ArticlePDF Available

Toward a New Typology of Vetoes and Overrides

Authors:

Abstract and Figures

This research develops a typology of vetoes and overrides based on the nature of the coalition that originally passed the legislation. We posit that both legislative and non-legislative objectives inform the strategies of the President and congressional leaders. The implications for measures of presidential and congressional power are considerable. Using veto and override data for the period 1969-98, we show that vote-switching between the passage of legislation and veto overrides is likely to occur only for a specific subset of legislation.
Content may be subject to copyright.
Toward
a New
Typology
of
Vetoes and
Overrides
RICHARD S. CONLEY and
AMIE KREPPEL
UNIVERSITY
OF FLORIDA
This research
develops
a
typology
of vetoes and overrides based on the
nature
of the
coalition
that
originally passed
the
legislation.
We
posit
that
both
legislative
and
non-legislative
objectives
inform the
strategies
of the
President and
congressional
leaders.
The
implications
for
measures
of
presidential
and
congressional power
are considerable.
Using
veto and
override data
for the
period
1969-98,
we show that
vote-switching
between the
passage
of
legislation
and veto overrides is
likely
to
occur
only
for
a
specific
subset
of
legislation.
The
power
of the President has been described as
the
"power
to
persuade"
(Neustadt 1960).
Defining
and
quantifying presidential
influence
is,
however,
a
difficult
task.
Scholars
have
variably
focused
on floor
success
rates
in
Congress,
aggregate party
or individual-level
legislative
support,
or agenda
success
(Edwards
1980. 1989;
Peterson
1990;
Bond and Fleisher
1990;
Jones
1994).
Recently
attention has been
paid
to the veto.
In
this era
of
nearly permanent
divided
government
in
the last two
decades,
vetoes and veto
threats have become
an
important
form of
leverage
for the
legislative
Presidency
(Cameron 2000).
In
addition,
the
propensity
of members
of
Congress
to
change
positions
in
support
of the President between
the
passage
of
legislation
and override
attempts
is an
important
consideration
in
assessing presidential
power
(Krehbiel
1998).
This research contends
that there are subtleties
in
the veto/override
game
that have
gone
unexamined.
Not all vetoes are created
equal.
Sometimes
Presi-
dents
veto
legislation originally passed
by strong bipartisan
majorities
and
face
the near
certainty
of a successful override.
In
other cases the
congressional
lead-
ership
may bring
overrides
to the floor on
legislation passed
by
thin
partisan
majorities
with
virtually
no
chance
of
success.
The
implications
for
measuring
presidential
and
congressional
power
are considerable.
Depending
upon
the
NOTE: This
article
was
originally
delivered
at
the Annual
Meeting
of the American
Political Science
Association,
September
1999,
Atlanta.
Political
Research
Quarterly,
Vol. 54,
No. 4 (December
2001)
pp.
831-852
831
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political Research
Quarterly
nature
of the
passage
coalition
of the
original legislation,
the President and con-
gressional
leaders
may
or
may
not need
or want to
lobby
members of
Congress
to
change
their
positions.
The
general
presumption
in
the
literature about vetoes and overrides
is that
the
actors involved are
always
focused
on
achieving legislative
outcomes. It is
assumed
that all actors have as their
primary goal
the
realization or the
blockage
of some
legislative
proposal
and
that
defeat
is
costly
(in
terms of
public,
institu-
tional,
and/or
political
prestige).
As a result
success
is measured
in
terms
of
leg-
islative outcomes and
power
or
influence
is
gauged
by
the
number of
legislators
who are convinced
to
switch
their
allegiance
from
one
side to the other.
We
argue
that this is an
oversimplification
of true
strategies
and
goals
involved.
Traditional
assumptions
conflict
with
empirical
evidence
on veto and override
attempts
brought
to bear
for
the
period
1969-98.
Voting
behavior
in
the
Congress, particularly
on contentious
issues,
is
one of
the few
real world
examples
of
near-perfect complete
information
for all of
the
actors involved. The vote breakdown
on the
final
conference
report, assuming
a
recorded
vote
is
taken,
is
public
information.
In
addition
both
political parties
and
the White House
have staff members whose
job
it
is
to
investigate
and
pre-
dict
how
members
will vote
once
they
receive
pressure
from
either
or both sides.
The President
has a
very good
idea
of the
potential
success
of
a veto
long
before
a
veto
is
cast.
The
same
is true of
congressional
leaders when
they contemplate
an override
attempt.
Yet the President's
ability
to sustain vetoes and
congressional
leaders'
attempts
to
trump
the
President's vetoes are
variably
unsuccessful.
We
posit
that most
of
these
unsuccessful
attempts
were
not miscalculations
or
failures.
Instead
these failed vetoes
and override
attempts
were
a kind of
position-taking
aimed
at
informing
the
public
and
building
electoral
support
rather than
affect-
ing
immediate
legislative
outcomes.
On
these
types
of vetoes and
overrides there
is little incentive
on the
part
of
the President
or
congressional
leaders
to
attempt
to
persuade
members
of
Con-
gress
to switch their
vote from one side
to the other.
Whenever
the President or
the
opposition
leadership
in
Congress
acts
to
change
the
voting
behavior
of a
member,
political capital
is
expended.
It would
not be
logical
to
expend
that
cap-
ital with the advance
knowledge
that efforts
to
persuade
legislators
will nonethe-
less result
in
almost certain
defeat.
By
the
same
token,
in
those
votes where
vic-
tory
is assured it
is equally
unlikely
that there
will be a great
deal
of effort
expended
in
attempting
to
sway
the votes of additional
members
if
their
support
is
unnecessary.
By
implication,
then,
current
measurements
of
presidential
and
congressional
power
and influence
that
rely
on
measurements
of
vote
switching
do
not
accurately gauge
the
actual
ability
of the President
or
congressional
lead-
ers to affect
change.
By assuming
that
legislative
goals
are
always
at the heart
of
presidential
vetoes and
congressional
overrides,
scholars
have
overlooked the
subtleties
of the veto tool.
The result
has
been
an
exaggeration
of the need
for
the
832
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward a New
Typology
of Vetoes and
Overrides
President and
congressional
leaders to
engage
in
coalition
building
to
sway
mem-
bers of
Congress
on an
important
subset of
vetoes.
We
develop
a
new
typology
to
classify
vetoes and overrides that better
cap-
tures the
complexity
of
inter-institutional
dynamics
and accounts for
legislative
and
non-legislative goals
of the President
and
congressional
leaders. We
perceive
three different
types
of
vetoes
(and
override
attempts)
based on the nature of
the
initial
passage
coalition
of
the
legislation:
(1)
partisan
vetoes where
the
original
conference
report
vote was
clearly
partisan
and
an
override
attempt
would most
likely
fail
(Type
I);
( 2)
contested vetoes on
contentious issues where
the outcome
based
on
the final
conference vote
is
unpredictable
(Type
II); and,
(3)
position-
taking
vetoes where the
original
legislation
was
passed
by
a
broad
super-major-
ity
coalition and an
override
attempt
would
likely
succeed
regardless
of
cam-
paigning
by
the President
(Type
III). We
begin
our
analysis
by
placing
our
typology
into
perspective
with current
literature
on
presidential power
relative to
vetoes
and overrides. We then
develop
a
model of
vetoes and overrides and test
the model's
implications
for
measuring presidential power
through
an
analysis
of
vote
switching
across
categories.
The
concluding
section
suggests
avenues
for
future research
based
on
the
findings.
PERSPECTIVES ON THE
VETO/OVERRIDE PROCESS
Most
scholarly
studies of
modern
(post-war)
Presidents' use of
the
veto
have
assumed
the
primacy
of
legislative goals
in
Presidents' calculations-that the Pres-
ident
normally expects
his
position
to
prevail
and works to
achieve
this
end.
With this frame of
reference
in
mind,
the incidence of vetoes has been
tied
to a
number of factors. Lee
(1975)
found that divided
government
contributes
to
the
use of the
veto,
and
Congress
is
more
likely
to
override Presidents who
lack
a
strong
electoral
showing. Copeland's
(1983) study
also
emphasizes
the
impor-
tance of
split-party
control of the
Presidency
and
Congress,
and finds
that
veto
use
increases
as the
President moves
through
his
term. Watson's
(1993)
analysis
dovetails with the latter
finding,
as
veto use
appears
to
increase
in
proportion
to
the President's loss of
congressional support
over
time,
presumably
as his
politi-
cal
capital
(Light
1982)
declines.
In
sum,
presidential
vetoes are the
product
both
of conflicts over institutional
prerogative
and
policy
concerns,
with variation
explained largely by
electoral forces and
partisan
configurations
between institu-
tions
(see
Ringelstein
1985;
Woolley
1991).
What
prompts
congressional
leaders to
pursue
overrides,
and do
they
simi-
larly
expect
to
prevail
in
most
cases? There
is,
interestingly,
little
scholarly
litera-
ture on
congressional
override
attempts,
but most studies
also
assume
legisla-
tive/policy
goals
as
the
primary
motivation. Rohde and Simon
(1985)
show
that
the
policy
area of the veto
does
not seem to affect whether an
override
is suc-
cessful,
but
congressional
leaders are more
apt
to
bring
about
attempts
to over-
833
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political Research
Quarterly
ride on
bills that
concern
social welfare and economic
policy.
Disaggregating
override
attempts
by
chamber and
extending
his
analysis
back to the
late
19th
century,
Hoff
(1992) discovers that
partisan support
for the President is most
important
in
determining
the first chamber's
success
in
overriding
a veto.
Most
interesting
is
Hoff's
finding
that a certain momentum builds after successful
pas-
sage by
the first
chamber,
implicitly limiting
the
impact
of
presidential
resources
to halt the override at that
point.'
