Arms Races, Arms Control, and Disarmament
| Author | Knut Midgaard |
| DOI | 10.1177/001083677000500102 |
| Published date | 01 March 1970 |
| Date | 01 March 1970 |
| Subject Matter | Articles |
Arms
Races,
Arms
Control,
and
Disarmament
1
Knut
Midgaard
I
Arms
races,
arms
control,
and
disarma-
ment
can
be
studied
from
many
points
of
view.
If
we
concentrate
on
the
reasons
and
conditions
for
each
of
these
processes,
we
quickly
come
to
realize
that
a
number
of
factors
are
relevant.
Both
political
and
military,
and
thereby
technological,
factors
should
be
taken
into
account;
furthermore,
the
characteristics
of
relevant
bureaucracies
may
constitute
significant
variables,
etc.
The
present
study
takes
as
its
point
of
departure
two
questions
which
a
state
has
reason
to
ask
itself
regardless
of
whether
it
is
offensively
or
defensively
oriented
and
regardless
of
which
factors
it
altogether
takes
into
consideration
when
appraising
its
military
needs
and
possible
choices:
(i)
Given
that
states
Ai,
A2, - - -
An
have
such
and
such
military
forces
and
armaments
in
situation
S,
what
type
and
how
large
military
forces
and
armaments
should
we
have?
(2)
Given
that
we
have,
in
situation
S,
such
and
such
military
forces
and
arma-
ments,
what
type
and
how
large
military
forces
and
armaments
will
each
of
states
A,,
A2, - - -
An
feel
a
need
for?
It
is
likely
that
most
states
are
immediate-
ly
more
inclined
to
deal
with
problems
of
the
first
type
than
with
problems
of
the
second
type,
and
that
is
perhaps
part
of
the
reason
why
armament
occurs
more
readily
than
disarmament.Whether
or
not
states
in
a
given
interaction
have
such
an
inclination,
it
is
important
for
those
who
wish
to
understand,
predict,
or
influence
their
armament
policies
to
have
a
theo-
retical
grasp
of
the
implications
of
asking
the
two
types
of
question.
The
pioneer
of
precise
and
systematic
studies
of
the
interaction
between
arma-
ment
policies
was
the
mathematician
and
physicist,
Lewis
Richardson,
who
in
the
inter-war
years,
worked
out
mathematical-
ly
formulated
hypotheses
about
the
inter-
relation
between
level
of
armament
of
states
or
groups
of
states,
and
who
tested
his
hypotheses
on
the
arms
races
preceding
each
of
the
two
World
Wars.2
2
His
basic
hypothesis
can
be
expressed
thus:
Given
two
states
A
and
B,
the
speed
with
which
A’s
armaments
increase
at
a
given
point
in
time
is
proportional
to
B’s
level
of
armament
and
the
grievances
A
has
against
B.
It
is
inversely
proportional
to
the
level
of
armament
A
has
already
reached
(since
both
security
and
the
eco-
nomic
burden
can
be
assumed
to
increase
with
one’s
own
level
of
armament).
The
same
applies
to
B.
We
cannot
discuss
here
under
what
conditions
this
hypothesis
is
reasonable.
What
we
shall
draw
attention
to
is
the
following:
Richardson’s
hypothesis
pre-
supposes
that
the
states
in
question
react
immediately
or
instinctively
to
the
existing
situation.
They
do
not
reason
strategically
about
the
interaction
which
their
possible
armament
decisions
constitute.
Neither
of
the
states,
in
a
given
pair,
takes
into
con-
21
sideration
how
the
other
will
react
to
its
decisions
-
and
in
connection
with
that,
what
strategy
is
the
most
profitable,
every-
thing
considered.
Although
such
instinctive
reactions
can
play
an
important
role
in
arms
races,
Richardson’s
assumption
entails
a
strong
limitation
of
relevance.
In
his
book
Conflict
and
Defense
(1962)
the
economist,
K.
