The Current Financial Crisis in Spain:
What Should We Learn from the Great Depressions
of the Twentieth Century?
Timothy J. Kehoe
University of Minnesota
and Federal Reserve Bank of Minneapolis
18th Barcelona Lecture
Barcelona G.S.E. and Banc Sabadell
February 2010
www.econ.umn.edu/~tkehoe

Modeling Great Depressions
not
Explaining Great Depressions
To explain something, I need to understand it.
To understand it, I try to model it.

Modeling Great Depressions
not
Explaining Great Depressions
To explain something, I need to understand it.
To understand it, I try to model it.
My colleagues and I am making progress on modeling Great
Depressions.
We have not yet arrived at understanding, much less
explaining.

Data show that the current financial crisis in the United States is not
unprecedented in the postwar period.
Furthermore, there are signs that a recovery may be in sight.

S&P 500 (log scale)
10000
1000
100
10
Jan-50 Jan-60 Jan-70 Jan-80 Jan-90 Jan-00 Jan-10

International data show a darker picture.

Source: Eichengreen and O’Rourke

Source: Eichengreen and O’Rourke

To sum up, globally we are tracking or doing even worse than the
Great Depression, whether the metric is industrial production,
exports or equity valuations. Focusing on the U.S. causes one to
minimize this alarming fact. The “Great Recession” label may
turn out to be too optimistic. This is a Depression-sized event.
That said, we are only one year into the current crisis, whereas
after 1929 the world economy continued to shrink for three
successive years.
Barry Eichengreen and Kevin H. O’Rourke (June 2009), “A
Tale of Two Depressions.”

A September 2009 update on Eichengreen-O’Rourke

World industrial production

World stock markets

World trade

What do the new data tell us?
Global industrial production now shows clear signs of recovering.
This is a sharp divergence from experience in the Great Depression, when the
decline in industrial production continued fully for three years. The question
now is whether final demand for this increased production will materialise or
whether consumer spending, especially in the US, will remain weak, causing the
increase in production to go into inventories, leading firms to cut back
subsequently, and resulting in a double dip recession.
Global stock markets have mounted a sharp recovery since the beginning of
the year. Nonetheless, the proportionate decline in stock market wealth
remains even greater than at the comparable stage of the Great Depression.
The downward spiral in global trade volumes has abated, and the most recent
month for which we have data (June) shows a modest uptick. Nonetheless,
the collapse of global trade, even now, remains dramatic by the standards of
the Great Depression.

Great Depressions
of the Twentieth
Century Project
Timothy J. Kehoe and
Edward C. Prescott
www.greatdepressionsbook.com

Cole and Ohanian, “The Great Depression in the United States
from a Neoclassical Perspective,” Federal Reserve Bank of
Minneapolis Quarterly Review, Winter 1999.
Federal Reserve Bank of Minneapolis Conference, October 2000.
Special Issue of Review of Economic Dynamics, January 2002.
Great Depressions of the Twentieth Century, July 2007.

15 studies by 26 researchers using the same methodology
Great depressions
1930s
United States, United Kingdom, Canada, France, Germany
Recent
Argentina (1970s and 1980s), Chile and Mexico (1980s), Brazil
(1980s and 1990s), New Zealand and Switzerland (1970s, 1980s,
and 1990s), Argentina (1998-2002)
Not-quite-great depressions
Italy (1930s), Finland (1990s), Japan (1990s)

Kehoe and Prescott define a great depression to be a large negative
deviation from balanced growth.
They set the growth rate in the balanced growth path to be 2
percent per year, the growth rate of output per working-age person
in the United States during the twentieth century.

Real GDP per Working Age Person in the United States, 1900-2009
3.00
800
2.00
400
index (1900 = 100)
200
1.00
trend
real GDP
100
0.00
-1.00
50
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
year

Real GDP per Working Age Person in Spain, 1900-2009
9.64
800
8.64
400
index (1900=100)
trend
7.64
200
real GDP
6.64
100
50
5.64
1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
year

Trend growth:
yti
ˆ t
ˆi
y0 , 1.02
Great depression:
D [t0 , t1 ] such that
t t0
1. There is some t in D in such that yti / yti0
ˆ 1 0.20.
t t0
2. There is some t t0 10 such that yti / yti0 1 0.15.
3. There are no t1 , t2 in D, t2 t1 10, such that
t 2 t1
yti2 / yti1 1 0.

Great depressions in the 1930s:
Detrended output per person
110
100
France Germany
Index (1928 = 100)
90
United States
80
70
Canada
60
50
1928 1930 1932 1934 1936 1938

Great depressions in the 1980s:
Detrended output per working-age person
110
100
Chile
Index (1980 = 100)
90 Brazil
80
Mexico
70
Argentina
60
50
1980 1982 1984 1986 1988 1990

Great depressions methodology
Crucial elements: Growth accounting and dynamic general
equilibrium model
Growth accounting decomposes changes in output per working-age
person into three factors:
a productivity factor
a capital factor
an hours-worked factor

Great depressions methodology
Crucial elements: Growth accounting and dynamic general
equilibrium model
Growth accounting decomposes changes in output per working-age
person into three factors:
a productivity factor
a capital factor
an hours-worked factor
Keynesian analysis stresses declines in inputs of capital and
labor as the causes of depressions.