We take
a
somewhat different
approach
to
the veto/override
process.
We are
interested
in
understanding
the
multiple
motives
behind
the
players'
objectives.
We
posit
that
legislative goals may
not
always
be
the
driving
force
behind
presi-
dential vetoes or
congressional
override
attempts.
As
we
explain
in
the
following
sections,
vetoes and overrides
may
sometimes constitute
an extreme form of
"going public"
(Kernell
1997). Players
in the
veto/override
game anticipate
losing,
but
have incentives
to embark
on
the course of action
for
symbolic
rea-
sons,
such as
to
distinguish positions
between the
parties' policy
positions
when
control of
the
Presidency
and
Congress
is divided
and
seek to derive electoral
benefit
among
their
supporting
constituencies.
OUR PERSPECTIVE: MISCALCULATION OR "GOING PUBLIC?"
If,
as is asserted
both
implicitly
and
explicitly
in
the
literature,
vetoes are a
tool
in
a
game
with
primarily
legislative goals
in
the short
term then the
empiri-
cal evidence
suggests
that
the
players
are
not
particularly savvy.
Between
1969
and
1998,
there
were 80
attempts
to
override the President's
veto
in
the House
of
Representatives, yet only
41
(51.3
percent)
were successful.
The odds of
a
suc-
cessful override
are about the same
in the Senate: Of
63
attempts during
the
same
time
period,
34 (54 percent)
have
been successful.
The
bicameral
nature of
the
override
process
would seem to
work to the President's
advantage. Only
30 times
from
1969-98
have
both chambers
of
Congress
overridden
presidential
vetoes.
Presidential success
in
sustaining
vetoes is rather variable.
President Ford was
overridden
12 times
in
28
attempts
in
both chambers.
President
Reagan
batted
.500 as congressional
leaders
overrode
nine of his
eighteen
vetoes.
And
Presi-
dents Bush and
Clinton
(1995-98)
were
overridden
only
a single
time
despite
numerous
override
attempts.
Which
factors
account
for the
fairly large
variation
in
both
the level of
presidential
success in
sustaining
vetoes
and
congressional
leaders'
success
on
overrides?
Given that information
is
very
close to
perfect
and
all actors have
access
at
the
very
least
to the
final
vote on
the
legislation,
why
have
there been so
many
"failures"
in
the veto and override
game?
Our
answer
is
simply
that
there is
more
than
one
game being
played.
The
standard
and familiar
"legislative"
veto
game
is
certainly
an
important
and
very
1
We
will
discuss
the
potential
influence of momentum further
below.
834
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward
a New
Typology
of
Vetoes and Overrides
FIGURE 1.
ELECTIONS
DETERMINE
WHO HAS THE
MAJORITY
IN THE CONGRESS
AND
CONTROLS
THE
PRESIDENCY
Pass-Partisan Pass-Bipartisan X Do not
pass
(I) Majority AIA
Veto Do
not
veto VetoDo not
veto
BABt CC2
(2) President
(1) Majority
Ovenide Do
not
override Override Do
not
override
DAD2 EE2F F1F2 G1G2
frequent
occurrence.
At
the
same
time,
what we
will call the
"position-taking"
veto/override
game
also
exists and
should be examined
independently.
The
goals
of the
actors
are
very
different
in
each case
and as a result
so
are
their
strategies
and actions.
It
is incorrect to assume
the
same
goals
for both
games
and
even
both
actors
(discussed
further
below).
In
fact,
these
assumptions
have
led
to mis-
interpretations
of
presidential
power
and influence.
In a simplified
world
with
only
one
chamber,
or a unified
majority
that
behaved
as
if
it were
one
chamber-and
assuming
complete
information-the
veto
game
can
be
interpreted
as the
game
tree
in
Figure
1.
Here
we have
"nature"
or the electorate
deciding
the
partisan
breakdown
of
Congress
and
the Presi-
dency
In
the
first round of action
a "majority"
in
the chamber
must
attempt
to
pass legislation.
This
can
have
three results.
The least
interesting
from our
point
of view
is that it fails and
no
legislation
is
passed,
at which
point
the
game
ends
(A1A2).
The two
more
interesting
options
are
that the
legislation
passes,
either
by
a
partisan
vote or
by
a
broad
supermajority
coalition
of at least 67
percent
of
the votes cast.
Note that this
is
already enough
to
guarantee
that a presidential
veto
can
be overridden
if
the
original
passage
coalition
holds. Because
the final
vote
on a
conference
report
is
public
there is
no information set.
At
the next
stage
in the
game
the
President must
decide to
veto the bill or
not.
In
the
vast
majority
of cases the
President does
not veto
legislation
(B1B2
and
C1C2),
either because
he
and the
majority largely agree
on
the
proposal
or
because
he knows
that he cannot
win a veto and
no
benefit
will be derived
from
a lost veto.
We are
primarily
interested,
however,
in
those
cases were
a veto
is
835
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political Research
Quarterly
issued.
We note that there are two
very
different situations
in
which
this can
occur;
(1)
a bill
supported
by
a
partisan majority;
and
(2)
a bill
supported by
a
supermajority
which
is
necessarily
at
least
partially bipartisan.
The
veto of a
par-
tisan bill is
substantially
less
likely
to be overridden
given
that the
President's
own
party
(which
has
never controlled less than a third
of
both
chambers
in
the
period analyzed
here)
supported
him in
his
opposition
to the bill.
On the other
hand,
a
presidential
veto
of
a
supermajority passed
bill
is
very likely
to be over-
ridden
given
that the bill has
already
demonstrated
its
attractiveness
to
members
of
both
parties
in
substantial
numbers.
At the
final
stage
of the
game
the
majority
must
decide
whether to
attempt
to override
(D1D2,
F1F2)
the
presidential
veto
or not
(E1E2,
G1G2).
Once
again
we have two
very
different
probable
outcomes
depending
on the
original major-
ity
which
passed
the
legislation.
If the
proposal
was
initially partisan
and
sup-
ported by
significantly
less than
the
qualified majority
needed to override
a
veto
it
is
unlikely
that the override
attempt
will
succeed
(D1D2).
On the other hand
if
the
original
majority
was above the
required
override threshold
the
likelihood
of
a successful
override
increases
significantly
(F1F2).
If
we
rely
on
the current
understandings
of the veto/override
process
as
a
primarily
short-term
legislative
game
we should
expect
the
following strategies
or
preference
orderings
by
the
President and the
majority,
assuming
a
disagreement
between the President and
the
majority
in
Congress.
Preferences
in
a
Legislative
Game:
President:
A2 > E2
> D2 > G2
> C2
> F2 > B2
Majority:
C1
> B1
> F1
> Al > El > D1
> G1
If
these
preference
orderings
are
correct
then
we
should
never
(or
rarely)
arrive
at the
"out
of
equilibrium"
outcomes
of
D
1D2
(the
majority
unsuccessfully
attempts
an
override)
and
F1F2
(the
President has
his veto
overridden).
These
represent
the least favored outcomes
of the
actors,
"mistakes"
(B1B2,
G1G2)
notwithstanding.2 Through
backward
induction
(possible
because
of
complete
information)
each
of
the actors would foresee
the outcomes
of each
possible
action and select
the
one
that
they
most
prefer.
Since each is
capable
of
prevent-
ing
the occurrence
of
their
least
preferred
outcome
(minus mistakes)
neither
D1D2
nor
F1F2
should
ever
occur,
yet they
both do.
2 These "mistakes" are outcomes
where the President or
the
majority
is
taking
actions that
lead to a
lesser-preferred
outcome
(again assuming
a
disagreement
between the
two).
These are
unlikely
out-
comes but must
be
included
because of
the
possibility
of a
"trembling
hand" which causes mistakes
despite
the
best
intentions of the actors.
These
outcomes
do occur
empirically,
though they
are rare
and
probably
have
to
do with the
perceived
costs
of
opposing popular
legislation
(B1B2)
or
a
pop-
ular
President
(G1G2).
These
may
also
occur
if
significant
numbers of members
have let it be
known that
they
will
switch their
votes,
changing
the
probable
outcome.
Vote
switching
will
be
dis-
cussed
further
in
sections 3 and 4.
836
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward
a
New
Typology
of
Vetoes
and
Overrides
Because
of the
significant
amount of information available to
the
actors the
perfect
information
assumption
is not that much
of a
stretch
from
reality.
In
each
of
the
above cases
the
likely
outcome would be
clear to all actors.
If the
President
vetoes
a proposal
that
has
already
received
the
support
of at
least 67
percent
of
the
legislators
it
is likely
that
he
will
be overridden.
In fact,
twenty-seven
of
twenty-nine
(93
percent)
of
the
successful overrides
from 1969-98
were
gleaned
on
legislation originally
passed
by
supermajorities.3
By
the same
token,
when the
majority
chooses
to
attempt
an
override
when it
clearly
lacked
the votes
to win
previously,
it is almost certain
to
lose.
Not a single
Type
1 (partisan)
veto
was
overridden
from 1969-98.
Given the odds
against
success
in
these cases
(D1D2 & F1F2),
we have to
assume either
that
the
potential
benefits
of
winning
were so
high
as to be worth
the
risk,
or that there were
benefits
inherent
in
making
the veto or
override
attempt
that
outweighed
the
potential
(and
probable)
costs
of
loss.
Judging
from
the
empirical frequency
of these
ostensibly
out of
equilibrium
outcomes
it
seems
highly likely
that there are indeed benefits
to
making
the
veto or
override
attempt
regardless
of
the
final
outcome.
Let us then reassess our
understanding
of
vetoes and overrides
in
light
of the
empirical
evidence.
