
2012
2013
2014
2008
τx accounts for all of fall in I
Brinca, Chari, Kehoe, McGrattan
Wedge Accounting
2009
2010
2011
2012
2013
2014
But τx makes C rise
Spring 2015
30 / 43
Decomposition Statistics
Want statistic fiy that measures fraction of movement in
output from wedge i
P
With fiy ∈ [0, 1],
fi = 1, fiy = 1 when var(yt − yit ) = 0
Our statistic
1/var(yt − yit )
fiy = P
j (1/var(yt − yjt ))
where yit is the component of output due to wedge
i = (A, τl , τx , g)
Similar statistics for other wedges
Next Decomposition Statistics for OECD countries
Brinca, Chari, Kehoe, McGrattan
Wedge Accounting
Spring 2015
16 / 43
Labor and Efficiency wedge contributions
100
Two wedges
account for 90%
for 70%
Contribution of Labor Wedge
90
80
70
60
50
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100
Contribution of Efficiency Wedge
Brinca, Chari, Kehoe, McGrattan
Wedge Accounting
Spring 2015
17 / 43
Labor and Efficiency wedge contributions
100
Two wedges
account for 90%
for 70%
Contribution of Labor Wedge
90
80
70
Labor-wedge
driven recessions
60
50
USA
ESP
KOR
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100
Contribution of Efficiency Wedge
Brinca, Chari, Kehoe, McGrattan
Wedge Accounting
Spring 2015
18 / 43
Labor and Efficiency wedge contributions
100
Two wedges
account for 90%
for 70%
Contribution of Labor Wedge
90
80
70
Labor-wedge
driven recessions
60
50
USA
ESP
Efficiency-wedge
driven recessions
KOR
40
30
20
AUT
10
JPN
0
0
10
20
30
40
50
NZL GBR CAN
AUS
DNK
ITA SWE
BEL
GER
CHE
NOR
FRA
FIN NLD
60
70
80
90
100
Contribution of Efficiency Wedge
Brinca, Chari, Kehoe, McGrattan
Wedge Accounting
Spring 2015
19 / 43
Labor and Efficiency wedge contributions
100
Two wedges
account for 90%
for 70%
Contribution of Labor Wedge
90
80
70
Labor-wedge
driven recessions
60
50
USA
ESP
Efficiency-wedge
driven recessions
KOR
40
30
20
AUT
ISL
10
JPN
0
0
10
20
30
40
50
NZL GBR CAN
AUS
DNK
ITA SWE
BEL
GER
CHE
NOR
FRA
FIN NLD
60
70
80
90
100
Contribution of Efficiency Wedge
Brinca, Chari, Kehoe, McGrattan
Wedge Accounting
Spring 2015
20 / 43
Summary
To explain Europe and Japan,
Need theory of endogenous productivity decline.
Not theory of the labor wedge.
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 8, 2015
1/2
Japan vs United States
Inflation and interest rates
Output growth
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 9, 2015
6 / 10
−.02
0
.02
.04
.06
Japan
1970m1 1975m1 1980m1 1985m1 1990m1 1995m1 2000m1 2005m1 2010m1 2015m1
Inflation
T−Bill
0
.05
.1
.15
.2
USA
1970m1 1975m1 1980m1 1985m1 1990m1 1995m1 2000m1 2005m1 2010m1 2015m1
Inflation
T−Bill
To account for Japan
Need theory of,
Endogenous fertility decline
Endogenous productivity decline
Will nominal rate = 0 be critical?
P. Kehoe
Discussion of Eggertsson and Mehrotra
June 8, 2015
2/2