·

26 Alliance Formation and Cohesion

Alliance Formation and Cohesion 27

TABLE 2. I Threats level of internal threat provided the necessary conditions for cooperation to
and Cohesion continue.
Low internal threat High internal threat Alliances that confront low levels of internal threat and low levels of external
Low External Moderate or low cohesion; threat provide the necessary, though not sufficient, conditions for cohesion to
Threat Low or no cohesion;
depends on which (internal tethering alliances. burgeon. Insights from institutionalist theory are relevant here. In the absence of
or external) threat is higher; conflicts of interest, common values, institutions, and goals may flourish. Examples
usually hedging alliances.
In these cases, insights here include the Partnership for Peace initiatives, which were undertaken with an eye
generated from liberalist to heightening transparency and deepening cooperation between NATO and the
theory will hold.
High External Partnership for Peace signatories.
Moderate to high Cohesion difficult though
Threat cohesion; balancing Alliances that have low levels of external threat and high levels of internal threat
not impossible to attain;
alliances. In these cases, depends on which (internal are pure tethering alliances. States have combined in order to reduce conflict rather
insights from realist theory
will hold.
or external) threat is higher. than to aggregate their capabilities. Conflicts of interest among member states will
These may be tethering,
inhibit them from effectively coordinating joint goals and strategies for attaining
balancing, or bandwagoning
alliances. those goals. In these cases, since it is the competing interests that bring states
together in the first place, cohesion will be difficult if not impossible to achieve. The
two Leagues of the Three Emperors, concluded by Germany, Austria-Hungary, and
the levels of external threat. The liberalist ideas imply that looking at the level of Russia in 1873-78 and I 88 I- 87, were alliances of this type. Austria-Hungary and
threat within an alliance is also important. Indeed, by looking at both, more Russia had intense rivalries in the Balkans, and the states formed these alliances in
nuanced propositions about alliance cohesion can be generated. order to manage and contain those conflicts of interest. The overriding objective of
The most straightforward way to analyze the relationship between internal and these alliances was to keep the peace among the signatories. The level of internal
external threats and alliance cohesion is to construct a two-by-two matrix. threat was quite high. There were no common enemies, so a low level of external
Additionally, the alliance types spelled out in the preceding section c_an b~ threat prevailed. The cohesion of both of these leagues was very low-throughout the
categorized accordingly. Table 2. I illustrates the hypothesized relationships duration of both leagues, the states were completely incapable of forming joint goals
between threats and cohesion.
and strategies for attaining those goals.
Alliances that confront a low level ofinternal threat combined with a high level of
external threat are pure balancing alliances· cohesion will be relatively e~ to
achieve. External threats do, in these' cases, provide member s_tates ':1th a reason
for being and continuing. An example of a balancing alliance is the Franco-Russian
Alliance, formed to counter the threat of Germany, Austria, and Great Britain. Possible Alliance Combinations
In cases that have high levels of external threat as well as high levels of internal
threat, states have combined to target a specific menace but are also t?reatened ~y Alliances are bilateral or multilateral agreements to provide some element of
each other. Cohesion will be difficult, although not impossible, to attain. These are security to the signatories. States have different reasons to enter into them. Even
alliances such as the Allied Powers during the Seco~d _World War. The threat of
when we restrict the examined motives to conditions of varying threat, We must still
the opposing coalition eclipsed the threat withm the alliance, but as soon as the
consider a range of alliance motivations. Some of the motives specified above are not
Axis Powers were defeated the wanime coalition fragmented. Indeed, this is
precisely why wartime, alliances often come to an end with the termination of a compatible; others require reciprocity; still others may be paired with similar
war-not simply be- motives. Further, the commitment level of the agreements emerges from the
cause the external threat d b al b motivations bringing states together in alliance.
. ecreases ut so ecause there are often internal
conflicts that undermine th allia 0h In this section, I outline the possible alliance combinations-those motives that are
e nee. n t e other hand, an alliance such
as NA1'0 could endure beyond the end of the Cold War because the low compatible and likely to emerge. I do so for only dyadic relations, since any
multilateral agreement can be reduced to the dyads of which it is composed. It is
actually most useful to do so, since a state entering a multilateral alliance may have
certain motives toward one state and different mo-
State A's motives State B's
motives
28 Alliance Formation
Hedge Hedge and Cohesion
Alliance Formation and Cohesion 29
Hedge Balance
tives in regard to another. For example, Russia had balancing motives a~ainst
TABLE 2.2
Hedge BritainBandwagon
Great when it formed an alliance with France in 1891/94, but it had Possible Dyadic Alliance Motivations under Conditions of Threat
tethering motives toward Great Britain in I 907 when it signed the AngloRussian
Convention. Hedge
Tether These two agreements, together with the Entente Cordiale between
France and Great Britain, were the treaties that formed the Triple Entente. So
Tether Tether
within the Triple Entente, Russia had balancing n:io~ives toward one ally and
tethering ones
Tether toward the other. I also offer predictions for the commitment levels
Balance
that will be manifest in the agreements. Commitment levels are driven by states'
motivations Balance
Balance to join an alliance and then the bargaining that takes place between the
signatories.
Balance
States hedging will invariably seek low commitment level agreements;
Bandwagon
balancing states will generally prefer high commitment levels. Tethering and
bandwagoning states will have variable preferences, depending on the threats they
are attempting to contain.
Under conditions of threat, two states may form an alliance to hedge their bets,
seeking low-level commitment agreements with potential friends and enemies. The
mutual threat level will be low to both and result in an alliance that has a low level
of internal threat. One state may seek to hedge while the other balances. These
alliances will also have a low level of internal threat, but the states will have
divergent goals, and cohesion will not flow automatically, although it should not
be too difficult to attain and maintain. The states will not have opposing or
conflicting or even incompatible aims.
One state may seek to hedge, while the other bandwagons. In these asymmetrical
alliances, the latter power is threatened by its ally, while the former is not. Terms
will be dictated by the hedging state, and cohesion will follow via coercion rather
than cooperation. Similarly, in a balance/bandwagon alliance, the threat will
emanate for one state from within the alliance, and terms will be dictated by the
stronger power. The primary difference between a hedge/bandwagon and a
hedge/balance alliance will be that the former agreement will have a lower
commitment level than the latter.
Tether/tether relationships will have high levels of internal threat and very low
cohesion. States form these alliances to manage their partners, not for coherent
external purposes. These are alliances that serve to keep the peace among the
signatories. Balance/balance alliances, in contrast, will have very high levels of
external threat and very high cohesion. States form these alliances for clear
external purposes. As long as the external threat remains, the alliance will endure
as well. Alliances that have a low level of internal ~h~~at in addition to a high
level of external threat may endure beyond their trutlal purpose, while those with a
high level of internal threat as well as an even higher level ofexternal threat may
collapse once the external threat has receded. Again, this framework helps us
understand why some wartime al-
 Characteristics of alliance