It
is
unlikely
that both
the
President
and the
majority
in
Congress
continuously
and
consistently misjudge
their chances
of
winning.
Instead,
it
seems
more
likely
that sometimes
making
the
public
statement
of
opposition
is
believed
to
hold some benefit
greater
than
the
perceived
cost of
publicly
losing
the
legislative
battle.
Despite
an almost
guaranteed
loss,
the
veto or override is
attempted
because
the
goal
is not to
change
the
legislative
outcome,
but
to
gain
the attention and
sup-
port
of the
public
or
some
segment
thereof.
If
this were
the case the
preferences
of
the
players
would
have to
be modified and
we would need to model
the
veto/over-
ride
process
as
two
different
games
depending
on the
preferences
of
the actors. It
should
be noted
that
only
one
of the
two
actors
(in
this
stylized game)
needs
to
be
playing
the
position-taking
game
for the nature
of the
game
to
change.
If
the
first
set of
preferences
described
a
"legislative" game
then
the
following
second set describes
a "position-taking" game.
Here
both
actors
will
be assumed
to want
to make a public
statement. This
is
in
fact
a very
unlikely
outcome.
In
most cases at
least one of the actors
will
play
the
legislative
game
(and
win).
Again,
assuming
a
disagreement
between
the President and
the
majority
in
Congress,
Preferences
in a Position-Taking
Game:
President:
A2 >E2
> D2 > G2
> F2 > C2
> B2
Majority:
C1
> B1
> F1 >
Al > D1 > El > G1
We include
here
legislation adopted
by
voice
vote or
unanimous
consent
agreements,
in addition
to
legislation originally passed
by
a roll-call vote
with
67
percent
or
more of the chamber
voting
in
favor
of
the
measure.
By
their
very
nature,
voice votes
and unanimous consent
agreements
evince
broad
agreement
and
lack of
controversy
837
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political Research
Quarterly
All that
changes
are the
preferences
of the actors in the
subgames
where
they
are
likely
to lose.
For the
President this is the
supermajority passage subgame,
and
for the
majority
it
is
the
partisan passage subgame.
For
both actors
their
pref-
erence is now to risk the veto or
override
attempt
knowing
that
they
will lose
because
they prefer
to make
the
public
statement then do
nothing
at
all.
This
changes significantly
our
anticipated
outcomes such
that we now
expect
to arrive at
D1D2
when the
congressional majority
wants
to
make
a
public
statement or at
F1F2
when
the
President wants
to make a
public
statement. Note
that the
majority
can
play
the
position-taking
game
only
when the
original pro-
posal
was
passed by
a
partisan majority
and
it is
unlikely
to
be able
to
muster
the
requisite
override
majority. Similarly
the
President
can
only play
the
position-
taking
game
when a broad
super majority
passed
the
original
legislation
and a
successful
override is more
likely
to occur.
In
other
words,
only
the
probable
loser
can
play
the
position-taking game
while the
winner is still
playing
a legislative game.
Understanding
that there are
likely
to
be
goals
other than the immediate
leg-
islative outcome
in a
veto/override
game
is
a fundamental
first
step
in
gaining
a
better
understanding
of this
part
of the
legislative
process.
The
implications
are
particularly important
in
measuring presidential power
if the
President does
not
always
expect
to
succeed
in
sustaining
his
vetoes.
Recently
the
success
of
the
President
in
convincing
members
to switch
their
allegiance
to his side has been
used
as a
measurement
of
presidential
influence and
power
(Krehbiel
1998).
This
assumes
that the
President
not
only always
wants to win a
veto
(which
we have
just
demonstrated is
false)
but
also that
he
always
is
trying
to
convince members
to
switch.
This is also an
over-simplification
of
the
process.
As we show
in
the
next
section,
the President
may
not need switchers
in
cases
when
strong
intra-
party
cohesion
prevails
on the
original legislation
and therefore
would
not
expend precious
political capital
on
attempting
to
gain
them.
UNDERSTANDING
THE
TYPOLOGY
It
is clear
that there exists more
than one
potential goal
in
casting
a veto or
attempting
an override.
It
is
also
likely
that the
strategies pursued
once the
Pres-
ident wields
his veto
pen
will
vary.
Non-legislative
goals suggest
that
examining
"switchers"
as a measure
of
partisan
strength
or
presidential
influence
is
ques-
tionable
(Krehbiel
1998).
If
the
goal
is
not
always
to win
a veto or an
override
attempt
then
clearly
the
assumption
that either side
would
actively
work to
obtain
converts does not
always
hold true either.
In
fact,
given
the
possible
sce-
narios
it
only
makes sense
to
attempt
to
gain
converts
in a small set
of
very
explicit
circumstances.
To
test this
assertion,
we
divided
the set of
all vetoes
(and
override
attempts)
into three distinct
categories.
Two
of
these
types
(I and
III) are
drawn
from the
previous
section.
Unlike
the
last
section
in
which
we
presented
a
simplified
uni-
cameral
model,
here we
present
a more realistic model
by
incorporating
both
838
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward a New
Typology
of
Vetoes
and
Overrides
chambers
into the
discussion.
For
simplicity's
sake, however,
we
will look
only
at
presidential
vetoes,
although
the same
arguments
hold
in the reverse for
majority
override
attempts.
We
focus
on the
period
1969-98
for several reasons.
In all but six
years
during
this
time
span
(1977-80; 1993-94),
control of the
Presidency
and Con-
gress
was
split
with
Presidents
of
both
parties
facing
an
opposition majority
in
Congress.
The
early
1970s witnessed
a variety
of institutional reforms
that
aug-
mented
congressional capacity
to
challenge
the
President
(Sundquist
1981;
Dodd
1979;
Dodd and
Oppenheimer
1977
). Democrats redoubled
efforts to shore
up
unity
in
the
party
in the
1980s and
the
trend
toward
a more
centralized
Congress
continued
well into the 1990s when
Republicans captured Congress
(see
Rohde
1991;
Sinclair
1992;
Hernnson
and Patterson
1995;
Davidson
and Thurber
1995;
Aldrich and Rohde
1997-98).
Additionally,
1969
represents
an
acceptable
histor-
ical
dividing
line
because
neither President
Kennedy
nor
President
Johnson
faced
a single
override
attempt
from
1961-68.
Finally,
extending
our
analysis
back
to
the 1940s and
1950s is
problematic
in
terms
of the
vote-switching analysis
that
follows.
Fewer votes were recorded and
much
legislative
business
was transacted
by
voice
vote or unanimous
consent,
rendering
a comparison
of
members'
posi-
tions on the
passage
of
legislation
relative to veto override
attempts
impossible.
The
first
type
of veto is the
"partisan
vote"
veto
(Type
I).
In
this case
the
Pres-
ident
is
vetoing legislation
which divided the
Congress
along party
lines
in
at
least
one chamber. Most
of the
President's
party
has supported
his
position
already
(in
one
or
both
chambers)
by voting
against
the
original
proposal.4
There
is
little or no chance
of a
successful
override
given
this
type
of
partisan
support.
The second
type
of veto
is
what we call the
"contested
veto"
(Type
II).
In
this
case the
vote
is
neither
clearly partisan
nor
is
it
clearly
bipartisan
in at least one
chamber
with the other chamber
being
either contested
or
supermajority
(Type
III
below).
We have set
the
support
interval
for on the
original legislation
for this
type
of veto between 60 and
74
percent
for
bills
originally
passed
via recorded
roll
calls.5
It is
important
to remember
that this is a
subjective
decision and
other
intervals could
easily
be
argued
for. We selected this
interval because
it
nicely
borders
the
required
67 percent
threshold
for a successful
override
(?
7) and
includes
the
probable
number
of
potential
switchers based
on our
empirical
analysis.
In
the House
it
would
be
necessary
to
gain
at least 30-35
votes,
again
assuming
100 percent
participation.
Since
on average participation
in both
Clearly
this is most
likely
to occur
during periods
of divided
government.
Since these have been
the norm for
most of the
postwar period
this is not a serious limitation.
It would
be
enough
for
the
proposal
to have
been
passed
by
a
partisan
majority
in
one chamber
to make a veto
comparatively
safe.
Our interval
is
very
close to
the 4th
"pivotal"
interval
described
by
Krehbiel,
which is between 60
and 80
percent
of the vote. We believe
this is a bit too
large
for reasons that
will be discussed
in
the text.
839
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political Research
Quarterly
chambers falls short of the 100
percent
mark the actual number of switchers
required
would be
somewhat lower.
According
to our
data,
during
the
period
1969-98,
on
average
the
President was able to convert
just
over
23 of
his
co-par-
tisans
in
the House who had voted for the
original legislation,
and
just
over 6
members
of
the
opposition party
Thus,
our threshold
nicely incorporates
the
empirical
realities
of
potential
switchers
in
this uncertain
category
of
vetoes.
Regardless
of the
specific
threshold
used,
the
Type
II veto is
clearly quite
con-
tentious and as a result the
probable
winner of
any
veto or
subsequent
override
attempt
is unclear at the
outset
of the
game.
The third and final
type
of
veto is
what we have
called
above the
"superma-
jority
veto"
(Type
III).
In
this case
a very
large majority passed
the
original pro-
posal
in
both chambers. We have set this
at
75
percent plus
for the reasons
given
above,
but
again
this is somewhat
arbitrary
and other boundaries
arguably
could
be
used.
In
this case a substantial number
of
the members
from the
President's
party
have
already
voted for the
legislation,
and therefore
against
him.
The
number of
defectors
is
higher
than
what we
should
normally expect
that the
President could
hope
to convince to return to his side
(switch).
This is
the
type
of veto that we have described as a
position-taking
veto
in
that we believe the
goals
of
the President
in
this
case are not to
impact
the
legislative game,
but
the
public
relations
game.