Balanced growth path
In the dynamic general equilibrium model, if the productivity
factor grows at a constant rate, then
the capital factor and the hours-worked factor stay constant and
growth in output is due to growth in the productivity factor.
Twentieth century U.S. macro data are very close to a balanced
growth path, with the exception of the Great Depression and the
subsequent World War II build-up.

Balanced growth path
1
Yt 1
Kt 1 Lt
A
t
Nt Yt Nt
When At 1 g 1 At
Kt Lt
and are constant
Yt Nt
Yt
grows at rate g 1, assume g 1 0.02 as in U.S.
Nt

Growth Accounting for the United States, 1970-2009
240
220
200
index (1970 = 100)
output
180
160
productivity
140
120 hours worked
100
capital
80
60
1970 1975 1980 1985 1990 1995 2000 2005

Growth Accounting for Spain, 1970-2009
240
220
200
productivity
index (1970 = 100)
180
160 output
140
120 capital
100
80 hours worked
60
1970 1975 1980 1985 1990 1995 2000 2005

Growth accounting for the United States 1929–1939
140
capital productivity
120
index (1929=100)
100
output
80
hours worked
60
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939

We use a dynamic general equilibrium model to model the
responses of households and firms — in terms of capital
accumulation and hours worked — to changes in productivity and
changes in government policy.
We take the path of the productivity factor as exogenous.
Comparing the results of the model with the data, we can identify
features of the depression that need further analysis.
Example: The Great Depression in the United States.

Growth accounting for the United States
140
capital productivity
120
index (1929=100)
100
output
80
hours worked
60
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939

Growth accounting for the United States
140
productivity
120
index (1929=100)
100
80
60
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939

Growth accounting for the United States
140
120
index (1929=100)
100
output
80
60
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939

Growth accounting for the United States
140
120 capital
index (1929=100)
100
80
60
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939

Growth accounting for the United States
140
120
index (1929=100)
100
80
hours worked
60
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939

Conclusions
A simple dynamic general equilibrium model that takes
movements in the productivity factor as exogenous can explain
most of the 1929-1933 downturn in the United States.
The model over predicts the increase in hours worked during the
1933-1939 recovery.
Need for Further Study
The decline in productivity 1929-1933
The failure of hours worked to recover 1933-1939

Lessons from Great Depressions Project
The main determinants of depressions are not drops in the inputs
of capital and labor — stressed in traditional theories of
depressions — but rather drops in the efficiency with which
these inputs are used, measured as total factor productivity
(TFP).
Exogenous shocks like the deteriorations in the terms of trade
and the increases in foreign interest rates that buffeted Chile and
Mexico in the early 1980s can cause a decline in economic
activity of the usual business cycle magnitude.
Misguided government policy can turn such a decline into a
severe and prolonged drop in economic activity below trend — a
great depression.

Growth Accounting for Mexico
140
120 capital
index (1980 = 100)
hours
100
output
80
productivity
60
1980 1985 1990 1995 2000 2005

A Decade Lost and Found:
Mexico and Chile in the 1980s
Raphael Bergoeing, Patrick J. Kehoe, Timothy J. Kehoe,
and Raimundo Soto
Similar crises in 1981-1983
more severe in Chile than in Mexico
Different recoveries
much faster in Chile than in Mexico
Why different pattern?

Real GDP per working-age (15-64) person
detrended by 2 percent per year
130
120
110 Chile
index (1980=100)
100
90
80
Mexico
70
60
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
year

Growth accounting and applied dynamic
general equilibrium model
Two numerical experiments with model:
Base case model: takes series for productivity factor as
given.
Model with tax reform: takes series for productivity factor as
given and imposes tax reform that lowers tax on capital
income in 1988 in both countries.

Applied dynamic general equilibrium
model
The representative consumer maximizes
t 1980
t log Ct (1 )log(hNt Lt )
subject to
Ct Kt 1 Kt wt Lt (1 t )( rt ) Kt Tt
where Tt t ( rt ) Kt is a lump-sum transfer.
Feasibility:
Ct Kt 1 (1 ) Kt At Kt L1
t .

Calibration
First order conditions:
1
1 (1 )(r )
C C t t
t 1 t
1 wt
.
hNt Lt Ct
Look at 1960-1980 data
Ct Ct
0.98, 1 1 0.45 in Mexico, 0.56 in Chile ;
(rt )Ct 1
Ct 0.30 in Mexico, 0.28 in Chile .
Ct wt (hNt Lt )

Numerical experiments
Base case:
t 0.45 in Mexico, t 0.56 in Chile , 1980-2000.
Tax reform:
t 0.45 in Mexico, t 0.56 in Chile , 1980-1988;
t 0.12 in Mexico, t 0.12 in Chile , 1988-2000.