Low commitment level; modest cohesion; low

internal threat
Low to moderate commitment level; limited
cohesion; low internal threat
Low to moderate commitment level; limited
cohesion; threat within alliance to State B
Low commitment level; moderate to intermediate
internal threat; low cohesion
Variable commitment level; low or no cohesion;
high internal threat
Variable commitment level, moderate internal
threat, low to moderate cohesion
High commitment level; high cohesion; high
external threat
Variable commitment level, though probably high;
moderate cohesion; internal threat to B is
high, external threat to A is high

· · duplicate
NOTE: For simpliciry, · outcomes ave h eenbropped e .. , d ( g Balance/Hedge
. and
. , . .Hedge/Bal-
. .
ance do not both appear). The Tether/Bandwagon combination drops out, smce tethering IS reciprocal -
both sides need to be of the same approximate power level. Bandwagon/ Bandwagon drops out, two states
will not bandwagon with each other-bandwagoning requires capitulation of one to another. The latter state
will therefore have other motives, i.e., hedging or balancing.

liances come to an end once the war is over and others endure. It illustrates why it
is important to look within the alliance at t~e thre_at level an~ not just at the level of
external threat. A summary of dyadic motives for alliance under conditions of threat
is offered in Table 2.2.

Summary of Propositions: Alliance Formation

HEDGING
and Alliance Cohesion
Formation At low levels of threat, states will mix their strategies by seeking
low-level commitment agreements with potential friends and e~emies. They will
do so with an eye to consolidating their power, blockmg
· fpaten
avenues of expansion or t err h · ra r ti I ivals while simultaneously .off
· h · ti
seekmg to curry favor to ensure t eir ac tons ar e not overly provocative.
· · fh d · /h d · alliances will be modest since

ohesio« Cohesion :o e gmg e gmg 1

C · ·· hi h agree let alone
strategies
h for
w re states can ,t ere. is no driving goal. on
attaining those goals. Hedging/balancing and hedgingv'bandwagoning
 30 Alliance Formation and Cohesion
Alliance Formation and Cohesion 3 r
alliances will also have limited cohesion, since each state will enter the
agreement with divergent objectives. The level of internal threat within the Peacetime versus Wartime Alliance Dynamics
latter type of alliance may frustrate cohesion. The aims of the alliance will be
dictated by the hedging state. Although this book is principally concerned with the dynamics of alliances -
formation and cohesion-during peacetime, it also seeks to answer a very important
TETHERING and all too often ignored question of how those dynamics are transformed during
wartime. It is somewhat surprising how thin the literature is on the dynamics of
Fo'!""tion At relatively high levels of threat, states will seek to ally with alliances during wartime; scarcer still are the studies that contrast those dynamics
their adversaries to with those at work during times of peace. Yet the complicating processes of military
mana th · h ili" . conflict inevitably alter the peacetime functioning of alliances in important ways.
ge err ost ty and conflicts of interest.
Cohesion The coh · fh· . . Further, no study of alliances is truly complete without attention to the ways in
. esion o tet enng alliances will be very low. States which wartime alters the theoretical relations that otherwise hold. This is especially
h will have entered these all'
. true for this study, as the alliances of the period are frequently viewed as the cause of
iances as a consequence of their differences, and
t ey will undermine th · fh the war itself. I therefore extend the empirical analysis through the First World War
. e capacity o t e members to agree on goals and
strategies for attaini th al Th in Chapter 6, assessing the ways in which threat affects the cohesion of the alliances
ng ose go s. e level of threat internal to the alli-
ance may be quite high fu h d during wartime.
..
, rt er un ermmmg members' attempts to agree on goals and strategies. The dynamics of peacetime and wartime alliances are quite different.
BALANCING
Questions of burden sharing become far more intractable during wartime than
peacetime, as the stakes are greater and the resources dearer. Division regarding
Formation At relatively hi h h
th I tg t reat levels, states will seek to secure resource allocation arises from the specificity and immediacy of the threats that
emse ves by forming all'
h iances to counter the state or states that states confront during wartime. During periods of peacetime, states can use promises
most
for bargaining leverage with their partners; they can also use promises to enhance
t reaten them.
relationships with their partners. During wartime, however, action is what really
Cohesion Balancing alliances ill b matters. Once war comes, states may not make good on the promises they have
made; commitments may become unrealistic or unfulfilled. It will thus be difficult to
. .
sustain cohesion under these circumstances.
agreeme ts . h . . w e qurte cohesive as states enter these
strategy f Wit . ~1milar goals, although there may be differences over Using the logic of the peacetime arguments in regard to external threats, we
or attairung these goals I bal . would anticipate alliances to be cohesive when they are losing-and threats are higher-
ship al will b . · n ancmg/bandwagoning
relation- and less cohesive when they are winning-and threats are lower. Empirically, the
s, go s e dictated by th b l ·
wa · . relationship between the cohesion of wartime alliances and the threat level states face
e a ancmg state. The threat to the band-
BANDWAGONING gorung state by Its partner f is more complex. The symmetry of threats confronting alliance members, the clarity
agree on al d . may rustrate cohesion, but the states may
go s an strateoies to tt · h of the threats involved, and the commonality of threats facing the signatories will
state. a am t em as directed by the balancing
Formation At very high i::r
1 1 play an important role in determining wartime alliance cohesion. Threats that are
they will seek to all . h e~e s of threat, when states' survival is at stake,
y Wit t e most threatening state. complementary during peacetime-emanating from a rival coalition, for example-may
Cohesi h be divisive during wartime as they culminate in different fronts and result in
strategies f b
on
.
T e goals and
disagreement regarding resource distribution and allocation. Further, during
divergent Because b d . 0
a andwagonmg alliance will be
· an wagonmg alli peacetime, if one state in an alliance is threatened more than another, it might not
one state to the oth h . ances represent the capitulation of
er, co esion may . t . affect cohesion at all. During wartime, however, it may once
This is true of hedgi lb d . exis , yet It may also be coercive.
alliances. ng an wagoning and balancing/bandwagoning
 32 Alliance Formation and Cohesion
Alliance Formation and Cohesion 3 3