We now have
a
typology
of three different kinds
of
presidential
vetoes which
we believe should be understood
to be very
different
in terms of
presidential
strategy.
The literature has
suggested
that we measure
partisan strength
and
pres-
idential influence
by examining
the
ability
of the President
to convince members
of
Congress
(and
particularly
those
in
his
own
party)
to switch
their votes
and
oppose
a potential
override.
We
agree
that this is
an
interesting
and
potentially
informative
approach
to
understanding presidential
influence,
but
we
must
not
continue to assume that all vetoes
are
created
equal.
In
fact,
not
only
is
it
likely
that the
goals
of vetoes
vary (legislative
vs.
posi-
tion-
taking),
but
also that the
strategies employed by
the President
vary depend-
ing
on the
type
of veto
we are
discussing.
Vetoes of
Type
I occur
when the
Pres-
ident
already
has
the
support
of the
bulk
of his
party
(a partisan
vote
on the
original legislation).
He is therefore
unlikely
to lead
any
attempt
to
win vote
switchers
if
he faces an override.
Clearly, convincing
a member
of
Congress
to
switch
her
position
on a
piece
of
legislation
is difficult since
this reversal will
nec-
essarily
be
public.
This means the President
must
expend
a fair amount of
polit-
ical
capital
in the
attempt.
Why
would
he do this
if
he
did
not need
to? In
addi-
tion,
if
most of his
party
(and
those who would be
most
likely
to
switch)
have
already
voted with
him,
whom would
he
try
to convince? While
it
is
true that the
President
is
able
to
get
switchers
from the
opposition party,
this is
more
difficult,
and
probably
more
costly
(Krehbiel,
1998).
Thus,
it seems
unlikely
that
that the
President would
expend
much
effort to obtain switchers
if
an override
were
840
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward a New
Typology
of Vetoes
and
Overrides
attempted.
His
veto
"staying power"
is
relatively
assured and the
override threat
is most
likely
an
empty
one
(or
a
position-taking
one).
This leaves
us with
presidential
strategy
on
Type
II and
Type
III
vetoes. Let
us examine
Type
II
vetoes
first,
as it is this set
of vetoes for which
we
expect
the
active
lobbying
of members has the
potential
make
a difference
in
the
eventual
outcome.
When the
original
vote
on the
legislation
was
in the 60-74
percent
range
it is unclear
which
side
will
be
successful
in an eventual veto and override
battle.
In
this case
everything
will be decided
by
the
ability
of both sides
to retain
their
previous
support
while at the same
time
lobbying
potential
switchers.
In
this case the President is
most
likely
to
bring
all of the resources
he
has
to bear
in an
attempt
to
prevent
a successful override.
It
is
in
these cases
that the relative
success
of a President
in
obtaining
"switchers"
(from
either
party)
can
fairly
be
considered
a
kind
of
proxy
for
presidential
power
or influence.
Our
final
category
is the
Type
III
veto situation.
In these cases
the
original
legislation
has
passed
in
both
chambers
by
a supermajority
(> 74 percent).
If
Presidents
choose to
veto
legislation
in
this
situation,
given
the existence
of near-
perfect
information we
might
assume that
they
are
doing
so for
position-taking
rather than
legislative
reasons.
In
other words
the
goal
is
not the immediate
leg-
islative outcome
(which
seems
a foregone
conclusion)
but
rather a more
long-
term contest
fought primarily
in the battlefield
of
public
opinion.
When Presi-
dents
are
simply positiontaking,
they expect
to lose
on the override
attempt
and
consequently
should
not be expected
to invest
time and
political
capital
in
a
losing
cause.
Thus,
we should
expect
to see
little to
no
attempt by
the President
to persuade
members
of either
party
to switch.
Here
again
then,
counting
"switchers"
would seem
to be
at
best
an inaccurate
method
of
measuring presi-
dential
influence.
But,
as
the
empirical
evidence
below
demonstrates,
an override
attempt
fol-
lowing
a
bicameral
supermajority passage
on the
original
legislation
may
be
more
or
less
complicated.
Eventual
presidential
veto success
can
depend
on a number
of variables
including
the
relative
strength
of the President's
partisan
base
in
Con-
gress
and the resources
he can
marshal
(popularity,
media
attention,
etc.)
to
per-
suade
members
to
change positions.
This is
in
part
because
of the
ability
of
the
President
to
adopt
a more
complex
"bicameral"
strategy
in
which
he
attempts
to
reduce
the
passage
coalition
on the override
attempt
in the first
chamber
to
slow
the momentum
and
halt a successful override
in
the second
chamber.
TYPES OF VETOES AND
OVERRIDES,
1969-98: AN EMPIRICAL
ANALYSIS
With what
frequency
do Presidents
cast different
types
of vetoes? Table
1
par-
titions
presidential
vetoes
from1969-98
into the
three
categories
discussed
earlier.
Several
trends
are
immediately
evident.
First,
the vast
majority
of vetoes
from
Presidents
Nixon
through
Reagan
were on
Type
III
bills-legislation
passed by
75
841
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political Research
Quarterly
TABLE 1.
PRESIDENTIAL VETOES OF
PUBLIC
BILLS
BY
TYPE,
NIXON-CLINTON
(1969-98)
President #
of
vetoes Type
1 Type
2 Type
3
Nixon 24 1 6 17
(4.1%) (25.0%) (70.8%)
Ford 47 5 6 36
(10.6%) (12.8%) (76.6%)
Carter 13 0 3 10
(0.0%) (23.0%) (76.9%)
Reagan 36 2 6 28
(5.5%) (16.7%) (77.8%)
Bush 29 6 12 11
(20.7%) (41.3%) (37.9%)
Clinton 24 12 6 6
(50.0%) (25.0%) (25.0%)
Column
Totals 173 26 39 108
(15.0%) (22.5%) (62.4%)
percent
plus
of both chambers
(or
voice vote/unanimous
consent
combinations).
Here we find
preliminary
evidence
that Presidents
are
undeterred
from
vetoing
legislation
that
may
result
in
a successful
override,
accentuating
that more than
legislative goals
may
be
at
play
in their calculus.
Second,
Type
II
vetoes,
or
legis-
lation
passed
by
a supermajority
in one
chamber but
not
the
other,
represent
an
intermediate
category.
Just
less
than a quarter
of all vetoes for
the
period
were of
the
Type
II
variety.
Third,
and
quite importantly,
Bush's
Presidency
would
seem to
mark an
important
turning
point
in the nature of
legislation
subject
to
vetoes.
The
trend
becomes
more
pronounced
under
Clinton. A fifth of the vetoes
cast
by
Bush,
and
fully
half
of
those cast
by
Clinton
from
1995-98,
were on
legislation
passed
by
partisan
majorities.
The
implications
are
particularly
important
in
considering
attempts
by
the
congressional majority
to
override under
divided
government.
Examining
the
frequency
of override
attempts
on the three
veto
types
tells
us much
about
the
nature
of
presidential-congressional
relations
under
divided
government
as
party
cohesion
in
Congress
has
heightened
over
the
past
decade
(Tables
2a/2b/2c).
Presidents'
"staying power"
on
vetoes is
greater.
And
they
pre-
sumably
need
to
do little
arm-twisting
of
co-partisans
to
thwart
overrides.
On
Type
1
vetoes,
congressional
leaders
bring
override
attempts
to
the floor for
posi-
tion-taking
or
symbolic
reasons.
Not
a single
override
attempt
on
Type
1
vetoes
was successful
in
either
chamber,
though
this has not dissuaded leaders
of
the
opposition
majority
in the House
or Senate from
bringing
attempts
to the floor
with
greater
frequency during
the Bush and
Clinton
presidencies.
842
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward
a
New
Typology
of Vetoes and Overrides
TABLE 2A.
OVERRIDE ATTEMPTS ON
TYPE
1
VETOES,
NIXON-CLINTON
(1969-98)
House House Senate Senate Total
President Attempts Successes Attempts Successes Overriden
Nixon 1 0 0 0 0
Ford 1 0 0 0 0
Carter 0 0 0 0 0
Reagan 1 0 0 0 0
Bush 2 0 3 0 0
Clinton 5 0 0 0 0
TABLE 2B.
OVERRIDE
ATTEMPTS ON TYPE
2 VETOES,
NIXON-CLINTON
(1969-98)
House House Senate Senate Total
President Attempts
Successes Attempts
Successes Overriden
Nixon 2 1 5 1 1
Ford 5 1 1 0 0
Carter 1 1 1 1 1
Reagan 2 1 2 0 0
Bush 7 1 4 1 0
Clinton 1 1 1 0 0
TABLE 2C.
OVERRIDE
ATTEMPTS ON
TYPE
3
VETOES,
NIXON-CLINTON
(1969-98)
House House Senate Senate Total
President Attempts
Successes Attempts
Successes Overriden
Nixon 10 4 10 6 4
Ford 19 15 18 13 12
Carter 3 1 1 1 1
Reagan 11 9 10 9 9
Bush 8 5 5 1 1
Clinton 2 2 2 1 1
Override
attempts
in
the
Bush and Clinton
presidencies
emphasize
how
non-legislative
goals may
inform the
opposition
majority's
decision
to
attempt
a
veto
override.6
Bush faced
two House
attempts
and
three
Senate
attempts
on
6 Nixon's
veto of
H.R.
10511,
the Urban Mass
Transportation
Act,
and
his
veto of S.
3418,
the Public
Health Service
Act,
are
not considered
in
the veto and override
analyses,
though
the
two
vetoes are
often
listed as successful overrides.
The
vetoes were the
subject
of
litigation
in which the courts
decided that
the President's
pocket
veto of
the
two
bills was unconstitutional.