Boul-
ding
introduced
a
significant
modification.3
At
the
same
time
as
demonstrating
the
fruitfulness
of
Richardson’s
model
in
a
number
of
areas,
he
introduced
actors
who
were
able
to
make
rational
choices
on
the
basis
of
preference
functions
or
indif-
ferent
curves,
i.e.
on
the
basis
of
long-
term
calculations,
and
who
have,
there-
fore,
the
possibility
of
attaining
more
advantageous
results.
He
pointed
out
the
parallelism
between
arms
races
and
price
wars.
The
importance
of
this
modification
or
extension
of
Richardson’s
model
was
strongly
emphasized
by
Thomas
C.
Schel-
ling
in
his
review
of
Boulding’s
book
in
World
Politics.4
Schelling,
moreover,
had
already
made
use
of
similar
analyses
in
The
Strategy
of
Conflict
(1960).
Through
his
lectures
in
the
autumn
of
1961
(from
which
the
present
author
benefited),
he
inspired
Colonel
Glenn
A.
Kent
to
write
a
monograph
’On
the
Interaction
of
Op-
posing
Forces
Under
Possible
Arms
Agree-
ments’
in
which
he
assumed
that
the
actors
based
themselves
on
strategic
considera-
tions
concerning
the
armament
game,
i.e.
the
interaction
which
is
constituted
by
their
possible
armament
policies,
each
of
the
states
pursuing
objectives
which
are
relative
to
the
opponent’s
strength.5
Kent
investigated
the
likelihood
for
stability
under
different
quantitative
limitations:
a
limiting
of
the
number
of
missiles,
of
the
yield
of
warheads,
and
certain
com-
promises.
A
later
significant
contribution
which
is
directly
linked
to
Richardson’s
and
Boulding’s
models,
is
Martin
C.
McGuire’s
book
Secrecy
and
the
Arms
Race
(1965).6
McGuire
develops
economic-strategic
models
for
the
analysis
of
the
nuclear
arms
race;
and
he
pays
special
attention
to
the
positive
or
negative
value
that
secrecy
can
have,
under
various
conditions,
on
the
state
in
question
itself
and
also
on
its
counterpart.
McGuire’s
book
also
contains
a
survey
of
the
arms
race
literature.
A
recent
contribution
which
helps
to
locate
Richardson’s
and
Boulding’s
approach
in
a
wider
mathematical
context
is
T.
L.
Saaty’s,
AIathematical
Models
of
Arms
Con-
trol
and
Disarmament
(i968).’
7
This
article
does
not
contribute
any
new
theory
of
significance.
Its
independent
contribution
consists
in
the
attempt
to
proceed
gradually
in
a
non-technical
man-
ner
from
the
simple
to
the
complex,
and
furthermore,
at
each
step,
to
make
clear
the
limitations
and
scope
of
the
assumptions
made;
finally
it
consists
in
some
game-theoretic
considerations.
Our
subject
is
the
armament
policies
of
states.
We
shall
discuss
how
one
state’s
armament
policy
can
be
influenced
by
another’s
actual
or
estimated
armaments,
armament
requirements,
and
armament
policy,
the
former
state
realizing
that
the
latter’s
armament
policy
is
or
can
be
in-
fluenced
in
a
similar
way.
Our
analyses
will
be
related
to
graphic
representations
of
the
two
states’
arma-
ments
and
armament
requirements.
Thus,
we
make
the
assumption
that
a
country’s
possible
armaments
can
be ranked
on
an
ordinal
scale.
This
means
that
when
con-
fronted
with
two
possible
instances
of
armaments
we
can
decide
whether
one
represents
a
higher
or
a
lower
level
than
the
other,
or
the
same
level.
Indeed,
ratio
scales
are
introduced.
22
As
a
measure
of
capability,
or
the
size
of
armaments,
authors
have
used
the
actual
armament
budget,
the
number
of
missiles,
the
yield
measured
in
megatons
TNT,
and
similar
items.
