8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
1/44
The National ureau of Economic Research
The Political Business Cycle after 25 YearsAuthor(s): Allan DrazenSource: NBER Macroeconomics Annual, Vol. 15 (2000), pp. 75-117Published by: The University of Chicago PressStable URL: http://www.jstor.org/stable/3585387 .
Accessed: 06/11/2014 05:21
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp
.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
.
The University of Chicago Press and The National Bureau of Economic Research are collaborating with
JSTOR to digitize, preserve and extend access to NBER Macroeconomics Annual.
http://www.jstor.org
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/action/showPublisher?publisherCode=ucpresshttp://www.jstor.org/stable/3585387?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/3585387?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=ucpress
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
2/44
AllanDrazen
UNIVERSITY OF
MARYLAND,
HEBREW
UNIVERSITY OF
JERUSALEM,
AND NBER
The
Political usiness
ycle
after5 Years
1.
Introduction
A
quarter
f
a
century
as
passed
since the
nitial
utburst f formal
theoretical
nd
empirical
workon
political
usiness
cycles,
hat
s,
on
political
eterminantsf
macroeconomic
ycles.
On the
empirical
ide,
therewasKramer's1971) nfluentialtudy f conomic eterminantsf
U.S.
congressionaloting,
ollowed
y
thework
fTufte
1975, 978)
nd
Fair
1978).1
Nordhaus's
1975)
pioneering
ormal
model
of
the
political
business
ycle
PBC)
due
to
opportunisticre-electoral anipulation
as
published xactly wenty-five
ears
go.2
Soon
after,
ibbs
1977)
pre-
sented
model
of
partisan olicymakers
that
s,
policymakersaving
different acroeconomic
oals)
n
an
environmentimilar o that fthe
Nordhaus,
but
where these
partisan
ifferences ere
the
key
driving
force. erhaps sinfluentialn timulatingesearch as the 972 residen-
tial
lection
n
the
United
tates,
n
which
ncumbent ichard
ixon
was
justifiably
iewed s
engaging
n
significantre-electoral
anipulation.3
I
wishto thank
my
discussants,
lberto
lesina,
arl
Walsh,
nd
conference
articipants
and
seminar
articipants
t
theHebrew
University
f
Jerusalem
nd the
Bank f srael or
helpful
omments,
nd StefanHubrich
or
xtraordinarily
ble research
ssistance nd
many
ery
seful
iscussions. his
research as
supported
n
partby
theMaurice alk
Institute
or conomic
esearch,
ebrew
University
f
Jerusalem.
1.
Early
work
on connections etween
olitics
nd
fluctuations
n
economic
ctivity
s
reviewednKramer1971).2. Kalecki
1943)
presented
n
early
xplicit
model of the
PBC;
the
political
ature f
economic
luctuationsas
recognized y Schumpeter
1939)
n
his
study
f business
cycles.
imultaneously
ith
Nordhaus,
indbeck
1976)
presented
similar
dea;
soon
after,
cRae
1977)
lso
presented
formalmodel f hePBC.
3.
Rogoff
1988)
called
Nixon the
all-time ero of
political
usiness
ycles,"
t least n
contemporary
.S.
history.
ufte
1978)
begins
his
famous ook on the
PBC
with
quotation
rom
814,
A
Government
s
not
upported
hundredth
art
o
much
y
he
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
3/44
76
-
DRAZEN
Subsequent
to
this
flurry
f
research,
here
has been a
large
amount
of
further ork.
Theoreticalresearch
has
concentrated
n
making
both
op-
portunistic nd partisanmodels consistentwithvotersbehaving ratio-
nally,
both
in
forming xpectations
bout
future
olicy
and
in
voting
on
the basis of those
expectations.
The
success of
opportunistic re-electoral
manipulation
was rationalized
by
assuming
that
there s
imperfect
nfor-
mation about
an
incumbent's
ompetence,
with
expansionary
policy
be-
fore
n
election
taken as an
indicatorof
high competence,
as
in
the
pio-
neering
work of
Rogoff
1990)
and
Rogoff
nd Sibert
1988),
and
in
papers
that
followed.
A
partisan
postelectoral ycle
was
argued
to
be consistent
with rational
xpectations
n
the
mportant
work
by
Alesina
(1987, 1988).
On the
empirical
ide therehas been
extensive
work
testing
he
original
and
subsequent
models,
and
more
generally,
ooking
for
empirical
evi-
dence of
political
determinants f
business-cycle activity.
n
his
NBER
Macroeconomics
nnual
paper
in
1988,
Alesina
presented
n excellent
um-
mary
of much of
the work
up
to
thattime.
It is
over
a
decade
since
Alesina's
paper
was
published.
It
now seems
like
a
good
time
to look
at the
past twenty-five ears
of
work
and to
evaluate
the
state
of the literature.What
is
our
current
tate
of
under-
standingof thePBC, boththeoreticallynd empirically? n whatpoints
is there
agreement
nd on
what
points
is there till
significant
isagree-
ment? How well
do the models
explain
the
data?
What
does
existing
theory
s well
as data
suggest
about directions or
future
esearch?
The
short
answer
to
these
questions
is
that we have learned
quite
a
bit,
with
agreement
on
a
number of
ssues,
but
still
ignificant
isagree-
ment on
others. On the
empirical
side,
there are
a
number
of
clear
electoral
effects
n macroeconomic variables.
However,
at
least
for
the
opportunisticmodel in developed countries,there is much less hard
evidence
than both the theoreticalmodels
and the
conventional
wisdom
about
the
prevalence
of
"election-year
conomics"
would
suggest.
Al-
though
there s
wide
(but
not
universal)
agreement
hat
aggregate
eco-
nomic conditions
affect
lection outcomes
in
the United
States,
there
s
significant
isagreement
bout
whether here s
opportunistic
manipula-
tion
that
can
be observed
in
the macro
data.
There
is
a
clear
partisan
effect
n
the United
States
(as
well as
in
some
other
countries),
with
economicactivity einglower
in
the first
art
of
Republican
than
Demo-
cratic
dministrations,
ut
still
disagreement
bout
the
underlying
driv-
ing
mechanisms.
On
the theoretical
ide,
many
of the
leading
models
have
been
criticized
for
implausibility
f
key
assumptions.
Two
key
constant,
uniform,
uiet prosperity
f
the
country
s
by
those
damned
spurts
which
Pitt
used to
have
just
in
the
nick
of
time."
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
4/44
ThePoliticalusiness
ycle
fter
5 Years
?
77
points,
as
I will
discuss
below,
are:
first,
he
assumption
of
seemingly
irrational
ehavior
by
the
public
in
some of the
models; and, second,
the
reliance on monetary urprises s thedrivingforce.
The
purpose
of this
paper
is twofold: first o
present
a short
review
and
critical ssessment of the
existing
iterature,
oth
opportunistic
nd
partisan
models,
the
principal
aim
being
to
point
out
what
we
know
empirically
nd
to
what
extent
existing
models
explain
the
empirical
regularities.