more lead to division over resource allocation and undermine the cohesion The final case study chapter of the book thus explores the similarities and
of the alliance. Finally, ambiguous threats during peacetime may actually serve differences between peacetime and wartime alliances. In particular, I investigate
to enhance alliance cohesion-they can be manipulated to serve the interests of the ways that the degree of symmetry, clarity, and commonality of threats to the
the member states. During wartime, however, ambiguous threats are likely to alliance members affect cohesion during the alliances' wartime years. I also
culminate in disintegrative forces. Since states during wartime must agree on
examine whether the level of threat internal to an alliance during wartime
how to concentrate their war-fighting resources, ambiguity may give rise to
undermines cohesion in the same way it does during periods of peace. I do so by
friction. If states perceive or interpret threats differently, then they will not
contrasting the experience of the peacetime alliances with their wartime
agree on necessary action to confront those threats. Hence th~ way threats are
counterparts.
perceived-their symmetry, clarity, and commonalitywill ~ affect alliance
cohesion during wartime very differently than during peacetime.
. Assessin~ wartime cohesion is also more complex than assessing peace-
Methodology: Concepts and Cases
ti~e cohesio~. Employing a behavioral conceptualization of cohesion implies
It is important to strive for methodologically rigorous tests for the theories
that alliance members fighting together during wartime must have s~me
we propose. It is similarly essential that we not sacrifice the richness that the
element of cohesion. In order to tailor the behavioral conceptualiza-
tion of cohesion to p · d f · . . empirical world offers us. 61 Qualitative analysis facilitates our ability to yield
erio s o wartime, I focus m the wartime chapter on effective insight into the power of our theories to explain the world. Using a
three ce~tral aspects of coordinated alliance behavior. First, I assess the straightforward case study approach to testing the theory I propose in this
level of cohes~on according to the capacity of the member states of the chapter, I provide a diplomatic and military history of each alliance and examine
alliances the theoretical linkages between the variables. I use process tracing in order to
to coordmate their war-fi ht' Th' . . . establish the causal relationships between the independent variable, threat level,
g mg strategy. rs rs a straightforward extension
of the peacetime analy · Th d and the dependent variable, alliance formation in the first place and alliance
. sis. e secon component of wartime alliance
cohe- cohesion in the second. In order to provide both richness and rigor, I examine in
sion that I examine is the bili f b .
a i ity o mem er states to agree on war aims. This depth the theoretical ideas in light of the empirical material for eight cases. I
element as with the fi t · · htf
. ' . trs , 1s a straig orward extension of the analysis of have conducted multicountry archival research in addition to mining the
co-
h hesion durmg peacet' I h . secondary literature in order to present thorough, structured, focused
ime, n t e peacetime cases it is important to explore
a t e states' abilities to agrr al I h ' .. comparisons of all eight cases.62
. ee on go s. n t e context of war, those principal
go s are allied war aims Th· d ffu d .
r. . · 1r , o n amental importance to coalition war-
h iare-and 1 alliance cohesio · h bili
.
h t ea tty of a wartime alliance to prevent
n-1s CONCEPTS
t e cone usion of a sep t Th' .
. .. ara e peace. 1s ts, therefore, the third
element of THREAT Threat is an inherently perceptual concept. While capabilities play
co esion I will mvestigate.
I have emphasized the i an important role in determining what is deemed threatening, the other essential
f ingredient is intentions. A threatening state has both the capability and the will
d to do harm to some important interest. Gauging threat thus has to do with
. mportance o threat to explain peacetime alliance
ynanucs, and I maintain th . . identifying those states that have the capacity to undermine one's interests and
di d b .
e same is true in wartime. Threats, however, are the perceived desire to do so as well.
me iate Y different fore d · ·
war co h es unng wartime than during peacetime. While decision makers have a tendency to infer intentions from capabil-
Once ities, it is clear that they do not do so uniformly. The degree of threat is nev-
mes, t reats loom large th c.
tegrity d . r an ever ror states, as states' territorial in- ertheless a function of capabilities most essentially. While it is critical that the
an even survival a t . k Yc
external th d intention to do harm exists, the level of damage that could be wrought is a
re a ns · et we know that this high level of
reat oes not au tom ti 11 1. . function of capabilities. Capabilities concern not only military expenditures,
wartime alli a ica Y cu nunate m cohesion; if it did, all
ances would be cohes·v hi d technological prowess, industrial capacity, population, and size of the
empirically is n t p i e-somet ng we know intuitively an
0
true. art of the re c hi · .
sist within alli ason tor t is rs that threats may still per-
an ance. Although t al h
ally sigru'ficantl I h ex em t reats generally loom larger, usu-
y arger, t an the int al
tend to the question of wh em ones, we n~ed nevertheless to at-
during wan. ,..., ftus ether the level of threat internal to an alliance
-ue trates cohesion . d .
as it oes dunng peacetime.
 rity promise levels. Ranging from the least to the most committed, the cat-
egories are: 66
36 Alliance Formation and Cohesion
. . ality in the event
Alliance of war (I);
Formation and Cohesion 3 7
34 Alliance Formation and Cohesion mamtam benevolent neutr •. .a. promise
. h to . li-
and coordinate accordingly, cohesion will be interpreted as high. In general te? Alliance
a promise to consult in the event of ilitary
·Conclusion Formationwit
rm hostilities andan imp 3
Cohesion
11_5, when states agree on mutual goals and strategies toward those ends, I will 5
military but also proximity. During the period under review, proxi~ty has to cation of aid (2); id . t of war but uni-
mte~ret cohesion as high. When states cannot come to any agreement ove_r th~ir
do with the closeness of the two countries and with their colomal ho~dings. Threat ..
The theoretical ·
framework thatndIother a1 mis even
propose based on '
a number of impor tant
goals and strategy, or if they undertake policies that undermine assistance a •. .promises di .of military
s specified (3)-
grows as military expenditures, military size, population, and mdustrial and insights in the literature. Instead ofd viewing these propositions as competing, I have
their allies I will interpret h · I lateral and without pre-prepare or explicit con ition all d r specific '
technological