Consequently,
there
was
no override
attempt
and the law
was
promulgated according
to
the
original
vote.
843
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political Research
Quarterly
Type
I
vetoes. The
subjects
of the bills
clearly
were
among
the most
contentious
of
Bush's
term,
including increasing
the
minimum
wage,
taxes,
civil
rights,
cam-
paign spending,
and
"motor
voter." Democratic
leaders knew
in
advance that the
overrides would
fail,
but
pushed
forth
anyway
to
publicly
demonstrate their
resolve.
Override
attempts
on
taxes,
campaign spending,
and
"motor
voter" were
aimed
at
symbolically
blaming
the President
for
apparent gridlock.
On the min-
imum
wage
and civil
rights
bills,
a compromise
was
eventually
reached on
sub-
sequent
legislation
that
won Bush's
signature, largely
because
the President
doubted he could sustain
his vetoes
upon
a second
round
of
override
attempts
(Congressional
Quarterly
Almanac
1989).
Clinton's
bout
with House
Republicans
on
Type
I
vetoes is a more
straight-
forward
case
of
the
majority opposition
in
Congress
attempting
to
paint
the
Pres-
ident
as the obstructionist
in
the court
of
public
opinion.
Four
of the five
over-
ride
attempts
on
Type
1
vetoes
in the
104th
Congress
were
on
budget
legislation
(the
other
was
product liability
reform).
The
Republican
spending
bills,
which
had been
passed
along
party-line
votes,
took aim at
cutting
a
variety
of
domestic
and
foreign
policy
programs
in
conjunction
with
the
goals
outlined
in
the
Con-
tract
with America.
From
the
outset,
Speaker
Gingrich
and
Republican
leaders
in
the House
knew the overrides had
little
hope
of
passing:
But
success was not the
point. Gingrich
hoped
to call the President's bluff
on
the
government
shutdown
and
turn
public opinion
against
Clinton. The
budget
brinkmanship
backfired.
What
Gingrich
and
his
team
of leaders
apparently
did not
realize was that the
vast media
resources
of
the
Presidency
gave
Clinton
a decided
edge
to shift the
tide of
public opinion
to
his benefit.
Moreover,
public
opinion
of
Congress
tends
to
dip
when
Congress
challenges
the
President
on vetoes
(Durr,
Gilmour,
and
Wolbrecht
1997).
As
a result,
it was the
Republican
leadership
that was held to
blame
for the shutdown
and was forced
to
compromise
on
spending
levels
to the
preference
of
Clinton.
Tables 2b and
2c
suggest
that the bicameral
process
of veto overrides
vari-
ably
works
to the President's
advantage
on bills that
have
stronger
bipartisan sup-
port
in one
chamber.
Recalling
that
Type
II
vetoes are
bills
passed
by
a
superma-
jority
in
one chamber but
not
the
other;
the total number
of vetoes
overridden is
quite
small relative
to
the
number
attempted.
If
congressional
leaders can muster
the
requisite
votes
to
override
in the
chamber
that
originally passed
the
legisla-
tion
along
strong
bipartisan
lines,
the
second chamber
override
attempt
most
often
fails
in the
absence
of
prior
consensus
among
legislators
in
that chamber.
Again,
we are
left with the
probability
that
congressional
leaders
suspect
that the
override
will
likely
fail in the other
chamber,
but
go forth with
an override
attempt
to make a public
statement.
A cursory
review
of
Type
II vetoes,
from
Reagan's
vetoes
of
industry protections
from
foreign competition
to Bush's
vetoes
of
family
leave and
abortion
counseling,
suggests
that this
is
generally
the case.
There
is,
however,
more
uncertainty
about
the outcome
of
Type
II overrides. A
844
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward
a New
Typology
of Vetoes and Overrides
separate
question,
addressed
momentarily
in an
analysis
of vote
switching,
is
how effective Presidents are
in
converting
enough
members
to either
stop
the
override
altogether
or
slow
the momentum
to override when the
legislation orig-
inally passed
one
chamber with
strong
bipartisan support.
Presidents
most often see their vetoes
trumped
on
Type
III
vetoes,
or those
bills
originally
passed by
supermajorities
in
both chambers. Ford
has the
poor-
est record of
all
six
Presidents,
no doubt
in
large
part
because
of the
sheer deficit
of
Republican
strength
in
the House and
Senate
following
the
1974
elections
that
brought
the Democratic
"Watergate
babies" to
Capitol
Hill
in
the wake
of
Nixon's
resignation.
Though Reagan's
early
legislative
victories
in
1981 are
often
the focus
of
analysis,
he has the second
poorest
record on
Type
III vetoes,
having
been
overridden a total
of nine times.
Why
did
Ford and
Reagan apparently
veto so
much
legislation
they surely anticipated
would be overridden?
An
examination
of
their veto
messages
reveals that the common
thread was often
a
desire
to make
a public
statement
about
allegedly
wasteful
spending by
the
Democratic
major-
ity.
Ford vetoed
a wide
variety
of
spending
legislation
(health
programs,
public
works,
school
lunches,
etc.)
on the
principle
that
deficit
spending
was inflation-
ary
and would
prove
deleterious
to an
already teetering
economy
The
short-term
costs of
austerity,
Ford
believed,
would
be far
outweighed by
the
long-term
ben-
efits of
lowering
inflation
and
unemployment.
But
the
popular
programs
never-
theless
resonated
with members'
constituents,
and with
only
a third of
the House
controlled
by Republicans,
the Democratic
majority frequently
overrode the
President's
vetoes.
In
similar
fashion
to
Ford,
Reagan
took issue with
allegedly
excessive
con-
gressional spending.
Reagan
vetoed the
Water
Quality
Act
in
1987,
calling
the
bill
"loaded
with waste and larded
with
pork."
He
appealed
to
Congress
to
reduce
outlays by
two-thirds,
but
legislators
instead
easily
overrode the
veto with
an
eye
to
various
pet
constituency projects.
In a similar fashion
Reagan objected
to
the
Surface
Transportation
Act of
1987,
which
he
regarded
as
wasteful
spending
for
congressional
pork
barrel
projects
with
some
$88
billion for 120
special
"demon-
stration"
projects.
But the President's
political
capital
had
waned
to such
a
degree
that not even
a personal
plea
to the Senate
could forestall
the override.
Reagan
also took rather
uncompromising
and confrontational
stances
toward
an
opposition
Congress
that,
while
symbolic
of
conservative
cause he
championed,
nevertheless
cost
him a variety
of losses-some of which were
rather
spectacular,
including
the Grove
City
civil
rights legislation
and South
Africa sanctions.
The South
Africa sanctions
veto
override
was
perhaps
the most
symbolic
and
damaging.
When
Congress
passed
legislation
to
impose
a host of
restrictions
on the
pro-apartheid
government
of South
Africa,
Reagan
vetoed the
bill on the
grounds
that the sanctions
would
only
add to the
despair
of
the black
population.
But
many
in
Congress argued
that
Reagan
was out
of touch with
public
opinion,
and
Democrats
rounded
up
a
substantial
number of
Republicans
845
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political Research
Quarterly
in
both
chambers
to
easily
override the veto. The
Congressional
Quarterly
Almanac
called the
override,
the first on a foreign
policy
issue since
passage
of the War
Powers
Resolution over Nixon's veto
in
1973,
"The
most
serious
defeat
Reagan
had suffered
on
a foreign
issue and one of the most
stunning
blows of his
Presi-
dency"
(Congressional
Quarterly
Almanac
1986:
359).
Table
2c shows that there is
also more
going
on
than meets the
eye.
Exam-
ining
Type
III
vetoes
in
the
past
decade,
there
is
some evidence that electoral
and
institutional
changes
that have boosted
party unity
voting
in Congress
have
indeed altered veto/override
dynamics beginning
with Bush's
term
(see
Rohde
1992,
1991;
Sinclair
1992;
Bond and
Fleisher
1996).
What
is
striking
is how few
overrides
were
successful
on
Type
III
vetoes for Bush and
Clinton.
Despite
a siz-
able
number of bills
passed
by
broad coalitions and
vetoed
(Table
1),
only
twice
since
1989
has
the
President
been overridden
in
both
chambers. Bush was over-
ridden
only
on
cable television
re-regulation;
Clinton
was
overridden
only
on
shareholder
lawsuits,
in
a veto that
puzzled
Democratic leaders who
thought
they
had
worked out technical issues to the President's favor
(see
Congressional
Quarterly
Almanac
1995,
2: 92). This
finding begs
the
question
of
whether
stronger parties
in
Congress
make the
President's
task of
persuading
members to
back his vetoes when
challenged
easier--even
on
legislation originally
passed
with
a
strong
bipartisan spirit.
A
brief
analysis
of vote
switching
casts
light
on this
and
other
questions
raised
by
Tables 1-2.
The
Meaning
of
Switching
Examining congressional
vote
switching
between
the
passage
of
legislation
and an override
attempt
is
a preliminary
means to assess
presidential
legislative
influence.
In the
following
analysis,
we
compare
vote
switching by type
of
veto,
according
to the nature of the
passage
coalition of the
original
legislation.
The
number
of
override
attempts
in the
vote
switching
analysis
(Tables
3a/3b/3c)
is
fewer
by type
than the number
reported
in
Tables 2a/2b/2c
for
two
reasons.7
First,
we limit the
analysis
to
those bills for which
a recorded
roll-call vote was available
on
the
original
legislation.
Bills that
originally passed
by
voice votes and
unani-
mous
consent
agreements
were excluded
from
the
analysis.
Second,
we confine
our
analysis
to override
attempts
first
begun
in
the House of
Representatives.