A
principal
conclusion
is that
models based on
manipulat-
ing
the
economy
via
monetary
policy
are
unconvincing
both theoreti-
cally
and
empirically,
hile
explanations
based
on
fiscal
policy
conform
much
better to the data and form a
stronger
basis for
a
convincingtheoreticalmodel of electoral effects n economic outcomes.
Second,
I
present
new
model of
political ycles
based on
Rogoff's
1990)
model of
political
budget cycles,
extended
to include
monetary olicy.
The
model
is
the first o
incorporate
both
monetary
nd fiscal
policy
in a
rational
opportunistic
rameworkwith
separate monetary
nd
fiscal uthorities.4
This
separation
of
monetary
policy
from the direct control of
elected
officials
s crucial for
number of reasons. It is both in
sharp
contrast
o
existing
BC
models
and
far
more
nstitutionally
ealistic
han
the
policy-
makingstructurenthosemodels. Moreover, t s crucialto thenature of
the electoral
ycle,
which
depends
on the
nteraction etween the ncum-
bent
politician
who can
influence
fiscal
policy
and
an
independent
cen-
tralbank
that
controls
monetary
ggregates
and
interest
ates,
but
may
be
pressured
to
accommodate fiscal
shocks. We also
present
some non-
parametric
mpirical
vidence
n
favor
of the
active-fiscal,
assive-monetary
(AFPM)
model of
the
opportunistic
BC.
The
roadmap
for he
paper
is as
follows.
n
the next
section
quickly
review the opportunisticPBC model based on expansionarymonetary
shocks and
present
a
conceptual
assessment. In
Section 3
the
empirical
work on
this
approach
is
summarized. n
Section
4
I1
move on
to
partisan
models
driven
by
monetarypolicy,
both
the
original
Hibbs
model and
Alesina's rational
partisan
model.
In
Section 5 the
empirical
vidence
on
partisan
effects n
macroeconomic
outcomes is
reviewed. In
Section
6,
I
sum
up
what I
consider to
be the
conceptual
and
empirical
problems
with
monetary-based
PBC
models and
present
evidence
in
favor of
a
fiscal-based
model.
In
Section
7
recentwork
on
fiscal
cycles
n
develop-
ing
countries
s
summarized,
both
theoretical
xtensionsof
the
political
budget-cycle
model of
Rogoff
1990),
and
empirical
results
supporting
the
importance
of
fiscal
nfluences n
political
business
cycles
in
a
wide
4.
Rogoff
nd
Sibert
1988)
present
a
model of
fiscal-basedPBC with
nflation
ffects,
ut
where
both tax
and
inflation
olicy
are chosen
by
a
single
authority.
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
5/44
78
*
DRAZEN
range
fcountries.
n
Section
,
twocentral
uestions
elated
o a
fiscal-
based PBC model
re
posed;
the
nswers
resented
motivate he
AFPM
model of Section9, which combines lection-influencediscalpolicy
with
accommodatingmonetary olicy.
n
Section
10,
I
take
a look
at
some
data
for
the United States that are consistent
with the
AFPM
model,
nd
I
present
oncluding
omments.
2.
The
Monetarypportunistic
odel
Beginning
with Nordhaus's
1975)
model,
early
models of
the
PBC,
whether pportunisticrpartisan,were based on monetary olicy s
the
driving
orce.
Expansionary
monetary
olicy
ed
to a
temporary
increase
n
economic
ctivity,
ollowed
with a
lag,
by
an
increase
n
inflation.Models
differed
n
the motivation
f
policymakers,
s well
as
in the
modeling
f
xpectation
ormation,
nd
these
differences
ed to
very
ifferent
ypes
f
politically
nduced conomic
ycles.
Nonetheless,
it is
useful
o review
monetary-based
odels
as a
group
n
assessing
their uccess
n
explaining
PBC.
All
are based
on some
variant f
a
basic three-equationramework,ne equationrepresentinghe pol-
icymaker's
bjective,
ne
giving
he relation
etween
hanges
n the
rate
f
money rowth
r
nflationn
the
one
hand
and economic
ctivity
on
theother
a
Phillips
urve),
nd
finally,
ne
specifying
ow
expecta-
tions
of
inflation
re
formed.
We
begin
with
a
brief
eview
of these
models,
rief
ecause
we
simply
want o
point
ut ome
of
heir heoreti-
cal
shortcomings
nd to
summarize
mpirical
estsof
their
bility
o
explain
olitical
usiness
ycles.
his
review,
ontained
n Sections
, 3,
4,
and
5,
is
based
on
Chapter
of
Drazen
2000a),
where fuller
reat-
ment
may
be found.
2.1 NORDHAUS'S
PPORTUNISTIC
ODEL
Nodhaus's
model
was
meant o
show
that
f
voting
werebased on
eco-
nomic
erformance
n
the
ecent
ast
nd f
xpectations
f
nflation
ere
backward-looking,
n
opportunistic
ncumbent
ho controlled
onetary
policy
would
find
t
optimal
o
nduce n
inflation-unemployment
ycle
corresponding
o
the
length
f
his
term,
with
a boom
ust
before
n
electionnd a recessionfterwards.
The
structure
f
the
economy
s summarized
y
a
nonstochastic,
expectations-adjusted
hillips
urve,
yielding
n
inflation-output
rade-
off.
=
eTM-
1T,
(1)
t
'Tt
t
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
6/44
The olitical
usiness
ycle fter
5 Years
?
79
where
xt
s thedeviation
f actualfrom
otential
utput
nd where
he
monetary
uthority
s assumed
o
control he nflationate
ri.5
Theobjective f thepolicymakers tomaximize isprobabilityfre-
election.
oting
ehavior s
retrospective,
n that t
depends
n
economic
performance
nder
he ncumbent
n
the
past.
Economic
erformance
n
a
period
s measured
y
the
behavior f nflation
nd
unemployment,
o
thatvoter
dissatisfaction
n
any
period
can
be
represented y
a
loss
functionfthe
form
(x,
)2
(t
-
2
~2,
=
C
+
2(2)
2 2
where
r
s the
electorate's
arget
ateof
nflation,
is the
target
ate f
economic
ctivity
relative
o
potential utput),
nd
a
is
the
relative
weight
he electorate
uts
on
output
fluctuationselative
o inflation
fluctuations.
n
opportunisticolicymaker
ill
choose
the
policy
hat
attractsmost
voters,
o
that
hese
parameters
ould be
thought
f as
representing
he
preferences
fthemedian
voter.
In
thebasicmodel, ne thenposits retrospectiveoting unctionoranelection t theend of
period
,
oftheform:
N,
N
8sSts
+Et,
(2a)
s=O
yielding
he number f
votes
Nt
as
a
function f voters's
well-being,
where
N'(.)