. prowess grow and/or as intentions be~om~ more belligerent. This outlined an argument that allows us to see those insights as complementary. I
' a promise to come to the active · · t ce
assis anof an Y un e
threat is exacerbated statecoin
when the sessments esion as ow. In
question ts the
will be made
wartime cases these as-
very
in dclose or hits , colonial examine alliance behavior under conditions of threat and propose a curvilinear
. . power. .' circumstances (4); . ofjoint aplan-
holdings .are close to one's own. The regar to tdata for military
e states expenditures,
coordmat1on military
of military strat- relationship between threat and alliance formation. Ih tpresent unifying theoretical
egy,
personnel, iron and steel production, their
ener~ agreement on · dh. . · . an
. uncond1tiona
1 . promise
f utual assistance,
O
m s or
war aims, an t err prevennon of a separate peace. framework of alliance formation that incorporates the traditional view but also
CASES
. . d f om J David ning, with
explains the division
anomalies of forces (5);ignored by
that are . that .approach.
h vent of attack
I argue that different
alliance behaviors emerge · f tual assistance in t e e Those behaviors include
consumption, and urban and total populat10ns are generate r · • an unconditional promiseat mu
O varying levelsd of threat.
h integration of forces
Singer and Melvin Small's study of National Military Capabilities from the hedging, d control
tethering, an balancing,
t e i with preplanned comman an
and bandwagoning.
Correlates of War project. 63 I also use their classifications to define the great One implication of the theory presented here is that statesf confront
powers of the era. The great powers' border data -both of their homelands and
and strategy ( 6). 11 in external
threats, as the traditional view suggests, but they also confront threats internal to
colonial holdings-I glean from Siverson and Starr.64 I qualitatively assess the . . h romised response o an a y .
alliances. By examining both, a better understanding Each of alliance
gradation cohesion is
represents an mcrease 111 t e P ity level the alli-
historical record as well to determine which states others found threatening and achieved. The motivations. that draw . states tates together
what secun in the first place determine
which states others considered to have hostile intentions. the event of war. Each cohesion d sectt~n 5. The first two levels represent
which of the following will be most pronounced:
ance meets, according to the above cntena. · d te commitment; the the level of threat within the
The level of threat, therefore, is a function of the following: alliance or the level of threat external resent to it. The level and source of the threat will
mo era
· military expenditures have a profound effect on whether the signatories are two
ablereto agree on goals and
make good on low commitment; the secon lways
· military size strategies to attainhithose goals.. · t States
do not a Further, by drawing . all' out the implications of alliance t ird formation for1 cohesion, we
· industrial resources in terms of iron and steel production and in comrmtmen · . 1 rly in tethering -high
two represent
also begin to understand another d important dimension of alliances-their
. hi hl likely that parncu a functioning
terms
P
during wartime. An implicit assumption underpins much ofInthe
t ese promises. alliance
fact, it rs ig literature-the
Y1 7'
of urban population h assumption that its arguments and findings are universally valid. Yet
· total population ances, the obligations willwe notknow
be
intuitively that wartime dramatically transforms the context of state interaction,
· proximity of the homeland honored. 6
. • f alliance
If there is growth in any one area, it is characterized as an increase in threat. which most assuredly will have an effect on the dynamics of alliances that hold
· proximity of colonial holdings during peacetime. The prevailing conventional ioral
h conceptualization
wisdom, thato alliances
. are more.
A reduction in any of these categories is a decrease in threat. Fundamental to the
· will intentions cohesive the higher the external threat, h implies that wartime alliances should always
analysis here be the ratio of external threat to internal threat that a state is h b ALLIANCE COHESION I adopt a be av
experiencing. The book's Appendix contains all of the above data for the six great . 1be
h cohesive -again, something we know is notreristics. true. I contrast
Cohesion theisfactors that
· es socia c arac
facilitate cohesion in peacetime with those that affect cohesion in wartime d in more
powers of Europe in the period under review. cohesion rather than one that emp asiz
and tactics, an co- explicit the similarities and differences that operate under these
depth, making
ALLIANCE FORMATION An alliance is defined here as any formal or in fonnal " oncontexts
is indifferent
contrast goals
to . is .strategy, L' ka t eThis
' of interaction. _a ility of member
comparison states to agree
of peacetime ,, ~ 8 This
and wartime
h ordtnate activity directed toward those ends. In my view, cohes10n is
agreement between two or more states intended to further (militarily) the national alliances provides a more comprehensive understanding 69 of alliance dynamics. I tum
security of the participating states. It is a continuing security association among Who distinguishes between cohesion and efficacy. f able
now to the testing of the theory by analyzing eight cases in the next four chapters. to discuss co-
member states with an element of forward planning and understanding to aid Alli . . It might be pre er not analytically1the .-
same as duration.
un Id be inherent Y
member states militarily or through benevolent neutrality. While there are a . . it as such wou .
it ble alternative.
number of different commitment levels that can be identified, adopting as broad a k esion
.
as a "we feeling," but assessmg i .
.. . 1
.
eptron ts a
definition of alliance as possible will be the most powerful conceptually and sut a pressionist1c.
. lli-
Adopting a behaviora cone t of attent10n as a
analytically. Thus, alliances are formed when two or more states conclude an · d the
agreement to advance their mutual security via an implicit or explicit ~greement to same amoun i: t' on
come to the other's aid or to maintain benevolent neutrality in the event of war. 65 I ance cohesion
not receive . alliance 1ormahas
i
identify six different secu- ts
f
·
 The cases that I will · ·h
L examme m t e next three chapters are the two
eagues of the Three Em b
R . fc . perors, etween Germany, Austria-Hungary,
and
uss1a, ormed m 1873 d ..
A an agam in 1881; the Dual Alliance between Aus- .H
tria- ungary and G f .
.H
ermany o 1879; the Triple Alliance between Germany,
usrna- ungary a d It 1 f
ad h T. 1. ' 11 ayo I882;theFranco-RussianAllianceof1891/94;
n t e np e Entente betw F R·
in 1 1 h . een ranee, ussia, and Great Britain concluded
ti · t en exanune both the Central Powers and the Triple Entente
907