We
expect
that
the
success
or
failure of second chamber override
attempts
is
heavily
conditioned
by
the
margin
by
which the first chamber overrode
the
President,
and
for our
purposes
here,
does not constitute the best test
of
presidential
influence.
7
We do not include
Clinton's
line-item vetoes
in
the
analysis.
Clinton vetoed
certain items
in HR
2014,
a military
construction
appropriations
bill
in
the 105th
Congress.
He was "overridden" on
the line-item veto via
HR
2624,
a bill of
disapproval,
which
required
only
a simple majority
vote,
not
a
supermajority.
The line-item veto
was,
of
course,
later ruled
unconstitutional
by
the
courts.
846
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward a New
Typology
of Vetoes and
Overrides
TABLE 3A.
AVERAGE NUMBER
OF VOTE
SWITCHERS
ON TYPE I VETO OVERRIDE ATTEMPTS
Partisan Partisan Opposition Opposition
President Conversions Defections Conversions Defections
Nixon 7 1 10 1
(n
= 1) (0) (0) (0) (0)
Ford 8 1 7 10.5
(n
=
2) (5.66) (0) (5.66) (2.12)
Carter 32 40 0 141
(n
= 1) (0) (0) (0) (0)
Reagan
Bush 1.33 .33 3.00 8.67
(n =
3) (1.53) (.58) (2.65) (15.01)
Clinton 7.50 .50 0 5.00
(n
= 2) (3.54) (.71) (0) (2.83)
Standard deviations
in
parentheses.
TABLE
3B.
AVERAGE NUMBER
OF VOTE
SWITCHERS
ON
TYPE
II VETO OVERRIDE
ATTEMPTS
Partisan Partisan Opposition Opposition
President Conversions Defections Conversions Defections
Nixon 27.33 4.00 4.33 6.00
(n
= 3) (28.31) (2.00) (.58) (6.08)
Ford 13.40 1.40 3.20 5.20
(n
= 5) (12.24) (1.67) (3.03) (5.36)
Carter - -
Reagan 20.33 2.33 3.33 9.67
(n
= 3) (24.09) (2.31) (4.93) (6.43)
Bush 9.20 3.00 2.60 3.00
(n
= 5) (15.58) (2.92) (.55) (1.58)
Clinton 7.20 1.00 .20 2.00
(n
= 5) (8.04) (1.22) (.45) (2.45)
Standard
deviations in
parentheses.
We lose very
few
cases
following
this
methodology
because most
override
attempts begin
in the
House
of
Representatives,
which is
more beholden
to con-
stituency
pressures
and
popular
politics
(Hoff
1992:
176).
From
1969-98,
69 of
the 97
(71.1
percent)
override
attempts
started
in
the House.
We make
only
one
847
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political Research
Quarterly
TABLE
3C.
AVERAGE
NUMBER OF VOTE
SWITCHERS
ON
TYPE III VETO OVERRIDE ATTEMPTS
Partisan Partisan Opposition Opposition
President Conversions Defections Conversions Defections
Nixon 60.00 1.00 11.00 .80
(n
= 5) (28.18) (2.24) (9.43) (.84)
Ford 29.73 1.55 10.55 3.00
(n
= 11) (28.64) (2.07) (12.74) (3.66)
Carter 38.00 1.50 19.00 .00
(n
=
2) (48.08) (2.12) (26.87) (.00)
Reagan 34.29 3.71 1.86 11.71
(n
= 7) (23.08) (8.96) (2.34) (27.55)
Bush 28.17 1.67 15.17 2.67
(n
=
6) (29.51) (1.37) (21.94) (2.42)
Clinton 4 5 0 0
(n
= 1) (0) (0) (0) (0)
Standard deviations
in
parentheses.
small
change
to the
"type"
of veto override: We
consider
Type
II
those
overrides
which
passed
with a margin
of between
60
and
74
percent
in the
House,
and
exclude consideration
of
the
passage
coalition
in
the Senate.
Consistent
with our
expectations,
far
less
vote
switching
occurs on
Type
I
vetoes
compared
to the other
types-a
fact that casts doubt
upon
the
utility
of
measuring
presidential
power
with the inclusion
of
this distinct
type
of
veto
override.
With a strong
base
of
support
on the
original legislation,
Presidents
need to
do
very
little
when
faced
with the override
attempt.
Few
partisans
defect
to
support
the override
after
voting against
the
original
legislation
and
few
parti-
sans
are
available
to
convert.
On
average,
only
about
eight
members
of the Pres-
ident's
party
convert
to the
President's
position
for
the
entire
period
(or
=
9.74).
Members
of
the
opposition
party
who
originally
voted
against
the
legislation,
however,
do
apparently
fall
under
leadership
and/or
constituency pressure
to
back
away
from
supporting
the
President on
the
override
attempt.
But their num-
bers
are too few to
make
any
substantive difference
in
the
final outcome. The
only exception
was
Jimmy
Carter's 1980
veto of
a bill to raise the
public
debt
limit,
which evidenced a
widespread
number of defections
from
Democrats and
Republicans
alike.
Much more
vote
switching
occurs
on
Type
II
and
Type
III
veto
overrides.
Presidents'
ability
to
convert
and
retain
support
is
more volatile on this set of
vetoes,
as the
high
standard deviations
for
virtually
all cases
suggest.
But there is
a
stronger partisan
effect on
Type
II
vetoes.
If
Presidents
are successful
in
halting
848
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward
a
New
Typology
of
Vetoes
and
Overrides
the
override
attempt
in
the
House,
it is
because
of
their
ability
to convert
way-
ward
co-partisans.
On the subset of vetoes
begun
in
the House
for which
the
original
legislation passed by
a
margin
of 67
percent
or
more,
Presidents
man-
aged
to
halt the
override
eight
times out of eleven
(72.7
percent).
Nixon
tops
the
six
Presidents
for the
average
conversion
of
partisans,
followed
by Reagan,
Ford,
Bush,
and
Clinton.
In
few cases were
any
of
the Presidents able
to
convert sub-
stantial numbers
of the
opposition
in
Congress.
Hence,
whatever
effort Presi-
dents
must
expend
to
stop
veto
overrides
on
Type
II
legislation
seems
particu-
larly
worthwhile
in
terms of
building in-party support.
As Table
1
shows,
Presidents are less successful overall
in
halting Type
III
veto
overrides,
and this set of votes
presents
something
of
a paradox
on vote
switching.
It is for
this
set of
legislation originally
passed by
bipartisan
superma-
jorities
(75
percent
plus)
that the
partisan
effect of
conversions
appears
the
great-
est. Nixon
again
tops
the
list,
converting
on
average
60 of
his
co-partisans
in
the
House. The other five Presidents
(excluding
Clinton)
have lower scores between
28
and
38
partisan
conversions,
and
occasionally pick up
a
few votes from the
opposition.
Of what use are the conversion
of
partisans
on
Type
III
vetoes? Do Presidents
really
expect
to
prevail?
The answer
is mixed
from
the vote
switching
analysis.
Nixon,
for
example,
had the
highest
partisan
conversion
rate.
In
two
of the five
cases the conversion of
enough partisans
crushed the
override
in
the House:
Rural water/sewer
projects
and
supplemental
appropriations.
But the
partisan
conversions
on the other three bills
were insufficient
to
halt the override
in
the
House,
and
apparently
did
not
slow
the momentum
in
the
Senate,
as he was ulti-
mately
overridden:
Railroad
Retirement
act,
hospital
construction/public
works,
education
appropriations.
And
despite
converting
an
average
of 34
partisans,
Reagan
was
ultimately
overridden on all of the
Type
III
vetoes
begun
in
the
House. But
there is still more than meets
the
eye.
There
may
be an
advantage
to
rendering
the
override
attempt
more
partisan
in
the
House,
even
if
the President
loses
in
that chamber.
We do not
present
data on the
Senate
here,
but it is worth
noting
that
while Bush was
overridden
in
the House
on four of
the five
Type
III
vetoes,
Senate Democrats
did not
manage
to override
any
of the four.
In
similar
fashion,
in
four of twelve
cases Ford was overridden
in
the
House,
but the Senate
could not muster
the
necessary
votes to override.
In
sum,
this
indirect evidence
implies
that House conversions
may help
to erode
the
appearance
of consensus
on
the
legislation
and
improve
the
possibility
that the President
can sustain
the
veto
in
the
upper
chamber.
The
analysis
nevertheless
suggests
that
Presidents do often cast
vetoes for
symbolic
reasons
to make
a
public
statement,
with
the
expectation
of
losing
on
Type
III
vetoes.
Assuming perfect
information,
the
President
has a fairly good
idea about his
capacity
to
persuade
based on the
nature of the
original legisla-
tion.
Only
in
rare cases
are
partisan
conversions
enough
to stifle
the override
849
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political
Research
Quarterly
attempt
in
the House.
Moreover,
there is often
a
great disparity
in the number
of
partisans
Presidents are able to
convert.
The
high
standard
deviations
in
Table
3c
tap
into this
phenomenon.
For
example, Reagan
converted
only
five
Republicans
on
the South Africa sanctions bill in 1986. On six occasions Ford
managed
to
convert less than ten
partisans
on
Type
III
legislation.
These votes
correspond
generally
to
the earlier discussion about the two Presidents' economic
stands,
which were
largely symbolic.
REPRISE AND FURTHER RESEARCH
The central
goal
of
this research
has been to
show that
typical
characteriza-
tions of the reasons for
presidential
vetoes
and
congressional
leaders' decisions to
attempt
veto overrides
mask
important dynamics.
Presidents
do not
always
veto
bills with
legislative
goals
in
mind.
In
a similar
vein,
congressional
leaders
may
attempt
veto overrides to
make
public
statements.