Such
measures
are
too
simple
unless
other
central
variables
are
taken
as
given.
In
a
speech
about
the
dynamics
of
nuclear
strategy
given
in
September
1967,
the
then
Secretary
of
Defense,
Robert
S.
McNamara,
said:
...
the
most
meaningful
and
realistic
meas-
urement
of
nuclear
capability
is
neither
gross
megatonnage
nor
the
number
of
mis-
sile
launchers,
but
rather
the
number
of
separate
warheads
that
are
capable
of
being
delivered
with
accuracy
on
individual
high-
priority
targets
with
sufficient
power
to
destroy
them.
8
Others
have
gone
even
further
in
speci-
fying
different
variables
that
are
indicative
of
a
state’s
power;
they
have,
for
example,
introduced
a
state’s
capability
to
acquire
relevant
information
about
its
counterpart,
and
to
use
it.
In
spite
of
the
actual
com-
plexity
it
can
nevertheless
be
useful
to
introduce
models
which
only
take
care
of
quantities
that
can
be
measured
fairly
easily.
The
states’
assumptions
about
each
other’s
armaments
and
armament
requirements
as
well
as
their
assumptions
about
such
as-
sumptions,
and
so
on,
are
important
variables.
In
section
II
we
shall
make
the
presumption
that
the
states
are
known
to
be
fully
acquainted
with
each
other’s
armaments
and
armament
requirements,
or,
put
another
way,
that
the
states’
armaments
and
armament
requirements
are
known
to
be
known.
This
means,
first,
that
the
states
in
question
know
each
other’s
armaments
and
arma-
ment
requirements;
second,
that
there
ex-
ists
no
doubt
that
such
is
the
case;
third,
that
there
is
no
suspicion
of
such
doubt;
and
so
on.
The
assumptions
we
have
made
above
are
rather
special,
and
it
might
be
objected
that
central
types
of
situations
will
be
ignored
or
distorted
in
so
far
as
analyses
are
based
on
them.
First
the
armament
policies
of
two
interacting
states
may
partly
be
due
to
considerations
related
to
a
third
state.
Moreover,
the
difficulties
of
getting
disarmament
under
way
have
part-
ly
been
due
to
the
fact
that
actual
or
pos-
sible
armaments
have
a
number
of
compo-
nents,
and
that
the
participants
to
disarma-
ment
negotiations
have
found
it
impossible
to
agree
on
how
much
military
force
one
distribution
over
these
components
repre-
sents
relative
to
another
distribution.
It
should
be
noted
here
that
a
distribution
Mi
over
the
different
components
of
a
state’s
armaments,
can
be
more
effective
than
another
distribution
M2
in
relation
to
a
distribution
Ni
over
the
components
of
the
potential
opponent’s
armaments,
but
that
M2
can
be
more
effective
than
M1
in
relation
to
a
second
distribution
N~
2
over
the
opponent’s
components.
In
such
a
case,
we
cannot
simply
say
that
M,
represents
a
higher
level
of
armaments
than
M2,
or
vice
versa;
the
various
possible
armaments
cannot
simply
be
placed
on
an
ordinal
scale.
A
third
objection
is
this
one.
States
often
-
or
in
general
-
are
un-
certain
or
are
mistaken
about
important
qualitative
or
quantitative
aspects
of
the
armaments
of
other
states,
and
similarly
about
their
armament
requirements.
New
technology
-
or
the
possibility
of
new
technology
-
is
likely
to
cause
uncer-
tainty
and
errors.
The
quality
of
organiza-
tions
can
also
be
afflicted
with
considerable
uncertainty.
A
fourth
objection
runs
as
follows.
A
state
often
cannot
be
considered
a
rational
player
with
a
consistent
con-
ception
of
the
situation
and
a
consistent
structure
of
preferences.
The
complexity
of
a
state’s
decision-making
structure
may
be
considerable,
e.g.
a
decision-making
To continue reading
Request your trial