rorn 1914 to 1918 in order to h 1· · ·

th' h assess t e va idity of the propositions
raised m
is c apter as well as to co t hd .
pe . . n rast t e ynarrncs of the alliances during their
aAcet1me y_ears with their experiences during wartime.
s mentioned in the pr · h
alliance f h evious c apter, I have elected to examine the
s o t e pre-World W: I d Wi 1
reasons First th ar an or d War I period for a variety of
· , ese are cases that h di · 1
sistent with' bal f ave tra Itlona ly been understood as con-
ance o power theo d
liances under · . . ry an are powerful examples of how al-
nune mternat1onal t b ili F h
tively long du ti . h 5
a 1 ty. urt er, these are alliances of rela-
ra ion Wit vari ti ·h .
will allow rn t 1 a ion mt e level of cohesion over time. This
e o exp ore wheth th 1 1 f
tion with the le 1 f h . er e eve o threat also varied in
conjuncve o co esion whil0 h Id"
planatory variables ibl ' ~. mg constant as many other ex-
as possi e In add t 1 hi .
her of alliances did · ion, t is is a period in which a num-
1 emerge, which ·11 II
threat that emanated b wi a ow me to examine the level of
etween the memb .
gether of the states · . . er states pnor to the coming to-
m question Fmall th b
me to contrast how th a11· · Y,. e out reak of war in 1914 allows
e iances functio d d . .
wartime operations. ne urmg peacetime with their
While the Europe of the past is dr
. .
today, in the conclusion I will amat1cally different from the Europe of
hold in the contempor also_ assess the extent to which the insights
ary era. I believe th t th .
resentative of present-da alli . a e cases in some ways are rep-
found differences as well YTh anc~ d~•na~cs, although clearly there are
proin Chapter 7. · ose sunilanttes and differences will be discussed
 The Two Leagues ef the Three Emperors
CHAPTER
3

THE LEAGUES of the Three Emperors are interesting cases for historical and
analytical reasons. Historically, Europe of the early 1870s looked dramatically
different than it had in the preceding decades. Germany was unified, Italy, too;
Austria-Hungary and France were badly defeated in the process. The wars of
unification ushered in a lengthy period of peace. From 1870 to 1914, the great
powers of Europe were frequently in conflict but never at war with one another.
During the Concert of Europe from 1815 to 18 54, the great powers deliberately
sought to keep the peace through consultation and compromise. From 1870 to
1914, the great powers ofEurope were committed to keeping the peace among
themselves, but the primary mechanism through which they chose to do so-
forming alliances with their rivalsculminated in disaster.
The first alliance constructed to keep the peace in this period was the first
Three Emperors' League, or Dreikaiserbund. Its analytical importance stems
from the fact that it was an alliance formed among adversaries whose principal
threats, for two of the members, came from each other. It was a mechanism
through which Austria-Hungary and Russia tethered, and Germany hedged its
bets between Austria and Russia at the same time that it balanced against
France. Above all, the Three Emperors' Leagues manifest well the strategies of
this period-ally with your enemies and keep the peace.