Detrended real GDP per working-age person
and productivity factor
140
Chile output
120
index (1980=100)
100
Chile productivity
80
Mexico output
60
Mexico productivity
40
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000

Detrended real GDP per working-age person:
base case model
140
Chile data
120
index (1980=100)
100 Chile model
80
Mexico data
60
Mexico model
40
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000

Detrended real GDP per working-age person:
model with tax refrom
140
Chile model
120
Chile data
index (1980=100)
100
80
Mexico data
60
Mexico model
40
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000

What do we learn from growth
accounting and numerical experiments?
Nearly all of the differences in the recoveries in Mexico and Chile result
from different paths of productivity.
Tax reforms are important in explaining some features of the recoveries,
but not the differences.
Implications for studying structural reforms story:
Only reforms that are promising as explanations are those that show
up primarily as differences in productivity, not those that show up as
differences in factor inputs.
Timing of reforms is crucial if they are to drive the differences in
economic performance.

Fiscal reforms
Chile:
tax reforms 1975, 1984
social security reform 1980
fiscal surpluses
Mexico:
tax reforms 1980, 1985, 1987, 1989
fiscal deficits
Important, but not for explaining the
differences!

Trade reforms
Chile: by 1979
all quantitative restrictions eliminated
uniform tariff of 10 percent
tariff hikes during crisis — tariff back below 10 percent in 1991
Mexico: in 1985
100 percent of domestic production protected by import licenses
nontariff barriers and dual exchange rates
Massive trade reforms in Mexico 1987-1994, culminating in NAFTA
Timing seems wrong!

Privatization
Chile
major privatizations 1974-1979
Mexico
major nationalization 1982
expropriated banks’ holdings of private companies
government controlled 60-80 percent of GDP
major privatizations after 1989
Timing seems wrong?

Banking
Chile: 1982 and after
took over failed banks
market-determined interest rates
lowered reserve requirements.
Mexico: 1982 and after
nationalized all banks
government set low deposit rates
75 percent of loans either to government or directed by government.

Private credit as a percent of GDP
90
80
Chile
70
percent GDP
60
50
40
30 Mexico
20
10
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
year

Bankruptcy laws
Chile had reformed the administration of its bankruptcy
procedures in 1978. In 1982 it reformed its bankruptcy
laws to look much like those in the United States.
Mexico reformed its bankruptcy procedures in a similar
way only in 2000. (Maybe not so similarly!)

Business bankruptcies in Chile
500
400
number per year
300
200
100
0
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
year

Studying the experience of countries that have experienced great
depressions during the twentieth century teaches us that massive
public interventions in the economy to maintain employment and
investment during a financial crisis can, if they distort incentives
enough, lead to a great depression.

Concluding observations

Comparison of the 2007–2010 U.S. recession
with the 1981–1983 U.S. recession

U.S. Real GDP
106
104
index (peak = 100)
102
2007 IV
100
98
1981 III
96
94
0 1 2 3 4 5 6 7 8
quarter following peak

U.S. Unemployment Rate
12.0
July 1981
10.0
8.0
percent
6.0
December 2007
4.0
2.0
0.0
0 2 4 6 8 10 12 14 16 18 20 22 24
month following peak

Comparison of the 2007–2010 U.S. recession
with the 1981–1983 U.S. recession
Using conventional measures — real GDP, unemployment rate —
the 2007–2010 recession is not noticeably worse than the 1981–
1983 recession.

Comparison of the 2007–2010 U.S. recession
with the 1981–1983 U.S. recession
Using conventional measures — real GDP, unemployment rate —
the 2007–2010 recession is not noticeably worse than the 1981–
1983 recession.
The recovery during the 2007–2010 recession has been slower, and
there is still a danger that the current recession could turn into a
“double-dip” recession.

Comparison of the crises in the United States and Spain

Comparison of the Crises in the United States and Spain:
Real GDP per Working Age Person
108
106
Spain
index (2005 I = 100)
104
102 United States
100
98
2005 2006 2007 2008 2009

Comparison of the crises in the United States and Spain
Looked at over the past five years, Spanish growth experience
compares very favorably with that of the United States.

Comparison of the crises in the United States and Spain
Looked at over the past five years, Spanish growth experience
compares very favorably with that of the United States.
The recovery in the United States seems to have started in 2009 III.
The recovery in Spain does not seem to have started yet.

Those who try to justify the sorts of Keynesian policies
implemented by the Mexican government in the 1980s often quote
Keynes’s dictum from A Tract on Monetary Reform:
“The long run is a misleading guide to current affairs. In the long
run we are all dead.”

Studying past great depressions turns this dictum on its head:
“If we do not consider the consequences of policy for productivity,
in the long run we could all be in a great depression.”