<
0. The
exogenous ength
f ime
etween lections
s
T
+
1
periods, < 8< 1is thefactor ithwhich oters iscount ast conomic
performance
a
"forgetfulness
oefficient"),
nd
Et
s
a
mean-zero
tochas-
tic erm
elating
conomic
erformance
o electoral utcomes.
he elec-
toral
mechanisms
not
made
more
pecific.
he standard
pportunistic
PBC
model
assumes
that
8
is
small,
n
the
sense
that
ecent
conomic
performance
ountsfarmore
heavily
n
influencing
oter
hoices han
economic
erformance
n
the
moredistant
ast.
The
stochasticlement
is
added to allow for
he
possibility
f n
incumbent
osing
heelection.
To close the model one must pecifyheformationfexpectations.
5.
In
order
o
reproduce
he
regularity
f
high
nflation
agging
he
monetaryxpansion,
one must
ecouple
money rowth
nd nflation.
simple
ssumption
long
hese
ines
is that
nflationeflects
oney rowth
n
the
previous eriod,
hat
s,
r,
=
-U,_,,
with
,
being
he
monetaryuthority's
ontrol
ariable,
nd with
he
divergence
f
ctual rom
potential
utput
epending
n
the
differenceetween he
ctual ate f
money rowth
and the
economy-wide
xpected
ate f
money rowth
A.
See
Chapter
.3 of
Drazen
(2000a)
for
recise
etails.
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
7/44
80
DRAZEN
Crucial to
the main
results
f the Nordhausmodel is some
form
f
adaptive
xpectations.
standard
ormulation
f
adaptive
nflationx-
pectationss:
IT=-_1
+
"+0(,t-1
"/'-t-Ir),
(3)
where 0
is a
coefficient
etween
0 and
1
representing
he
speed
with
which
xpectationsdapt
to
past
nflation. hat s
crucial
n
theforma-
tion of
expectations
s that
wr
oes
not
depend
on the
expectation
f
future
olicies,
o
that
xpectations
renotrational.
t s this
haracteris-
tic
combined
with he absence
of
any
other onnections
etween
eri-
ods)
which
gives
the incumbent
olicymaker
n
exploitable
rade-off
between
nflation
nd
unemployment
n
the
attempt
o affect
lection
outcomes.
Voter ehavior
n
the
Nordhausmodel
s
backward-looking
n
two
dimensions:
otingdepends
on
past
incumbent
erformance,
nd ex-
pectations
f
moneygrowth epend
only
on
past
inflation
ates.
The
incumbent
policymaker
lected
at
t-3
chooses
inflation
ates
r7Tt_,
t-2,
Srt_1,
and
irt
o
maximize
is
expected
vote
in the nextelection.
This
simple tructureields hefollowingehavior f ncumbents howish
to
maximize
he
probability
f
remaining
n
office.
mmediatelyreced-
ing
an
election
he
government
timulates he
economy
ia
expansion-
ary
monetary olicy.
The levels
of
monetary xpansion
nd
economic
activity
re those
that
maximize oter atisfaction
n
an
election
eriod
taken
lone.
In the
period
mmediately
fter he
election,
he
govern-
ment
everses ourse.
t
engineers
recession
ia
contractionary
one-
tarypolicy
to
bring
down
inflationary
xpectations.
he
incumbent
keeps economic ctivityow to keep expected nflationow untilthe
period
immediately
efore
he next
election,
o thata
given
rate
of
economic
xpansion
induced
by
a
monetary
urprise)
an be obtained
at
a
relatively
ow
rateof
nflation.
n
thenext
lection
ycle,
he
same
behavior
s
repeated.
Hence,
we
have
a
simple
xample
n which
the
possibility
f
nfluencing
he
probability
f
re-election,
ombined
with
the structure
f
the
economy,
ields
cycle
n economic
ctivity.
The
exact
olution
may
be
found
n
any
treatment
f the
Nordhaus
model,
for
xample,
razen
2000a, . 233-236).]
2.2 CONCEPTUAL
RITIQUE
There
are
three
general
conceptual
riticisms
f the
basic
Nordhaus
model
s
a
tool
for
xplaining
PBC.
First,
t
assumes
hat he
president
controls
monetary
olicy,
n
assumption
hat s inconsistent
ith
the
independence
f
the
Federal
Reserve.
Although
ome
observers
rgue
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
8/44
The olitical usiness
ycle
fter
5Years
?
81
that
decisions
on
monetary olicy
n
the United States are
strongly
nflu-
enced
by
the executive
branch,
the notion
that the
president
can
easily
use monetary policy as an electoral tool does not fit the institutional
facts.
A
more subtle
argument
s that
an
independent
Federal Reserve
may
be
especially willing
to
accommodate the executive branch's
pres-
sures
for
monetary olicy during
election
years
n
orderto
prevent harp
movements
n
interest ates
which would lead the Fed to be criticized.
We
return o this
argument
below.
A
second,
more serious
problem
with
the
Nordhaus
model is its reli-
ance on irrational
ehavior on the
part
of
voters. Voters are
naive,
not
simply
n
the
way they
form
xpectations
f
nflation,
ut also
in
the
way
they
assess
government erformance.Any
voter
who
has
lived
through
an
election
cycle
in
Nordhaus's
world
should not be fooled into
voting
for
an
opportunistic,
manipulative policymaker.
He
will
know that
the
pre-election
period
of
low
inflation
nd
high
economic
activity
will
be
followed
by
a
postelection
period
of
both
high
inflation nd
high
unem-
ployment.
He should
therefore
unish
rather han reward an
incumbent
who
engages
in
pre-electoral
manipulation.
Finally,
and more
generally,
one
may
question
the central role
as-
signed tomoving along thePhillipscurveto reduce unemployment ia
inflation
urprises.
Fiscal
policy plays
no role
in
the
PBC
in
the
model,
though
transfers
nd
other
ypes
offiscal
policy
appear
to
play
an
impor-
tant
role
in
some
episodes
of
pre-electoral
olicy
manipulation.
3.
Empirical
ests
f
he
Nordhaus
odel
There have
been
many
econometric ests of the
monetary pportunistic
PBC, bothfor conomicoutcomes and forpolicy nstruments. he most
common form f
econometric
est of
these models
in
terms
of
outcomes
is
to
run an
autoregression
of an
economic
performance
measure
on
itself,
small
set of economic
variables,
and
political
dummies to
test a
specific
heory.
Consider a
regression
of
the form:
Yt
=
aY,
+
bo
+
1bIX
+
dPDUMt
+
Et,
(4)
i=1
j
where
Y
is
an
outcome
variable
such as
GDP,
the
Xj
are other
economic
variables that
may
also
affect
Y,
such
as world
economic
activity,
nd
PDUM is a
political
dummy
variable
or
set
of
variables)
meant
to
repre-
sent a
given
political
model.
