M.I.T.
LIBRARIES
-
DEWEY
Digitized by the Internet Archive
in
2011 with funding from
Boston Library Consortium Member Libraries
http://www.archive.org/details/consumptionrecesOOblan
working paper
department
of economics
massachusetts
institute of
technology
50 memorial drive
Cambridge, mass. 02139
Consumption and the recession of 1990-1991
Olivier Blanchard
MIT and NBER
No.
93-5
M.I.T.
LIBRARIES
Jan. 1993
-
DEWEY
J.T.
UBRAR
2 1993'
RED:; -•
AUG
1
Consumption and the recession of 1990-1991.
NBER
*
looks at the recession of 1990-91.
It
Olivier Blanchard,
MIT
and
January 15, 1993
Abstract
The paper
Not
all
recessions have the
reaches two conclusions:
same dynamics. Decreases
in
output which come from
other sources than consumption shocks tend to be short lived, and followed by sharp
recoveries. Decreases in output
in
which come from consumption shocks, decreases
consumption given income, tend
to be longer lasting,
and followed by weaker
recoveries.
In contrast to
its
predecessors, the recession of 1990-91 was caused primarily by a
consumption shock. Thus, conditional on the nature of those shocks, the slow
covery comes
as
no
surprise.
As
of the end of 1992, the
economy was
on the track given by a dynamic forecast based on information up
of 1991.
re-
nearly exactly
to the first
quarter
Recession
2
During the
two quarters of 1990 and the
last
first
quarter of 1991, the
experienced negative growth. These three quarters have
the 1990-1991 recession by the
NBER. But both
year following, growth was anemic as well.
has averaged 0.7% at an annual rate, a
In contrast to
cause.
And
its
in the year
Growth
far cry
now been
from
US economy
designated as
preceding and the
since the beginning of 1989
its
post- 1973
mean
of 2.2%.
predecessors, this recession does not have an obvious proximate
precisely because of that, explanations abound, ranging, to cite only
the leading candidates, from a long expansion dying of old age, to consumer depression, a
mix of debt overhang and the
realization of lower
growth prospects,
to
the credit crunch, the combination of imprudent behavior by financial institutions
and the tightening of regulation,
to the
end of the Cold War and the decrease
defense spending, to structural adjustments required by global competition
The purpose
structure
of this paper
to look at the data, putting just
is
on the econometrics
causes of the recession.
The
to pinpoint
story
if
consumption
in relation to
shocks are long
its
is
relatively clear.
"consumption shock"
to
animal
By
far,
the
a decrease in
effects of
such
why, in contrast to previous recoveries,
weak one. The
"consumption shock" was due
enough economic
normal determinants. Because the
lasting, this also explains
the recovery was a slow and
a
*.
not the deep, at least the proximate
which emerges
main proximate cause of the recession was
in
issue left
unanswered
is
whether
this
spirits/taste shocks, or simply to antici-
pations of the slow growth to come. Circumstancial evidence points to a role of
animal
spirits,
but will not convince
all.
Recession
I.
3
Looking
at the
A. Estimating a
Components
is
to estimate the joint
look at the residuals.
services (C),
(CD), residential investment
(IR),
I
non
by trend output, obtained by
(Y), as the
sum
of consump-
residential (INR) investment, govern-
CD,
log-difference C,
GNP
consumption expenditures on durables
(G), inventory investment
achieve stationarity,
NX
behavior of the components of GNP, and
thus decompose real
I
on non durables and
ment spending
GNP.
VAR.
A simple first pass
tion
of
fitting a
(1NV) and net exports (NX), lb
IR,
INR and
G, and divide
INV and
broken exponential trend to GNP, with
a break in 1973.
I
then run a
VAR
seven transformed variables,
for the
for the period
1992:3, with 3 lags of each variable, a constant and a post-73
1959:1 to
dummy
each
in
equation. Issues such as treatment of trend, or the incorporation of cointegrating
relations, while important
import
when
the focus
if
we were
as here,
is,
to look at impulse responses, are of little
on the
residuals to
each equation.
Given the 7 estimated equations, an
auxiliary equation
the behavior of GNP and
Given the non-linear transformations of the
its
residual.
needed
is
to characterize
individual series used to induce stationarity, there are two ways of doing
first is
to construct
residuals in
The
other
side variables of the
is
the one
residuals for
each period
is
to regress log-differenced
VAR, and
use below.