The Dreikaiserbund, r 873 -78

The .first League of the Three Emperors of Germany, Austria-Hunga~, a~d

iuss1a was created in I 873. This alliance heralded the era of Bismarckian dip
omacy conducted by the newly unified Germany. The League was the first
 40 Two Leagues ef Three Emperors
Two Leagues ef Three Emperors 41
in a series of secret treaties between the great powers of Europe in the four
that derived from the rules embodied in the agreement-made the alliance
decades prior to the outbreak of the First World War. _ appealing.
The Dreikaiserbund represented the coming together of the conservative
The alliance also served to advance Austrian eastern interests vis a vis Russia.
elements of central and eastern Europe at a time when revolutionary socialism
The Balkan question had been reopened, and with it the need to tether Russia as
in western Europe threatened dynastic control. 1 It was an alliance forged
well. It would be easier within the context of the alliance "to repel Russian
among rivals in order to keep close watch on the other member states; it was
encroachments." 5 It was clear that Russia and Austria had conflicts of interest,
an alliance designed to keep the peace among the signatories.
but management of that threat provided the key incentive for the states to ally.
The threat posed by Russia to Austria had several dimensions. First, Austria
FORMATION
feared war with Russia in the Balkans because it would likely be devastating.
Russia was very powerful-its share of total system capabilities was some IO
Austria
percent, and it possessed an army that could deal a severe blow to the Habsburg
Austria's defeat by Germany in 1866 left a legacy of hostility, suspicion, and Empire. Second, a war with Russia would sorely tempt Italy. Italy had not
distrust. Austro-Prussian rivalry had dominated the states' relations since 1848. relinquished hope of detaching Trento and Trieste from the Austro-Hungarian
The Seven Weeks' War in 1866, ending in the Treaty of Prague, perpetuated Empire; a war between Austria and Russia could easily provide an opportunity
Austrian animosity, as its defeat was absolute. Despite the fact that the peace for Italy to try to satisfy its irredentist desires. The need to neutralize the threat
terms embodied in the Treaty of Prague were not unduly severe, the ofltalian intervention in the event of war with Russia was one factor that
humiliation of the defeat led Austria-Hungary to support France whole- ultimately led to the conclusion of the Triple Alliance; the desire to neutralize the
heartedly in the Franco-Prussian War in r 870. The total defeat of France in the Russian threat in the event of war with Italy heightened the utility of the Three
Battle of Sedan on September r, 1870, however, prompted the Austrians to Emperors' League. As Peter Alexandrovitch Saburov, a special emissary to
rethink the orientation of their policy and ultimately led them to alter it. When Berlin and subsequently ( I 880 ~ 84) Russian ambassador to Berlin, observed
it became clear that France would lose its war with Germany, Austria no some years later in a letter to Nicholas Karlovitch de Giers, assistant minister of
longer wished to tie its fate too closely to France, especially in light of Foreign Affairs from I 875 to 1882 and thereafter minister of Foreign Affairs
repeated Prussian attempts to reconcile. (until 1894), "For Austria, that Alliance has become a matter of safety, for could
The threat posed to Austria by the newly united Germany ultimately made a she be at rest a single day with on the one hand the Italian aspirations, and on the
tethering alliance appealing. The conflicts between them underpinned Austrian other those of the Slav populations, towards whom she has not the ability to
motives. As Austro-Hungarian foreign minister Friedrich Ferdinand Beust know how to
reported to the emperor regarding a Prussian offer of friendship: tnake herself
sympathetic?" 6 .
We must not forget that this offer is made at a time when our neighbor has in - We may also identify the tactical side benefits accruing to Austr~a-Hungary
creased in power to gigantic extent, when the only other European State that de-
serves to be powerful has shown itself friendly to Prussia and hostile to us, and that
for entering the Three Emperors' League. Tethering both Russia and G_ermany
this offer of friendship comes at a time when our domestic affairs could easily offer was essential so as to prevent a Russo-German coalition fro~ formmg
the German government a pretext for hostile action. 2 ·
against Austria. Russia and Germany were tra di · men
itiona 1 c: · ,ds· an alliance be-
tween them was easy for Austria to imagine. Such a combination would have
Beust continued, arguing that Austria-Hungary should "keep a sharp watch"
been threatening to Austria-Hungary, given its antipathy toward Germany and
but "conc_eal any mistrust."3 By early 1871, the foreign minister believed that
its rivalry with Russia. Further, the threats Austria confronted resonat~d within
Austria needed to tether the German threat. Without an agreement, he feare~
the empire. The Dual Monarchy faced a plethora of challenge_s to its control.
that Bismarck would foment revolution among the Austro-Germans. 4 Austna
These came in the form of pan-Slavism and revolutionary liberalism. The
was thus drawn into the Three Emperors' League in order to keep close watch
commitment to monarchical solidarity offered by the League of the Three
o_n -~ennany. It was an effective tethering mechanism and offere~ the
Emperors would provide a bulwark against the rising tide of domestic agitation.
possibility of keeping the peace among them. The institutional incentives-the
clarity of expectations regarding the member states' behavior Iro ·. '· all f m Germany n sum, the threats Austria faced emanated prmc1p Y
Russia, and Italy. Austria entered the first League of the Three Emperors m
·