The
autoregressive pecification
or
Yt
is
adopted
as a
parsimonious
representation
f
the
time-series
ehavior
of
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
9/44
82"
DRAZEN
Y,,
nstead
of
using
a structural odel. For
example,
s
a
test
ofthe
Nordhausmodelon
quarterly
ata,
Alesina
nd Roubini
1992),
Alesina,
Cohen, ndRoubini1992), ndAlesina,Roubini,nd Cohen 1997)use
a
dummy
ariable hat
quals
1
in
the
election
uarter
nd in the
T-1
quarters
efore
he
lection,
nd 0
otherwise,
here
T
may
qual
4, 6,
or
8. As
the measure
f economic
ctivity they
ake the
year-over-year
growth
ateof GNP or
an
unemployment
easure,
he exact
pecifica-
tion
depending
n the
model nd data set.
3.1 THE
EFFECT
F
ECONOMIC ONDITIONS N ELECTIONS
Prior
o
discussing
he ffect
f lections n macroeconomic
ariables,
ne
must onsider he effect feconomic onditions n elections.A crucial
assumption
n
the Nordhaus
model,
or
in
any
model
of
pre-electoral
manipulation,
s that
voters
vote on
the basis of
economic
variables.
Kramer
1971)
regressed
otesreceived
y
the
ncumbent
arty
n
U.S.
congressional
lections
n
two
measures
f
performance
n
the
year
of
the election-the
growth
ate
of
real
per
capita
ncome
nd the rate
of
inflation
n
that
year-and
found
hey
were both
significant
etermi-
nants
fvote otals.
he
mportance
f
conomic onditions
or
oting
n
congressionalelectionswas confirmed yTufte 1975).6
The
most
nfluential ork
was
probably
hat fFair
1978)
updated
n
Fiar
1982,
988)],
who found imilar
esults or
he
United tates.
n
his
original
rticle,
air
ooked
at
presidential
lections
rom 916
through
1976,
arguing
hat
f
voters
hold
the
party
hat
holds
the
presidency
accountable
or conomic
vents,
heir nfluence hould
be seen most
strongly
n
presidential
lections.
Fair
found
that
the
change
n
real
economic
ctivity
n
the
year
of the
election,
s measured ither
y
the
change n realpercapitaGNP or thechange nunemploymentn the
election
ear,
oes
appear
o have an
important
ffectn votes
for
resi-
dent.
Specifically,
1%
increase
n
the
growth
ate
ncreases
he ncum-
bent'svote total
y
about
1%.
(Further
vidence
uggests
t
may
be the
growth
f
real
per
capita
GNP
in
the second
and third
uarters
f the
election
year
that s
important,
ut data limitations
revent
air from
drawing
ny
definitiveonclusions
bout
what
part
f
the
election
ear
is most
mportant
n
determining
oter
ehavior.)
Given
the
growth
f
economic ctivity,thermeasures f macroeconomicerformanceon-
tribute
ittle;
he
most
mportant
f the othermeasures
s
the nflation
rate
n
the
two-year
eriod
before he
election,
s
measured
by
the
change
n
the
GNP
deflator. second
keyfinding
f
Fair's
s that oters
6.
Though
most
studies
confirm
he
basic
results,
Stigler
1973)
concluded
that
congres-
sional
election
results
are not affected
y
economic
fluctuations.
ee
also
Okun's
(1973)'
comment
on
Stigler,
s well as
Arcelus and
Meltzer
1975)
and Bloom
and
Price
1975).
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
10/44
The olitical
usiness
yclefter
5
Years
?
83
appear
to have
a
high
discount
ate n
past
economic
erformance;
hey
don't ookback
more han
year
r
two.7
Numerous ther rticles ind imilar esults nthe mportancefpre-
election onditions n
voting
patterns
n both
the UnitedStates
and
other
ountries.
ooking
t
voting
r
popularity
unctions,
ewis-Beck
(1988)
found hat
he
sort f results
hat
Kramer
nd
Fair
report
or he
United tateshold
n
Britain,
rance,
West
Germany,taly,
nd
Spain
as
well. Madsen
1980)
reported
imilar esults or
Denmark,
orway,
nd
Sweden.8We summarize his
s:
REGULARITY Aggregateconomic onditionseforen election,pecifically
percapita utput
r income
rowth
and
to a lesser xtent
nflation),
ave
significant
ffect
n
voting atterns
n theUnited tates
nd
other ountries.
3.2
ECONOMICACTIVITY
Numerous conometric
ests
rovide
ittle
upport
or he
political
ycle
in
economic
ctivity redicted
y
the
Nordhausmodel.Studies
for
he
United
States
began
with
McCallum's
1978)
study
of
unemployment
fluctuationsefore lections.Altand Chrystal1983)summarizearly
empirical
tudies as
showing striking
ack of
support,
point
re-
inforced
y
results ummarized
n
Alesina,
Roubini,
nd Cohen
1997).
Faustand Irons
1999),
using
more
ophisticatedechniques,
ome
to
a
similar
onclusion.
igure
,
showing
meanrates
fGNP
growth
season-
ally adjusted)by quarter
f
the
president's
erm
n
the United tates
from
948
o
1998,
llustrates
he
point.9
Similarly,
o evidencewas found n
developed
conomies utside he
United tates or Nordhaus-styleBC for nemploymentreconomic
growth
Paldam,
1979;Lewis-Beck,
988).
Alesina,
Roubini,
nd Cohen
7.
One should
distinguish
aggregate
from ndividual
economic conditions on
voting.
Lewis-Beck
1988)
argues
that
ndividuals
vote on the basis of national
economic
perfor-
mance
(sociotropicvoting)
rather han
their
own
personal
economic situation
"narrow
pocketbook"
voting).
8. What
about
the effect f
economic conditions on the
timing
of elections when
govern-
ments
can
call
early
elections?
to
(1990)
finds evidence that
governments
n
Japan
do
not
manipulate
policies
in
anticipation
f
upcoming
elections,
but
that
hey pportunisti-
cally manipulate
the
timing
of
elections to
take
advantage
of
autonomous
economic
expansions.
Specifically, igh growth
ignificantly
ncreasesthe
probability
f an elec-
tion,
while
high
inflation
significantly
educes
it.
Chowdhury
(1993)
reports
similar
resultsfor
ndia,
with
the
government
more
ikely
o
call
early
elections when economic
times
are
good.
On the
other
hand,
Alesina, Cohen,
and
Roubini
1993),
argue
that
for
sample
of 14
OECD
countrieswith
endogenous
election
timing,
here
s
no
evidence
of
such an
effect
n
countriesotherthan
Japan.
9. A
plot
of
median
growth
rates,
or of
other measures of
aggregate
economic
activity,
or
the United
States would tell a similar
tory.
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
11/44
84 DRAZEN
Figure
MEAN GNP GROWTH
RATE,
948-1998
10
8
S6
0
a.