In
both
cases,
the
players
in
the veto/override
game
seek to derive
ancillary
electoral
benefits for themselves
or their
party,
and/or
gain
greater bargaining
leverage
vis-a-vis the other branch.
The
typology
of vetoes
and overrides we
develop captures
more of the com-
plexity
of inter-institutional
dynamics
and carries
important
implications
for
future
research on
presidential
veto
power.
First,
the
analysis
casts doubt on the
usefulness
of
using
vote
switching
on
Type
I
vetoes
as a means of
assessing pres-
idential
power
since there are so
very
few
changes
in
legislators'
positions
from
the
passage
of
the
original
legislation
to the override
attempt,
and more
impor-
tantly,
little need for switchers
at all.
Second,
our
unicameral
analysis
of vote
switching
on
Type
II
vetoes shows
that
Presidents
are
generally
adept
at mar-
shaling enough partisan
support
to
halt
overrides
that
begin
with
supermajor-
ity
status
in
the House
on
the
original legislation.
Further
confirming
evidence
with data on the
Senate and
an
examination of
presidential
lobbying
of individ-
ual
legislators
might support
our
claim that Presidents
are most
likely
to
lobby
more
actively
on
Type
II vetoes,
consistent
with Krehbiel's
(1998)
argument
about
the
importance
of
"pivotal"
voters.
Finally, Type
III
vetoes/overrides
chal-
lenge
us to define
better when
Presidents
have
legislative
versus
symbolic goals
in
mind.
Again, analysis
of
vote
switching
is not
consistently
a
particularly
good
measure
of
presidential power.
At a
very
minimum,
future
vote
switching analy-
ses
must
sort
out whether
the President
vetoes
the
legislation
for
symbolic
rea-
sons
with the
expectation
of
losing,
whether
he undertakes
an
extensive
lobby-
ing
effort to
stop
the
override
in
the first
chamber
considering
the
override
(most
often
the
House),
or
whether he adopts
a more
subtle,
bicameral
approach
in an
attempt
to
reduce
consensus
in
the
first chamber and
ultimately
prevail
in
the second.
Future
analysis
might
also
examine
veto threats
using
the
typology
devel-
oped
in
this research.
If Presidents can
predict
the nature
of
passage
coalitions,
veto
threats should be most
potent
on
Type
I
legislation
and
less
effective on
Type
850
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Toward a New
Typology
of
Vetoes
and
Overrides
II and
Type
III
bills.
In
addition,
it
may
be useful to examine more
subtly
the
characteristics
of
members
of
Congress
who switch votes across veto
types
to
supplement
Krehbiel's
analysis.
Focusing
on contextual
and electoral
factors,
such as the President's
public approval
and
legislators'
seat
safety,
in
addition
to
ideology,
holds
great
promise
to ascertain
when
presidential power
to
persuade
and convert members on override
attempts
is the
greatest,
and
why.
REFERENCES
Aldrich,
John
H.,
and David W Rohde.
1997-98.
"The Transition to
Republican
Rule
in
the House:
Implications
for Theories of
Congressional
Politics."
Polit-
ical Science
Quarterly
112:
541-68.
Bond,
Jon
R.,
and
Richard
Fleisher. 1996.
"The
President
in a More Partisan
Leg-
islative Arena."
Political Research
Quarterly
49: 729-48.
. 1990. The President
in the
Legislative
Arena.
Chicago: University
of
Chicago
Press.
Cameron,
Charles. 2000. Veto
Bargaining:
Presidents and
the Politics
of Negative
Power
New York:
Cambridge
University
Press.
Congressional
Quarterly
Almanacs,
various
years. Washington,
DC: Congressional
Quarterly.
Copeland,
Gary.
1983.
"When
Congress
and the President Collide:
Why
Presi-
dents Veto
Legislation." Journal
of
Politics 45: 696-710.
Davidson,
Roger
H.,
and
James
A.
Thurber,
eds. 1995.
Remaking
Congress: Change
and
Stability
in the 1990s.
Washington,
DC: Congressional Quarterly
Dodd,
Lawrence
C. 1979.
"The
Expanded
Roles of
the
House
Democratic
Whip
System:
The 93rd and
94th
Congresses." Congressional
Studies
7:
27-56.
Dodd,
Lawrence
C.,
and
Bruce I. Oppenheimer.
1977. "The House
in
Transi-
tion."
In
Lawrence C.
Dodd and Bruce
I.
Oppenheimer,
eds.,
Congress
Recon-
sidered.
New York:
Praeger.
Durr,
Robert
H.,
John
B.
Gilmour,
and
Christina
Wolbrecht.
1997.
"Explaining
Congressional Approval."
American
Journal
of
Political Science
41: 175-207.
Edwards
III,
George
C. 1989. At the
Margins:
Presidential
Leadership of Congress.
New
Haven,
CT:
Yale
University
Press.
. 1980. Presidential
Influence
in
Congress.
San Francisco:
WH. Freeman.
Herrnson,
Paul
S.,
and
Kelly
D. Patterson.
1995. "Toward
a More
Programmatic
Democratic
Party?
Agenda-Setting
and
Coalition
Building
in the House of
Representatives."
Polity
27: 607-28.
Hoff,
Samuel
B. 1992.
"Presidential
Support
and
Veto
Overrides,
1889-1988."
Midsouth
Political Science
Journal
13: 173-89.
Jones,
Charles
0. 1994.
The
Presidency
in a Separated
System. Washington,
DC:
Brookings.
Kernell,
Samuel. 1997.
Going
Public:
New
Strategies
of
Presidential
Leadership,
3rd
ed.
Washington,
DC: Congressional
Quarterly
851
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
Political
Research
Quarterly
Krehbiel,
Keith. 1998. Pivotal
Politics:
A
Theory
of
U.S.
Lawmaking.
Chicago:
Uni-
versity
of
Chicago
Press.
Lee,
Jong
R. 1975.
"Presidential
Vetoes from
Washington
to
Nixon."
Journal
of
Politics 37: 522-46.
Light,
Paul
C. 1982. The President's
Agenda:
Domestic
Policy
Choice
from
Kennedy
to
Carter
Baltimore:
Johns
Hopkins University
Press.
Neustadt,
Richard.
1960. Presidential
Power
and
the Modern Presidents. New York:
Free Press.
Peterson,
Mark A.
1990.
Legislating Together:
The
White
House and
Capitol Hillfrom
Eisenhower
to
Reagan.
Cambridge,
MA: Harvard
University
Press.
Presidential
Vetoes,
1789-1988.
1992.
S.
Pub. 102-12.
Washington,
DC: U.S. Gov-
ernment
Printing
Office.
Presidential
Vetoes,
1989-1996. 1997.
S.
Pub. 105-22.
Washington,
DC: U.S. Gov-
ernment
Printing
Office.
Ringelstein,
Albert.
1985.
"Presidential Vetoes: Motivations
and
Classifications."
Congress
and
the
Presidency
12:
43-55.
Rohde,
David W 1991. Parties and Leaders
in the
Postreform
House.
Chicago:
Uni-
versity
of
Chicago
Press.
. 1992. "Electoral
Factors,
Political
Agendas,
and Partisanship
in Con-
gress."
In
Roger
H. Davidson, ed.,
The
Postreform Congress.
New
York:
St.
Martin's Press.
Rohde,
David
W, and Dennis
M.
Simon.
1985. "Presidential
Vetoes and Con-
gressional
Response:
A
Study
of Institutional Conflict."
American
Journal
of
Political
Science 29: 397-427.
Sinclair,
Barbara.
1992. "The
Emergence
of
Strong Leadership
in the
1980s
House of
Representatives."
Journal
of
Politics 54:
657-84.
Sundquist,
James
L. 1981. The
Decline and
Resurgence of Congress.
Washington,
DC: Brookings.
Watson,
Richard
A.
1993. Presidential
Vetoes and
Public
Policy.
Lawrence: Univer-
sity
Press of
Kansas.
Woolley,
John
T. 1991.
"Institutions,
the Election
Cycle,
and
the Presidential
Veto." American
Journal
of
Political Science
35:
279-304.
Received:
August
23,
2000
Accepted
for
Publication:
December
21,
2000
rconley@polisci.ufl.edu
kreppel@polisci.ufl.edu
852
This content downloaded from 128.227.239.220 on Mon, 31 Mar 2014 13:53:48 PM
All use subject to JSTOR Terms and Conditions
... There is no question that veto power grants the president influence with Congress in lawmaking, as scholarship on veto bargaining and veto threats focuses heavily on formal theory and predictive models to demonstrate rational behavior in order to develop generalizable explanations applicable across time (Lewallen 2017;Cameron and McCarty 2004;Conley and Kreppel 2001;Cameron 2000). This research provides insight and support for causal mechanisms that can measure influence, a difficult concept in and of itself, as well as what outcomes presidential influence has provided, whether that is in policy agenda setting or policy outcomes. ...
Preprint
Building from Azari and Smith’s (2012) work on informal institutions, we understand the veto bargaining function as informal, operating within the formal rules and constraints of the legislative development process, as there are no formal rules to govern presidential bargaining with Congress. The president’s power to persuade becomes contextual and situational to the issue, individual, and moment in time. The Freedom of Information Act (FOIA) is employed as a case to examine a policy issue that presidents do not want, as it serves as a congressional check on executive power. Examining the development of FOIA, we can examine how and why presidents choose to employ a private bargaining strategy. Using the same policy issue across three administrations – Lyndon Johnson, Richard Nixon, and Gerald Ford – provides consistency in examining the political contexts on an issue each president wanted to avoid but was forced to engage with by Congress.