4244Two LeaguesefefThree
TwoLeagues ThreeEmperors
Emperors
Two LeaguesefefThree
TwoLeagues ThreeEmperors
Emperors43
45
order to tetherand
England Germany and Russia.
Russia-France The Schonbrunn
respectively, Conventionofprovided
the impossibility the
reconciling
formalAustrian,
rules Russian
designed andtoBritish
dictate interests
behavior,in theactivities,
East will andlead expectations
to war .... Germany
of the It was driven by the desire keep theTABLE peaceJ.Ibetween it and its adversary. The
would be called upon day after day to act as umpire between the two hostile groups League Member
would enableStates' Motivations
Russia to keep for its
Joining
eye the
on Three
both Emperors'
Austria and League,
Germany1873 and to
member states. It was designed to hold two of those threats at bay. The Italian
at the Congress; the most thankless task that could
threat was subsequently managed via the Triple Alliance, signed in May I 8 82. The fall to us· and as we could never manage the Balkan
Coun question in a way that would avert war with another great power.
Motivation for allying
see our way from the outset to seize and bind fast
first League, though informal and at a low commitment level, was the first alliance either ~f the two parties, the most The League,
try in sum, offered an institutional context in which Russia could keep
probable result for us would be that our three friends, Russia, Austria, and England, watch over Austria-
its allies. It served to reduce the
Tether conflicts Russia faced with Austria-
of this period to be used as a strategy of conflict management. The conflicts were Hungary
would leave the Congress in a state of annoyance against us. 11 Hungary by providing informal and Russia
formal Tether
rules that would alter the behavior of the
not resolved, but they could, at least for some time, be tabled. Russ Germany
Bismarck thus sought to prevent such a situation from arising. Above all, he member states.
ia Tether Austria-Hungary
wanted to continue a policy of rapprochement between Russia and Austria, despite Germany Hedge between Russia and Austria-
Russia difficulties. 12 Germany needed to stand between Austria and Russia
the attendant Germany Hungary Balance France
"like a man between two vicious dogs who would fly at each other as soon as they
andGermany,
For Russia's interests
in contrast derived
to Austriaprimarily from each
and Russia, other.Emperors'
the Three GermanyLeague was most
In 1872,
were Austro-Hungarian
unleashed." 13
emperor Franz
If Austro-Russian Joseph were
antagonism agreed lefttounmanaged,
meet German kaiser
Germany
threatened by the possibility of a Russian-French
served as a hedging device. Above all, Germany wanted to preserve good relations coalition. This was a threat not
Wilhelm in Berlin. have
would ultimately The meeting
to choose of the two emperors
between would be
them, creating a asituation
historic inmoment
whichina
wholly dispelled by the alliance of the three emperors.
with Russia and to improve its relations with Austria. Germany also was dedicated
improving
friendly power relations between
would their respective
necessarily become countries. The newsTo
an adversary. of this
most impending
German
Each member
to improving state was a relations.
Russo-Austrian major power, and to
In order they all shared
generate contiguous
cordiality, borders.
agreement
visit, when it
statesmen, notreached
Bismarck Tsaralone,
Alexander the idea II ofofRussia,
conflictwas greetedRussia
between with suspicion
and Austria-and
apprehension. His immediate thought was "The that the Power
among and
them geographic
all was proximity
necessary. were
Keeping certainly
both states sufficient
on good to be
terms-with threatening.
Germany In
Hungary was intolerable. As Bulow ar~ed, mostemperors
unpleasant of possibility
Austria-Hungary of all,-
and Germany would be coming to agreement 1873,
and France
each had
other-would over IO percentthem
eliminate of theas share
possibleof system-wide
alliance capabilities;
partners for France Germany
and
which we happily consider unhk~ly-would be on an alliance
a war betweenagainstRussia Russia.
and AustriaThe
tsar's reaction to the news wa~ w?at ~ne might expect in the context of tethering had
wouldaboutspare theGermany
same, asfrom did Russia.
makingAustria-Hungary
a difficult and perhaps possessed about
perilous 5 percent
choice and
between
especially if the latter were invaded. · .. [S]ince a Russo-Austrian war involves
alliances- VII, Italy
them.about 3 percent.thusFurther,
This approach
15
allowedasGermanydiscussed to in the its
hedge previous
bets in section,
two ways. theFirst,
Austro-
the
danger forheus,sought we aretoforced mvite to himself
do ourtoutmost Berlin.and Heforemost
asked Princeto stop Henry
it by Reuss
one means
the German ambass~dor at St. Petersburg, "Hasn't anyone in Berlin written to you Hungarians
alliance were
guaranteed wary
that of the
Germany intentions
would not of
havebothto Germany
take sides and
between Russia,
the two most
or another." 14