4--
2
0
0 1
2
3
4
5 6
7
8
9 10
11
12
13
14
15
Quarter
fter
aking
ffice
(1997)
reject
n
opportunistic
ycle
n
real
activity
or
sample
of 18
OECD
countries
ver
he
period
1960-1993.10
Wesummarizehegeneral onsensus hat heopportunisticBC re-
ceives ittle
upport
n
the
pre-electoral
ehavior
f GNP
or
unemploy-
ment s:
REGULARITY
Theres no
significant
ncrease
n
aggregate
conomic
ctivity
prior
o lectionsn
either
he
United
tates
r
other ECD
countries.
3.3
INFLATION
Thepostelectoralncreasen nflationredictedythe
Nordhaus
model
receives
upport
n
some countries
nd
not
n
others.
Alesina,
Cohen,
and Roubini
1992)
and
Alesina,
Roubini,
nd Cohen
(1997)
test
for
political
ycle
n inflation
measured
s
the
growth
ateof theCPI
over
the
previous
quarters), sing
he
ame
data set and
methodology
hey
used
for
GNP
growth,
nd
defining political
dummy equal
to
1 in
the
election
uarter
nd
in
the3
quarters
ollowing
he
election,
nd 0 other-
wise.
n a
pooled
cross-section,
ime-series
egression,
hey
ind
highly
10. If aggregateeconomicperformances importantn determining heway people vote
and
governments
want to
win
re-election,
why
don't
we
observe
a
clear
opportunistic
PBC?
Lewis-Beck
1988)
argues
that t
s because it
s
exceedingly
hard to time
conomic
manipulation.
Monetary
and fiscal
policy
can
be used
only
with
great mprecision,
o
that
politicians
cannot
expect
to
time
the
aggregate
stimulus
to come
right
before
an
election,
while the
risks ssociated
with a
mistimed
xpansion
are
high.
Another
xpla-
nation
s that
opportunistic
oliticians arget
ransfers
o a fraction
f
voters
with minor
effect n
aggregate
economic
activity.
he AFPM model
in Section
9
includes
both of
these
possibilities.
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
12/44
The olitical usiness
ycle
fter
5 Years
?
85
Figure
MEAN
NFLATION
RATE
CPI),
1960-1978
8
6-
4-
2
0-
0
1 2
3 4 5 6 7 8 9 10
11 12 13 14 15
Quarter
fter
aking
ffice
significant
oefficient f
the correct
sign
on the
political
dummy;
in
the individual
country
regressions, they
find
the
coefficient
s
of the
correct
ign
in
almost
all
the
regressions,
nd
significant
t the
10% or
higherlevel forDenmark, France,
Germany,
taly,
nd New Zealand.
Overall,
they
conclude the
PBC
effect n
inflation s
widespread
across
OECD
countries
on
the
basis of their
pooled regression)
nd
on
a
much
stronger mpirical footing
han
the
effect
n
GNP
and
unemployment.
The evidence for
he United States
s
less
clear.
n
similar
ests
to those
described
above,
Alesina, Roubini,
and
Cohen
(1997)
reject
he existence
of a
postelectoral
urge
n
inflation ver
the
period
1947-1994.
However,
the behavior of
nflation fter lections
changed
over
this
ample period.
After 979there s no evidence of a political nflationycle,whichcorre-
sponds
to
the
timing
of
the
change
in
Federal
Reserve
policy
rules
in
1979.
(See,
for
xample,
the estimated
policy
rules
in
Clarida, Gali,
and
Gertler,
000.)
Prior o this
however,
there
s more
evidence
of
a
possible
postelectoral
ncrease
in
inflation.
his
is
consistent
with
other
studies,
and
is
illustrated
n
Figures
2 and
3,
showing
mean
annualized CPI
inflation
seasonally
adjusted)
from
1960
to
1979 vs.
1979 to
1998
by
quarter
of the
president's
term.
A
graph
for1948-1979
looks
very
simi-
larto 1960-1979,but thelatter s used forbetter omparabilitywith ater
figures.)
To
summarize:
REGULARITY
In
many
OECD
countriesheres
a clear
postelectoral
ncrease
in
inflation.
n
theUnited
tates,
here
s
evidence
f
uch
postelectoral
ncrease
in
inflation
rior
o
1979,
butno
evidence
hereafter.
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
13/44
86
-
DRAZEN
Figure
MEAN NFLATION
ATE
CPI),
979-1998
7
6
cc5
4
2
1
0
0
1
2 3 4
5 6 7 8 9
10 11 12 13
14
15
Quarter fter akingoffice
3.4 MONETARYNSTRUMENTS
Not
surprisingly,
he
results
or
xpansionarymonetary
olicy
before
elections
mirrorhose
for
nflationfter
lections.
sing
he ame
politi-
cal
dummy
hey
did for
nflation,
lesina,Cohen,
and
Roubini
1992)
find
significant
olitical
ffector he
yearly
M1
growth
ates
n
pooled
cross-section,ime-seriesegressionsntheirample fOECD countries,
with
money
rowth
einghigher
or he
year
o
year-and-a-half
efore
elections.
n
the
country
egressions,
he
results
re
ess
strong, hough
a
num'ber
f
countries
isplay
ignificant
ffects.
For heUnited
tates,
he
ensitivity
f he
nflationesults
o the ime
period
onsidered
s seen
n
money rowth
ates
s
well.
Alesina, ohen,
and Roubini
1992)
find
nlyvery
weak
evidence f
politicalmonetary
cycle
n the
postwar eriod,
conclusion einforced
n
Alesina,
Roubini,
andCohen 1997)for heperiod 949-1994.ncontrast, rier1989) nd
Beck
1987)
oth
find
ignificant
upport
or n
office-motivatedodel
f
monetaryolicy
n
the
United tates ver he
ubperiod
960-1980.
rier,
using
U.S.
quarterly
ata from
961
o
1982,
egresses
M1
growth
n its
previous
alue,
he
full-employment
eficit,
nd
a
political ummy
peci-
fied s a
fifteen-quarter
econd-degree
olynomial
istributed
ag
on
a
dummy
which akes value of
one
n
the
lection
uarter
nd zeroother-
wise.
The
polynomial
istributed
ag
s chosen
o
conserve
n
degrees
f
freedom.) e finds hat hetimingf n electionignificantlynfluences
money rowth,
ven
whenfluctuations
n
output,
nterest
ates,
nd the
deficit
re held
constant. eck
1987)
also
finds
political
ycle
n
the
money upply
n the
United
tates ver
he ame
period.
Figures
and
5
present
mean
M1
growth
ates
seasonally
djusted)by
quarter
f the
president's
erm ver he
periods
960-1979
nd 1979-1998.
nterestingly,
Beck
finds o similar
ycle
n
monetary
nstruments,
uch
as reserves r
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
14/44
The
olitical
usiness
ycle
fter
5Years
.
87
Figure
MEAN
M1
GROWTH
RATE,
960-1978
8
(f6-
0
4
X.