The
weighted average of the
GNP
ratios of
on the
each compo-
set of right
hand
use the residuals from this equation. This second
yields results very similar to the
I
as a
each equation, multiplied by the time varying
nent to GNP.
approach
GNP
this.
first,
and
is
simpler to implement. This
4
Recession
B. Identifying the Shocks.
Denote the
residuals to
each equation by lower case
letters, for
example
c for
the consumption equation, y for the -auxiliary- output equation. These residuals
are simply forecast errors.
They
are,
not surprisingly, generally positively corre-
lated across equations. This reflects their joint
ing shocks, as well as their direct
consumption responds
sumption
We
residual,
c,
to
dependence on each
income within the quarter
reflects also the
two identifying assumptions. The
that, for
GNP
example,
separately
on
ir,
c
is
underly-
To the extent that
example, part of the con-
to the underlying shocks by
assumption
is
depends only on y -the residual
making
that, within the quarter, the
depend on each other only through
inr and so on.
spending residual,
first
for
other.
common
shocks to other components.
can make some progress and get closer
components of
dependence on
to the
The second assumption
is
GNR
Thus,
I
assume
output equation-, not
that g, the
government
exogenous. Under those two assumptions, g can be used as an
instrument to estimate the effects of y on each of the other six residuals. Denote
the residuals from these instrumental variable regressions by
consumption
are
still
for
e,
thus using
e c for
example, and refer to them as the "shocks". These new residuals
cross-correlated, but less so than the original ones.
The two
identifying assumptions are crude,
and forecast
errors in
government
spending are not a very powerful instrument. Nethertheless, the estimates of the
contemporaneous
effects of y
on
its
components make
sense. Using sample
mean
values to go from estimated elasticities to derivatives, the estimates imply that a
one dollar increase
in
GNP
increases consumption of
non durables and
by 12 cents, durable consumption by 11 cents, residential investment by
non
residential investment by
6 cents, net exports by
-8 cents,
services
-1
cent,
and inventory
in-
Table
1.
The
source of shocks
Accumulated normalized shocks
Uv/°v)
E(Cc/^c)
EO.rMr)
1989:
0.00
0.00
0.00
1990:2
2.32
-2.50
0.25
1990:3
1.38
-2.29
-1.15
1990:4
0.79
-4.07
-1.61
1991:1
-0.39
-5.42
-2.87
1991:2
0.18
-5.35
-2.17
1992:3
0.46
-6.12
-0.15
Quarter
The
estimation of the shocks, the
series follow
random walks with
example, starting
mean
zero
at
e's, is
zero
described in the text.
mean and
0.00 in 1989:1, the distribution of the
and standard deviation \/l4
=
3.74.
The accumulated
unit standard deviation.
sum
in
Thus
for
1992:3 has
5
Recession
vestment by -15 cents.
The next
step
is
then to look
at
the sequence of shocks, the € x 's, and look for
unusually large individual shocks, or for sequences of shocks of the same sign.
A
convenient way
to
do so
to look at the
is
components, each normalized by
some date which choose
I
first
column
to the various
estimated standard deviation, starting from
its
to be the first quarter of slow growth, 1989:2.
results are presented in table
The
accumulated shocks
gives the
The main
1.
sum
of normalized
GNP
despite the fact that output growth was below
its
residuals, y/cr y
mean
in
.
It
shows
that,
1989 and early 1990,
residuals to the output equation were, during that period, small but positive. Thus,
the
VAR
explains the slowdown of output pre-1990:3 through the internal dy-
namics of the economy rather than through adverse shocks.
following this finding through;
The second and
tion of
These
third
services,
and
for those,
than 2 standard deviations).
first
it
yet.
to residential
to
consump-
investment respectively.
two components of GNP which show large negative shocks
though, interestingly, even
nates the
have not done
would be worth
columns give the sum of accumulated shocks
non durables and
are the only
I
It
A
no
single residual
tion dominate the next two quarters.
Two
While the
the recession, the whole period since 1989:2
is still
larger in absolute value
negative shock to residential investment domi-
quarter of the recession, 1990:3.
consumption. The sum
is
(al-
is
large in 1992:3.
negative shocks to consump-
largest
shocks take place during
dominated by negative shocks
The
those shocks, their nature and their implications.
rest of the
to
paper focuses on
6
Recession
II.
Looking
at
Output and Consumption
A. From 7 to 2 Variables.
7-variable
VAR's
are unwieldy, their dynamics hard to understand
describe. Fortunately, examination of the
and harder
dynamic structure of the
VAR
to
above
suggests a short cut. Tests of the significance of each variable in each equation
(which, because of space constraints, are not reported) show the strong predictive
power of non durables and services consumption
for nearly all
components of
GNP; and the weak predictive power of most other components. This suggests
we may not
that
lose too
consumption. This
I
is
much
what
I
do
by focusing
on
a bivariate system in output
and
in this section.
thus estimate a bivariate system in the logarithm of consumption of non durables
and
services,
and the logarithm of GNR There
that the ratio of
consumption
to
terms of first differences, allowing
income
is
is
weak evidence of cointegration,
stationary.
for three lags of
I
thus specify the
VAR
in
each first-differenced variable,
the lagged value of the log consumption to income ratio, a constant and a post-
1973
B.