thatG~rmany
they wishsought to havetome thereits at the particularly
adversaries; the latter.
second, theRussia,
alliance too,
ensured wasthat leery of Austria,
neither would beless freeso of Germany.
to ally with
hedge betssame time as the
by securing an Emperor
alliance of withAustria? Do you
both Austna-
think this and
would be agreeable to theassured
king?"the 7 Apeace
formal invitation Germany
France. The alliance also allowed Germany to secure itself against France. 10France,
was not particularly threatened by either-its principal threat,
Hun~ary Russia. This strategy between the foll
two owe~ shortly
_adversaries
~hereafter. Although emanated
Hedging served from outside
to ensure theGermany's
alliance. position in the system.
and precluded either the fromThreeallyingEmperors'
with France.LeagueThis wasalsonot co?stituted
mamfested until Franz
Germany's
Joseph
desire toreciprocated
balance France. and Ininvited
other words,both Alexander
the allianceand Wilhelm ernb
for Germany to Vienna
di d th ·the·· TheThe specific
motives threats tothe
underlying the decision
member states may betoidentified
of Germany forge theasDreikaiser
follows. bund
following
·d year, it does reveal how the League was born in enmity rather than Austria-Hungary, as mentioned above, was vulnerable
were manifest in Bismarck's reasoning to Bernhard Heinrich von Bulow in a letter to Italian irredentism and
o ie reat motivations toward Austna, Russia, an Slavic nationalism. The combination was very threatening for a couple of reasons.
amity,Fr~nce.in suspicion
There were andtactical
distrustside rather thanaccruing
benefits confidence and faith:asTrusting
to Germany well. Theneither about holding a congress on the eastern question.
Germany nor Austria, ultimately the best strategy for Russia was to stay close to First, Italian irredentism and Slavic nationalism offered challenges t~ ~he
alliance further represented the conservative ideal; it embodied a commitment t? I consider it dangerous to the Kaiserbundniss, to peace, and to Germany's rela-
them both. centralized authority of the empire. Once the empire's Slavic nationalities began
the monarchical fraternity that would serve as a physical and psychological buffer tions towards her allies, for a Congress or a Conference to be held. · · · The dan-
One critical
toalliberal function the L dcR· · ·· voicing
ger todemands, political control over thethem would becomeinterests
more difficult.
. .threats emanating. fromeague the west.serve 10r ussia, and one of its prmci- the Kaiserbundniss rests on fact that Austria's lie much nearer
pGe~any
motivations Moreover
to those of England than of Russia, and that a Congress will by
other minorities would be encouraged to follow suit making
sharply their own
accentuate
in theforearlyaccedmg I 870s to the alliance,
sought was to manage
to maintain the threat
its newfound to it
dominance ~nd
that emanated from A ·R· h · demands for autonomy
this difference, seeing and self-determination.
that Austria will be obliged This Was true, too,
to choose of thethe two ab-
between
uru~. In seeking . to secure thatuss1a
ustna. end, itsoug hedged
ta truceits bets
withby forging an al- in
Austria-Hungary
liance
the west. with boththe
While A talli · n d R · inl did . di · irredentists
solutely opposed
in Trentointerests
and Trieste.of England
Any development
and Russia.that Austria
opened willthebe door
forced to to
. . . a ce certa1 us rrayan The alliance
ussia.completely
1 not ispel thehowever
suspr- was
unable
cion to and distrust upon whi h it bd·· · h declare herself
discontent with the for one orauthority
central the otheroffor thethe time being.
~mpire could Each
be Cabinet will
dangerous. Second,havethe
to
withsr d th ifir hi ' c 1 was ase ' , It did serve as a mechamsm throug nego~iate notwasunder the mediation of one equally friendly to both, andand itself
h
wt eich a~ . e centrr gal
~oth states coulddemise forces at work
keep close in the east, which ultimately led to domestic threat matched by an International one coming from Italy Russia.
Dre1kaiserbund's 878watch
inI 1all _ on the other, while tending to their
respective domestic itu · As interested
these states In keeping the peace, as is the
nurtured rebellio · · · h 1 16 The German Cabinet, but will be hampered
. si anons. t owed for the preservation The motivations of theforstatus
Austria H ns agamst authonty,
by themterference they assumed
of other participants in threatemng
the business, status
whose t emse
object ves.will ~e to
quo m the east and ensur d th b h . fact that
R. 8 e at war etween .t .e. two adversaries would not disturb theRussia
friendlysupported
relations the SlavsRussia
between in attaining freedoms
and Austria. froma the
France, Ottoman
competitor
th f In the words of Ts
result.
.
r Al- ungary,
a
d "Id ussia, and Germany
exan er,
. for jommg
o not wish to
. ~mpire
for implicitly
Russia's posed
friendship, a grievous
will bid threat
up the to Austria-Hungary's
claims of Russia and control
the over
personal
oppose e rrsr
the Leagiie
views of the Three Emperors are summarized in Table 3. I. Its Slav1· of· Gortchakoff;
impo'.tance ·· England realising that her interests no longer
EXTERNAL VERSUS INTERNAL THREAT c minorities.
f o Austria. I will ther f h . . coincide with those of Russia, will try to induce Austria
. e ore go to t e end m order to maintain intact the en- Relations between Austria-Hungary and Russia were to followstrained
further England's by lead
the
tente Q trois," 9 and Count Andrassy Will find it extremely difficult to withstand the pressure of
The level of threat within th D ikai b fact that any expansion by either power would be at the expense of the other.
__ 1 thr . . • •· Russia wasund
e rei ser motivated
was greater fc the f h
to allthan circumstan.ce~. Up to present the Drei-Kaiser-Bundniss has been the security for
level of Russia, with its interest in protecting its trade route through the
. ext,.£Uill . eat. Wtthir.l the alli . h . . . ' peace. If It is we~kened and relaxed by the deliberate rapprochement to one
ance, t e pnnc1pal threats to Austria-Hungary s another of Austria-
. Y or many o t e same reasons as was Austna.