2
0
I
0
1 2 3
4
5
6
7 8
9 10 11 12 13 14 15
Quarter
fter
aking
ffice
Figure
MEAN
M1
GROWTH
RATE,
979-1998
12
10
8-
0o 6
4-
2
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Quarter
fter
aking
ffice
the
federal funds
rate,
a
point
made clear
in
Figure
6,
giving
the mean
federal
fundsrate
by quarter
of
termfrom 959 to
1998. The difference
n
results forthe behavior of
money
growth
and
instruments
f
monetary
controlwill be centralto our model of the PBC presentedbelow. We
summarize
these
results
as:
REGULARITY
Theres evidence
f
pre-electoral
ncrease
n
money
rowth
ratesn
many
ountries.
n
theUnited
tates,
heres
a
pre-electoralffectrom
1960
o
1980,
ut
one
hereafter.
heres
no
vidence
or
heUnited
tates
f
n
electoral
ycle
n the
ederal
unds
ate.
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
15/44
88
-
DRAZEN
Figure
MEAN FEDERAL
FUNDS
RATE,
959-1998
8
7
6
c5
'*
4
C3
2
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Quarter
fter
aking
ffice
4.
Monetary
artisan
odels
The
basic
partisan
model startswith the
observation hat
right-wing
nd
left-wing arties
have different
ositions
on
economic issues
and
hence
differentmacroeconomic
objectives.
In
terms of the
objective
function
(2), theyhave differentreferences ver inflation nd unemployment,
both
in
inflation
nd
unemployment
argets
nd the
relative dislike
of
inflation s.
unemployment.
4.1 THE
BASIC
HIBBSMODEL
The
partisan
PBC
model
was
introduced
by
Hibbs
(1977).
To
represent
the
difference
n
interests,
we
replace
the
social loss function
2)
by
a
partisan
oss
function:
(xt
-
Xi()2
+
(
-r
T-
)2
'
=
aI
+
(5)
2 2
for
party
,
where
i1r
s
party
's
target
ate
of
nflation,
j
is
party
's target
for conomic
activity,
nd
d
is the
relative
weight
put
on
output
fluctua-
tions relativeto
inflation luctuations
y
party
.
There are two
parties,
left-wing
arty,
enoted
L,
and a
right-wing arty,
denoted
R. The
two
parties are characterizedby the following possible differencesn their
objectives.
First,
the
left-wing
arty
may
have a
higher target
for eco-
nomic
activity
han the
right-wing arty.
Second,
the
left-wing
arty
may
assign
a
larger
ost
to
deviations
of
economic
activity
rom ts
target
level
than
to deviations
of nflation
rom he
target.
Finally,
he
eft-wing
partymay
have
a
higher
nflation
arget
han the
right-wing arty,
nde-
pendent
f
the effects
n
economic
activity
ia
the
Phillips
curve,
which
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
16/44
The
olitical
usiness
ycle
fter
5 Years
?
89
could reflect ther
effects f
nflation iewed
differently
y
the two
par-
ties.
To summarize
the difference etween the
parties:
5-L tR
aL
>- aR
(6)
To obtain the
partisan cycles,
at least one
of
these
must hold
with
strict
inequality.
Fluctuations
n
economic
activity
nduced
by
these
partisan
differ-
ences are
generated
in
the
basic Hibbs
model
by
movements
along
an
exploitable
Phillips
curve,
where it s
assumed,
as inthebasic Nordhaus
model,
that
expectations
are not rational.
Thus,
the
left-wing arty
will
pursue
a
more
expansionary
monetary policy throughout
ts
term.n
How
long
these effects ast
depend
on
the exact
specification
f
expecta-
tions.
In
an
adaptive expectations
framework,
he more
slowly
nflation
expectations
adjust
to
actual
inflation,
he
longer
will be
the
partisan
effect.
A
basic criticism f the
original
Hibbs
model
is
the same as the one
that was applied to the Nordhaus model, namelythat t relieson mis-
taken
expectations
of
what
policy
will
be
in
order
to
get
real
effects.
Hence,
to the extent hat
t
s assumed that
monetary olicy
s
used to
hit
partisan
unemployment
nd
growth
argets,
he
explanation
fthe
politi-
cal
business
cycle
s
unsatisfactory.
4.2
ALESINA'S
RATIONAL-PARTISAN ODEL
Alesina
(1987, 1988)
introduced
rational
expectations
nto
a
monetary-
based PBC, influenced ythecriticismf models based on an exploitable
Phillips
curve.
In
his
partisan
model with
rational
expectations,
only
surprise
inflation
ffects
utput,
leading
to
Alesina's
terming
he
ap-
proach
the
rational-partisan
odel. The
rational-partisan
model can
be
represented
by
a
similar
three-equation
model to that
used
by
Nord-
haus,
retaining
he
expectations-augmented
hillips
curve
1)
but
chang-
ing
the
other
two
components.
First,
following
Hibbs,
the
motivation
f
policymakers
s
quite
different han in
the
Nordhaus
model:
they
are
purely partisan,
with
no
opportunistic
motives
and
hence
no desire to
manipulate
outcomes. To
represent
the
difference
etween
economic
effects
n
the
early part
and
the latter
part
of
an
incumbent's term
of
office,
Alesina
divides a
term f
office nto two
periods
and
assumes that
11.
As
in
the
Nordhaus
model,
the
key
assumption
here
is
that,
n
spite
of the
Federal
Reserve's formal
autonomy
n
the United
States,
monetary
olicy
reflects he
adminis-
tration's
macroeconomic
goals.
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
17/44
90
DRAZEN
there
is an
election
every
other
period, say
at
t,
t+2,
t+4,
.
. . .
It
is
assumed
hat
party
ares
only
bout tsown term f
office,
o
that
he
objective unctionfparty at timetmaythenbe representedyan
extended
ersion
f
5),
namely
A
(xt
-
2j)2
+(Vt.-
jCj)2
Al
aI
+
(7)
2 2
+
ai
xt+l
+
_tj)2
(t-
1
)2
for
party ,
where
?7
nd
?j
are
the
partisan argets,
i
is
the
relative
weight ut
on
output
eviations
y
party
,
and
P
is the
discount
actor.
These
are
characterized,
s
in
theHibbs
model,
by
6)
above,
where,
n
order
o obtain
he
cycles
n
the
rational-partisan
odel,
t leastone of
the
nequalities
n
6)
must e strict.
The other rucial
hange,
relative o
both the Nordhaus
nd
Hibbs
models,
s
thatAlesina
eplaces
he
ssumption
f
daptive
xpectations
byrationalxpectations,o nstead f 3),expectednflation7r sgiven y
rt
=
Etl(Tt).
(8)
In
determining
he
evolution
f
nflation
nd
unemployment
uring
term
f
office,
ay
t
and
t+ 1,
the
key
variable
n
the
model s
expected
inflation
n those
periods,
his
xpectation
eing
formed
efore
he
lec-
tion
n
period
.
Conditional
n
expected
nflation
n each
half
erm,
he
partynpowerchooses tsoptimal olicy, ymaximizing7) subject o
(1).