... One of the variables that would most affect the presence of legislative override is the size of the executive political party in the legislature. So, to the extent that the percentage of seats in the president's party is greater, the chances of a legislative override would tend to fall (Cameron, 2009;Conley, 2002;Conley & Kreppel, 2001;Hoff, 1992;Rohde & Simon, 1985;Wilkins & Young, 2002). In fact, in countries where qualified majorities of two-thirds of the total vote are required for a legislative override to proceed, if the president has a seat of at least 30 per cent it would be almost impossible for a veto to be defeated. ...
Article
This article addresses, in a comparative perspective, the conditions under which legislatures find greater incentives to overcome a presidential veto. The authors' framework points out that the legislative override is used selectively, essentially when the presidential veto is in particularistic bills, and in the context of unstable democracies. To test their theoretical argument the authors used an unpublished database on bills processed in Ecuador between 1979 and 2019. Though theories of veto bargaining predict the override success rates should be 50 per cent, the authors show that, similar to the US, in Ecuador the override success is relatively low. Although the effective number of political parties is a statistically significant variable too, the empirical findings are more consistent with the first hypothesis.
... Chiou and Rothenberg (2017) elaborate further-theoretically and empirically-in a substantively rich, well-crafted book, The Enigma of Presidential Power. On presidential power, see also Conley and Kreppel's (2001) analysis of veto types and veto switching. Deering and Maltzman (1999) use the pivotal politics framework for executive orders. ...
Article
Full-text available
Pivotal Politics: A Theory of U.S. Lawmaking was conceived as an instance of normal science within a nascent but nebulous paradigm. In the two decades following publication of the book, three disciplinary changes are noteworthy. First, the paradigm in which much political science is conducted has become more clearly identifiable. Second, more scholars are working within the emerging paradigm. Third, adherence to the paradigm’s norms of good research practice has become more consistent. This essay explicates four conceptually distinct but implicit components of the paradigm. Labeled the Components of Institutional Analysis, the framework has many uses. Several of these are illustrated with reference to Pivotal Politics and to the subsequent, constructively critical literature. The preponderance of this scholarship is normal science in the best sense. In its entirety, paradigm-guided research on institutions comports well with William Riker’s prognosis for a science of politics and Gerald Kramer’s requirements for scientific understanding.
... Second, empirical studies of vetoes indicate a variety of influences on their use but have not looked at the impact of veto-proof tallies. Researchers have established that vetoes occurred more frequently when divided government prevailed (Cameron 2000), when the president's party coalition in Congress diminished (Rohde and Simon 1984), or when party reputations, blame games, and bicameral differences were involved (Conley and Kreppel 2001;Gilmour 1995Gilmour , 2002Gilmour , 2011. None of these analyses dealt explicitly with the size of a bill's majority as a factor in presidential decisions. ...
Article
Super-majorities have occurred frequently in Congress but have escaped scholarly attention. This paper employs new measures of positive agenda control and a unique data set of 3,407 nontrivial bills from 1981 to 2008 to answer two questions: how did legislative leaders construct veto-proof coalitions, and what did presidents do with them? Legislative leaders, we argue, deployed procedures to expand and sustain veto-proof coalitions, despite increasing polarization. The resulting history, which signaled members’ commitment to a bill, provided information to the president that reduced uncertainty about possibilities for interbranch bargaining and the likely success of a veto. We find that positive agenda control increased the probability of vote tallies of two-thirds or more, especially after the 1994 election. In addition, we demonstrate that presidents concentrated veto activity on bills with outcomes of less than two-thirds, rejected some veto-proof bills for reputational gains, and deployed signing statements strategically. The analysis suggests that congressional leaders paradoxically gained capacity for nurturing large, bipartisan alliances as the institution became more polarized. Moreover, it demonstrates that strategic activity by legislative leaders is critical to explaining variation in presidential options for veto bargaining and signing statements.
Article
The executive veto is one of the most important instruments of presidential power within the legislative arena. Because of constitutional requirements, every regular veto and many pocket vetoes are accompanied by a short letter to Congress from the president describing—often in detail—the reasoning behind the decision to veto. As such, these messages form a unique window into interbranch and partisan conflict over time. While analyses have been conducted on the broad patterns of veto use, questions remain concerning presidential motivations for vetoing as well as how these motivations have evolved over time along with changing political institutions. Using a data set of approximately 2000 unique veto messages from the first in 1792 to the last of President Trump's vetoes in 2021, we examine the content of veto messages to uncover the number and kind of arguments used by presidents when addressing Congress over vetoed legislation. Our analysis uncovers significant variation, with the evolution of arguments within veto messages mirroring broader political trends.
Article
This paper uses assumptions about position taking and credit claiming to help predict when Congress will challenge and override a presidential veto. Using assumptions about position taking and credit claiming to generate measurable hypotheses, I find that vetoes on spending bills are 13.0 percentage points more likely to be challenged and 13.1 percentage points more likely to be overridden. Furthermore, spending vetoes are more likely to be overridden when congressional elections are nearing. The results confirm that the electoral connection not only explains individual behavior, but can also help predict institutional outcomes.
Article
This article uses presidential veto overrides as a site upon which to test whether the context of “political time” acts as a determinant of institutional politics. Using an original database, I transpose political time theory to the study of the veto process. While the concept of political time has mainly been used in qualitative case studies, I argue that it also allows for large N testing. Because political time creates observable measurements, it can be used to explore various institutional outcomes. One such institutional site is overrides. With this in mind, I pit political time against the standard variables in the vetoes literature (e.g., approval, divided government). Finding that political time is a significant determinant of overrides, the article discusses how the theory fits with the existing literature. The article concludes with the broader implications of more widely incorporating political time—as a quantitative independent variable—into the study of American political institutions.
Article
Veto override attempts offer an ideal opportunity to study the president's influence in Congress. A bill's content is identical during final passage and veto override votes, yet some members of Congress vote differently on these two roll-calls. This article focuses on those members of Congress who switch their votes to join, or defect from, the president's coalition during veto override votes. The empirical analysis investigates veto override attempts from 1973 to 2011 to determine which members of Congress are most likely to change their votes during veto overrides. I find that the president's ability to win, or keep, members' support on veto override votes is determined in part by members' electoral incentives and the president's strength in members' constituencies. In particular, the president's strength in members' constituencies makes presidential party members more likely to join the president's veto override coalition and also makes opposition party members more likely to defect from the president's veto override coalition. Accounting for the influence of constituencies and electoral incentives augments party-based and ideology-based theories of congressional behavior and thus helps build richer, more comprehensive, and more accurate theories of member behavior.
Article
Full-text available
span style="font-size: 100%; font-family: Arial; color: #000000;" data-sheets-value="{"1":2,"2":"Previous empirical studies of congressional responses to presidential vetoes have attempted to identify elements which explain whether an override takes place. However, they have failed to distinguish between initial and final reconsideration, and have included private bill vetoes in the analysis. This research employs a presidential support model to examine factors influencing probability of override at both the first and second house juncture, as well as strength of successful final override vote, over the last century. The results substantiate the value of the model for determining legislative reactions to public bill vetoes."}" data-sheets-userformat="{"2":2111744,"11":0,"14":{"1":2,"2":0},"15":"arial,sans,sans-serif","16":10,"24":{"1":0,"2":3,"3":0,"4":3}}">Previous empirical studies of congressional responses to presidential vetoes have attempted to identify elements which explain whether an override takes place. However, they have failed to distinguish between initial and final reconsideration, and have included private bill vetoes in the analysis. This research employs a presidential support model to examine factors influencing probability of override at both the first and second house juncture, as well as strength of successful final override vote, over the last century. The results substantiate the value of the model for determining legislative reactions to public bill vetoes.</span
Article
Full-text available
Despite the large number of analyses of vetoes in recent years, many puzzles persist. This paper treats vetoes as a form of institutional conflict. This conflict is analyzed as being jointly determined by congressional and presidential behavior. The role of elections, partisan conflict, institutional change, and variation in presidential resources are examined as determinants of the likelihood of institutional conflict. Veto conflicts are disaggregated into conflicts over major and minor issues. Minor vetoes are shown to reflect pure institutional conflict and the electoral cycle; their frequency is, substantially, a reflection of the frequency of institutional interaction. Major vetoes reflect a more complex form of institutional conflict; in this conflict, partisanship and presidential resources are central, and conflict is no more likely in election years.
Thesis
The President's domestic agenda is a critical force in the national policy process. Despite recent increases in congressional initiative, the President remains a powerful agent in the distribution of political benefits. This research rests on the analysis of the five most recent administrations: Kennedy, Johnson, Nixon, Ford, and Carter. The study is restricted to the President's domestic agenda--to the exclusion of foreign affairs. The end aim of this research is to isolate those variables which can be said to have an impact on domestic choice. In the effort, the research concentrates on two primary sources of data: (1) 126 interviews with past and present members of the five presidential staffs; and (2) Office of Management and Budget legislative clearance records. This study hinges on a resource definition of presidential power. According to this theory, Presidents are constrained by the level of both internal and external resources. The internal resources involve at least four separate entities--time, energy, information, and expertise--while the external resources focus on congressional support, public approval, and electoral margin. Without these requisite resources, presidential bargaining is not seen as having an impact on policy outcomes. Furthermore, these resources rise and fall over the term of office, creating two distinct policy patterns. The cycle of declining influence appears as time, energy, and congressional support drop, while the cycle of increasing effectiveness arrives as information and expertise grow. Initially, this research was concerned with the basic parameters of the President's domestic program. What does the domestic agenda look like? When is the agenda set? Here, each agenda item can be separated into three components: the issue or problem orientation, the alternative or programatic solution, and the priority or level of White House concern. Using both the staff interviews and the OMB clearance data, it is clear that the president's domestic agenda is set very early in the term and repeated often. The study also asked why