2
dummy
.
From Residuals
to
Shocks
Estimation of the system yields residuals
respectively.
1
follow the
using the unexpected
regression of c
on
y.
same approach
component
This yields an
an increase of consumption of
income.
for
consumption and income,
as before to get to
c
and y
consumption shocks,
of government spending as an instrument in a
elasticity of
.
1
2 of consumption to income, thus
7 cents within the quarter for a dollar increase in
The "consumption shock",
ec
,
is
defined as the residual of this regression.
1
o
03
o
O
i
i
o
i
O
i
o
i
•vl
IS3
en
O
b
o
CD
p
CO
L
8
S
3
3
I
m
o
o
o
o
-fc.
N3
1
1
o
o
o
o
o
o
o
N3
l
1
I
o
o
o
o
en
CD
l
Tl
o
m
o
>
-
>
O
—1
c
>
I
i
—1
i
U3
r
O
*
t
1
i
|
i
|
/
|
8
I
t
t
:;
...
.
I
':/y
~M
0-0*:
/
W+M
:M
::
I
}\
\
1
\
1
I
\
1
cz
70
<n
I N>
•
\
\
\
1
V
1
\
>i
f
I
\
\
\
\
\
\
\
\
\
\
\
\
J
\
\
v
I
\
1
\
\
\
I
\\
\
o
o
1
1
o
o
o
o
o
o
o
o
o
o
o
o
o
-b.
en
CO
o
70
m
O
>
>
o
—i
c
>
in
8
i
V
\
A
':
V
\
4
/
t
<>>
O
m
Recession
A
7
regression of y in turn
on
an
e c yields
elasticity of .5
1
(thus an increase of 96
cents in income for an increase in consumption of a dollar, a multiplier slightly
below one within the quarter), and an "income shock"
marizing, the
mapping from
residuals to shocks
c
=
A2y +
y
=
.51e c
An examination of the sequence of realized
in table
The
1.
shock, which
vestment.
why,
we know from
is
Sum-
ec
+
ey
shocks gives a picture similar to that
is
dominated by an income
can be traced mostly
to
non
residential in-
quarters are dominated by consumption shocks. This
why
the key to understanding
proceed in two steps,
I
earlier
.
-given by:
quarter of the recession, 1990:3
The next two
finding
last
first
is
as the residual, e y
first
the recovery has been so weak. To see
looking at the dynamic effects of both income
and consumption shocks, and then looking
at
dynamic
forecasts of output as of
1991:1.
C. Impulse Responses, and the
The
are
to
Recovery.
consumption and income shocks
effects of one-standard-deviation negative
drawn
in Figure
income have only
largely
1.
The
figure has
transitory effects
gone within two years
long lasting,
tion
Weak
hump
3
.
two main
features.
The
first is
on consumption and income;
The second
is
that shocks to
that shocks
their effect
is
consumption have
shaped, effects on output, and to a lesser extent
on consump-
4
.
Now consider Figure 2. Figure 2 gives both actual output and the dynamic forecast
8
Recession
for
last
output based on the bivariate
VAR
using information up
and including the
quarter of the recession, 1991:1 (Thus, both lines are the same up to and
including 1991:1).
What
is
since 1991, but also that a
striking
is
not only
weak recovery
is
also
how weak
the recovery has been
what would have been predicted
by the bivariate model as of the end of 1991:1. Indeed the actual recovery has
been
slightly faster
than the dynamic
forecast...
(Note that the dynamic forecasts
are independent of our identification of the shocks.).
For comparison purposes, Figure 3 performs the same exercise for the previous
recession, giving output, actual
sion, 1982:4.
Note
and forecast
that, in that case, the
the strong recovery which followed
as of the last quarter of that reces-
same
bivariate
model
largely predicts
5
.
Given our decomposition of the shocks, and the characterization of impulse
sponses, the key to Figure 2 and Figure 3
is
easily given.
More
re-
so than in pre-
vious recessions, the decrease in output in 1990-91 was due to adverse shifts in
consumption. Those
dynamic
shifts
effects explain
have
why
long,
hump shaped
effects
on output, and
the recovery has been slow. There
is
their
an important
lesson: recessions are not necessarily followed by fast recoveries; this
depends on
the nature of the shocks which triggered the recession.