We retain
he
assumption
rom
arliermodels
hat
he
government
has
perfect
ontrol
ver nflation.
n
turn,
xpectations
f
nflation
e-
pend
on
the
expectation
fwho
willwin the
upcoming
lection.
fout-
comes
were
fully
nown,
herewouldbe
no
cycle,
ince
party's
olicy
wouldbe
fully
nticipated
nd hence
have no
effect
n real
activity.
The existence
f
a
cycle
hus
depends
on
uncertainty
bout
election
outcomes.
xpected
nflation
or he half
erm fter
he elections
the
weighted
um
of
the
two
parties'policies,weighted y
the
probabilitythat achwillwin theelection,
amely,
7;
=
qLTr
+
(1
-
qL)rt,
(9)
where
L
s the
probability
hat
he
eft-wingarty
will
win the
election,
and
where
IrL
and
irr'
are the
optimal
policies
of the
two
parties
n
the
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
18/44
The Political usiness
ycle
fter
5 Years
?
91
first
alf
fthe
term,
which
depend
not
only
n
their
olicy references
(6),
but
also on
the
election
robability
L
itself,
s
optimal olicy
de-
pendson7rr.ince he eft-wingarty ollows morenflationaryolicy
once
in
office
han
the
right-wingarty,
xpected
nflations
between
these wo
values.
Hence,
there
s a
positive
nflation
urprise
f
he
eft-
wingparty
wins the
election,
mplying nemployment
elow the
natu-
ral
rate,
nd a
negative
nflation
urprise
f
he
eft-wing
arty
wins the
election,
mplying nemployment
bove thenatural ate.
n the
econd
half fa
president's
erm,
here
re
no
fluctuations
n
economic
ctivity,
as
the
dentity
f he
party
n
power
s knownwhencontracts
re
signed
(in
the first
art
of the
term).
n
contrast,
ibbs's
partisan
model
sug-
gests
higher
conomic
ctivity
n
eft-wing
dministrationshan n
right-
wing
dministrationsver he
ife fthe
erm.12
4.3
A
CONCEPTUAL
SSESSMENT F
THE
RATIONAL-PARTISANODEL
The
theoretical
tructure
f he
ational-partisan
odel aises number f
questions
bout
the
underlying
riving
orces.
irst,
nd
most
difficult,
theres
the
uestion
f
whetherhe
underlying
icroeconomic
tructure,
namely ominalwagecontractsigned efore lections,makes ense n
the
context f the
model. The
question
of
microfoundationss often
raised about
models
in
which
policymakersxploit
n
expectations-
augmented
hillips
urve,
utthe
mportance
f
lectoral ffects
ives
t
special
mportance
ere.
A
standard
rgument,
sed also
by
Alesina,
s
that
ominal
wage
contractsre
igned
t
discrete
ntervals,
here omi-
nal
wage
increases eflect
ationallynticipated
nflationt
thetime
he
contracts
signed,
o that
urprise
nflation
etween ontract
ates can
have realeffectsvenwhenagents rerational. he basicproblem,s
Rogoff
1988)
oints
ut,
s
that,
n theone
hand,
lectionsre
an
mpor-
tant
ourceof
fluctuations
ue
to
their utcomes
eing
ess than
fully
anticipated,
ut,
on the
other,
he
lection ate s
fully
nown.
The
mag-
nitude f
he
hanges
n
nflation
nd
unemployment
he
model s
meant
to
explain
re
sufficientlyarge
hat
here houldbe
a
large
tility ayoff
to
eliminating
he
uncertainty
hat
eads to these
luctuations.ut hat
s
easy
to do. To
theextent
here
s
a
significant
ffect
n
unemployment,
12. Hibbs (1994)
presents
uch a
theory
f
adjustment
of
partisan
objectives
contingent
n
economic
outcomes
and
learning,
which
predicts
that
unemployment
nd
inflation
outcomes across the
two
parties
may
diverge
more
in
the first
art
of their erms
han
in the
second,
though
not
because
of
uncertainty
bout
electoral
outcomes. The
key
to
Hibbs's model of
changing
objectives
and
to the result
on
time-varying
utcomes)
is
that
policymakers
re
uncertain about the
structure f
the
economy
and
the effects f
policies.
They
use
outcomes
to
refine heir eliefs about
attainable
argets,
eading
to a
feedback
from
utcomes to
partisan objectives
and
thus
policies.
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
19/44
92
-
DRAZEN
old
contracts
hould
be timed to
expire
and
the
signing
of
new
contracts
postponed
until
ust
after n
election,
o that
they
an
reflect he
election
results.Hence, themain drivingforceofthe model would seem to de-
pend
on behavior of
workers
nd
unions that
s less
than
rational,
not
in
the formation ftheir
xpectations er
se,
but n their
abor-supply
ehav-
ior.
A
simple
change
in
the
timing
f contract ehavior would eliminate
the
political ycle.
Garfinkel
nd
Glazer
(1994)
present
mpirical
vidence
that for abor contracts
f
less
than
two
years
signed
in a
presidential
election
year,
there s a clear
tendency
to
delay
the
signing
of
abor con-
tracts ntil after he election.13
A
second crucial
question
concerns the electoral
uncertainty
hat
drives the model. The
magnitude
of the
cycle
depends
on the
degree
of
electoral
uncertainty,
s
well
as
on the
difference
n
the
parties'
desired
inflation ates.
One
problem
s that
these
key
driving
forces
re
exoge-
nous.
Far more troublesome is the
predicted positive
correlationbe-
tween the
extent of
the
electoral
surprise
and the
size of
postelectoral
movements
in
real economic
activity.
As the
key probability
qL ap-
proaches
zero or
one,
the
magnitude
of the fluctuations
will
approach
zero,
with fluctuations
eing
maximal
(all
else
equal)
for
q'
=
1.
Hibbs
(1992), among others,has argued that thisprediction s not consistent
with
the
empirical
evidence for he United States. Consideration
of
ndi-
vidual elections reveals
the
problem.
For
example,
the outcome
of
the
1964
presidential
lection
s
probably
the closest
we have seen to
a
sure
thing
n the
postwar
era,
with
Lyndon
Johnson's
victory
widely
antici-
pated.
Yet the rate
of
real
GNP
growth
n the
first wo
years
of the
Johnson
administration
veraged
5.8%
per
year,
the
highest figure
of
any
Democratic
administration.
n
contrast,
mong postwar
Republican
victories hroughRegan's first lection,Nixon's victoryn 1968was the
closest
and least
certain,
but
corresponds
to the smallest
drop
in
real
output
n the critical
econd
year
of
the administration.
Alesina,
Roubini,
and Cohen
(1997,
Chapter
5)
construct
n
index
of
electoral
surprise
for
the
U.S.
presidential
elections
from
1948 to
1992,
with
Republican
victories
ntering
s
negative surprises.
They
use differ-
ent
variants
s an
explanatory
ariable in
a
real-GDP-growth
egression
of the
form
4)
and
find
that the coefficient
n the
surprise
variable
is
significantlyositive,meaning
that
arger
Democratic
Republican)
sur-
prises
imply
higher
lower)
postelection
eal
growth
rates.The construc-
13. Garfinkel
nd Glazer's results
may
be
interpreted
n two
ways.
One
is that
postpone-
ment
of contract
igning
indicates that electoral
uncertainty
s
important
n
forming
inflation
xpectations,
onsistent
with
the basic
thrust f
the
rational-partisan
model.
The
other s
that
n industrieswhere
this s
true,
ontract
igning
s
postponed,
under-
cutting
he
empirical
relevance
of
the main
driving
force
f the model.
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
20/44
The olitical usiness
ycle
fter
5 Years
93
tion of the variable is
complicated,
so that
t
is not
easy
to
see
why
the
results of the
regression
nd
of
the
simple
case
study
do not
agree.
The
relation ofpre-electoral ncertaintynd postelectoralfluctuations s an
important uestion
deserving
further esearch.
A final
question,
which can be
applied
to
all
the models discussed
so
far,
s
the central
role
assigned
to
moving along
the
Phillips
curve
to
reduce
unemployment
via inflation
urprises.
That
is,
even
though
real
effects f
monetary
policy
are
consistent
n this
approach
with
rational
expectations,
the
reliance on
monetary policy
as
the
driving
force
of
cycles
is
inconsistent
with
the
evidence
on the
important
ole of
fiscal
policy
n
PBCs.
We
return
o this
point
n
Section
6.
5.
Empirical
ests
f
artisan odels
The
partisan
PBC
has been
tested less
than
the
opportunistic
model.
There is
general
agreement
on the
existence of
partisan
effects
er
se,
especially
on
economic
activity.
owever,
there
s far
ess
consensus on
the
mechanism at
work.
5.1 ECONOMIC ACTIVITY
Perhaps
the
strongest
egularity
n
the U.S. data
was
first
ointed
out
by
Alesina
(1988),
with
Faust and Irons
(1999)
confirming
he
effect ver
a
longer
time
period
using
more
sophisticated
econometric
techniques:
For
the
United
States,
real
GDP
growth
s
substantially
higher
under
Democrats than
Republicans
in
years
2
and
3
of their
administrations.
Alesina,
Roubini,
and
Cohen
(1997)
report
hat
over the
period
from
he
first
uarter
of
1949
through
the second
quarter
of
1994,
growth
rates
duringDemocraticand during Republican administrationssharplydi-
verge
starting
bout
the third
quarter
after
he
election.
The
quarterly
growth
rate
averaged
over
Democratic
administrations
ises from
bout
3%
per
annum in
quarter
3 to about
6%
per
annum
by quarter
6 or
7 in
the
administration's
erm
of
office,
nd
falls from
he same
level
to
zero
by quarter
6
or
7
in
the
administration's
erm
averaged
over
Republican
administrations.
eal GDP
growth
ates
then
mprove
under
Republican
and
worsen
under
Democratic
administrations,
o that n
the
fourth
ear
of
the
administration,
he
growthperformance
nder
the
two
parties
s
identical.
Unemployment
hows
analogous
partisan
patterns
n
the
ex-
pected
direction.
Alesina,
Roubini,
and
Cohen
(1997)
present
more
for-
mal
econometric
ests for he
United
States to
confirm
his
result,
using
autoregressive
quations
like
(4)
in
quarterly
data from
1947:I
through
1993:IV
with a
political
dummy
that
equals
+1 in
the
first
part
of a
Republican
administration,
1 in
the
first
art
of a
Democratic
adminis-
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
21/44
94
*
DRAZEN
tration,
nd
0 otherwise.
hey report
esults avorable
o
the
rational
partisan
heory
or
ealGDP
growth
nd
for
nemployment.
hey
find
significantolitical ummy ver hewhole ife f n administration,ut
by dividing
he
variable
ntofirst
nd
second
halves of
the
administra-
tion,
hey
eject
Hibbs's
version f
the
partisan
heory.
hey
run imilar
tests
n
a
sample
of 18 OECD countries
ver
he
period
1960-1993,
lso
finding
upport
or he
rational-partisan
odel
and
lack
of
support
or
both
heHibbs
and theNordhausmodel.
Faust
nd Irons
1999)
find
imilar
artisan
ifferences
n
both
utput
growth
nd
unemployment,
hich
re
strongest
n thefirst
alf
f
the
term.
However, hey
find his
partisan
differenceemains
ven
after
controlling
or bservableconomic ariables nd for
olitical
ffectss
in
partisan
models,
uggesting
hat
hedata do
not
give support
o
any
partisan
model.
Graphs
of
quarter-after-inauguration
ffects
imilar
o
those
presented
ere
may
be
found
for
large
group
of variables.
he
key mpirical
egularity
n which
here
s wide
agreement
s
REGULARITY
There
s a
clear
partisanffect
n economic
ctivity
n
the
United
tates,
with
conomic
ctivity
eing
ignificantly
igher
nder emo-
cratshan epublicansnthe irst alff heirerms.
5.2 INFLATION
AND MONETARY
OLICY
There
re
partisan
ifferences
n inflation
as
measured
y
the
rate
of
change
n theConsumer
rice
ndex),
hough
hey
o notconform
im-
ply
to
the
partisan
heory,specially
he
rational-partisan
heory.
emo-
cratic
dministrations
ave
lower
verage
nflation
han
Republican
d-
ministrations
n
the first
alfof their
erms,
ut
that nflation
s
rising
underDemocratsndfallingnderRepublicans,findingeportedoth
by
Alesina,
Roubini,
nd Cohen
1997)
and
by
Faust
and Irons
1999).
Hence,
the
basic
inflation
ata
for
he United
tates
do not
support
monetaryartisan
model,
whereby
he
evel
f
nflation
hould
be
higher
under
Democrats
han
Republicans.
In
interpreting
hese
results,
lesina,Roubini,
nd Cohen
argue
hat
the
differences
ound
n
changes
n
inflation
ates re
consistent
ith
their
heory,
hough
he
rgument
s
only
partially
onvincing,
ince
he
rational-partisanheory
ased
on
the
expectations-augmentedhillipscurve s builton the rateof
nflation,
oton
changes
n thatrate.The
econometric
ests
for
nflation
ycles
n the UnitedStates
re
far
ess
favorable
o
partisan
models,
paralleling
he
nonparametric
ests
dis-
cussed
above.
Alesina,
Roubini,
nd Cohen
(1997)
find
hat fter
973
(and
the
move
to
floating
ates
fter he
collapse
ofBretton
Woods),
he
difference
n
average
nflation
ates etween
emocratic
nd
Republican
This content downloaded from 152.118.148.226 on Thu, 6 Nov 2014 05:21:42 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp
8/9/2019 The Political Business Cycle After 25 Years_allan Drazen
22/44
ThePolitical usiness
ycle fter
5 Years
?
95
administration