HI. Foresight, or Animal Spirits
I
have so
far
?
established that the recession was associated with large negative
"consumption shocks", that such shocks have long
that this explains
set of questions.
why
lasting effects
on output, and
the recovery has been so slow. This raises however the next
What
are these "consumption shocks"
and where do they come
9
Recession
from
One
And,
?
where did those of 90-91 come from
in particular,
interpretation, call
"foresight",
it
reflection of anticipations by
income.
The
is
that
consumption shocks are simply the
consumers of other shocks and
interpretation of impulse responses in Figure
place;
first
consumption shocks
pretation however
sumption not due
that
is
to
from increasing prudence,
to
outside of the usual maximizing
ing, panic,
1
show how
on future
then that consump-
Another
movements
inter-
in
con-
income. Reasons may range
in intertemporal preferences, to -stepping
model- sudden
realizations of past overborrow-
and so on. Under that interpretation,
responses in Figure
tion of
reflect in part
in expectations of future
changes
1 is
are a mirror, not a cause.
consumption shocks
changes
their effect
income which triggered them
tion shocks are simply followed by the changes in
in the
?
shifts in
call
it
"animal
spirits",
consumption lead through
dynamic multiplier and accelerator
effects to a
hump shaped
a
impulse
combina-
response of
output.
Can one
tell
these two interpretations apart
?
To some extent, one can. Under
the animal spirits interpretation, and under the plausible assumption that animal
spirits
to a
have
little
or
no long run
effect
on output, the impulse response of output
consumption shock should eventually return
in Figure
1,
to zero.
This
is
clearly violated
suggesting that consumption shocks must reflect in part foresight of
shocks with permanent
effects.
The
foresight interpretation
on the other hand
imposes constraints on the relation of the consumption response
to the subse-
quent response of income, and of the shape of the consumption impulse response
itself.
Following the route of putting more theoretical structure on the data to try
to disentangle foresight
and animal
spirits
would take me too
drop out of the above econometric framework, and
offer
my
far.
I
shall instead
speculations
on the
Recession
1
nature of the negative consumption shocks of 1990-91:
I
see the overall evidence, from the timing of monthly declines in the index of
consumer confidence (from the Michigan Survey of consumers)
,
in the
leading indicators, and in commercial forecasts of output, as pointing to
consumer
index of
more than
foresight at work. First, in contrast to earlier recessions, the decline in
confidence was largely prior to -and
much
stronger than would have been pre-
dicted by- either the decline of leading indicators or commercial forecasts of the
recession. Second, perhaps not coincidentally, the
in
first
August 1990, was associated with an important but
large decline in confidence
largely
non economic event,
the invasion of Kuwait by Irak. Third, after having dropped, consumer confidence
remained very low
in the following
two
years,
much
lower than would have been
predicted on the basis of historical relations with aggregate variables. Each piece
of evidence can, with
together however,
I
some
find
effort,
be reconciled with the foresight interpretation;
them suggestive of a
role for a
from more than the expectation of the recession.
drop in confidence coming
Recession
I
References
Cochrane, John, "Permanent and Transitory Components of
Stock
Prices",
mimeo, University of Chicago,
Sinai, Allen,
ing,
GNP
and
July 1992.
"What's Wrong with the economy
?",
Challenge, forthcom-
1992
Walsh, Carl,
"What caused
the 1990-1991 recession
Reserve Bank of San Francisco, 1992.
?",
mimeo, Federal
Recession
1
Footnotes
*
Department of Economics, MIT, 50 Memorial Drive, Cambridge
I
thank the
NSF
for financial support,
Roger Brinner, Ricardo Caballero,
Rotemberg, Allen Sinai and Lawrence Summers
Julio
MA 02 139.
for useful discussions
and comments.
1
Two
useful discussions of potential causes are given by Allen Sinai
Walsh
2
.
Such systems have been estimated by many
ticular focuses also
come
The
on the
different
dynamic
shocks. His approach to identification
from that used
3
and Carl
others.
John Cochrane
effects of consumption
is
however somewhat
in par-
and
in-
different
in this paper.
fact that the
long run effect of a negative shock
embarassing, but this long run effect
is
is
positive
is
mildly
not significantly different from zero.
emphasized by Cochrane.
4
This point
5
Results for the post -recession episodes of 1971 and 1975 are qualitatively
is
also
similar to those for 1982, with fast predicted
each recession.
and actual growth following
'78
2
Date Due
291994
Lib-26-67
MIT LIBRARIES DUPl
3
TOAD 0Dfl32T0M
^