Trading on Preconceptions: Why World War I Was Not a Failure
of Economic Interdependence
Erik Gartzke, Yonatan Lupu
International Security, Volume 36, Number 4, Spring 2012, pp. 115-150
(Article)
Published by The MIT Press
For additional information about this article
Access provided by University of British Columbia Library (18 Sep 2013 03:44 GMT)
http://muse.jhu.edu/journals/ins/summary/v036/36.4.gartzke.html
The relationship between
economic interdependence and military conºict is among the most
studied and debated in the international relations literature. Scholars who argue
that economic ties reduce the likelihood of conºict have struggled to reconcile
this view with the outbreak of World War I. The conventional wisdom
among political scientists is that World War I constituted a failure of economic
integration to maintain peace.1
Even prominent advocates of liberal theory
view World War I as an unfortunate contradiction to the general argument that
trade inhibits conºict.2
These arguments are important for several reasons. First, they bear on scholars’
understanding of the causes of war and peace and on more pragmatic issues
of policy in an increasingly globalized, but still fractious, world. Whether
interdependence encourages states to resolve differences diplomatically or not
Trading on Preconceptions
Trading on
Preconceptions
Erik Gartzke and
Yonatan Lupu
Why World War I Was Not a Failure of
Economic Interdependence
Erik Gartzke is Associate Professor of Political Science and Yonatan Lupu is a Ph.D. candidate in political
science, both at the University of California, San Diego. The authors contributed equally to this article, and
their names are listed alphabetically.
The authors wish to thank Mariano Bertucci, Chris Fariss, Eric Hamilton, Jack Levy, Keir Lieber,
Norrin Ripsman, Arthur Stein, William Wohlforth, and the anonymous reviewers for helpful comments
and suggestions.
1. Kenneth N. Waltz, Theory of International Politics (New York: McGraw-Hill, 1979); Barry Buzan,
“Economic Structure and International Security: The Limits of the Liberal Case,” International Organization,
Vol. 38, No. 4 (Autumn 1984), pp. 597–624; Dale C. Copeland, “Economic Interdependence
and War: A Theory of Trade Expectations,” International Security, Vol. 20, No. 4 (Spring
1996), pp. 5–41; Dale C. Copeland, The Origins of Major War (Ithaca, N.Y.: Cornell University Press,
2000); Paul A. Papayoanou, “Interdependence, Institutions, and the Balance of Power: Britain,
Germany, and World War I,” International Security, Vol. 20, No. 4 (Spring 1996), pp. 42–76; Paul A.
Papayoanou, Power Ties: Economic Interdependence, Balancing, and War (Ann Arbor: University of
Michigan Press, 1999); Norrin M. Ripsman and Jean-Marc F. Blanchard, “Commercial Liberalism
Under Fire: Evidence from 1914 and 1936,” Security Studies, Vol. 6, No. 2 (Winter 1996/97), pp. 4–
50; David M. Rowe, “World Economic Expansion and National Security in Pre–World War I
Europe,” International Organization, Vol. 53, No. 2 (Spring 1999), pp. 195–231; Jack S. Levy, “War
and Peace,” in Walter Carlsnaes, Thomas Risse, and Beth Simmons, eds., The Handbook of International
Relations (London: Sage, 2001), pp. 350–367; David M. Rowe, “The Tragedy of Liberalism:
How Globalization Caused the First World War,” Security Studies, Vol. 14, No. 3 (Spring 2005),
pp. 407–447; John J. Mearsheimer, The Tragedy of Great Power Politics (New York: W.W. Norton,
2001); Robert Jervis, “Theories of War in an Era of Leading Power Peace,” American Political Science
Review, Vol. 96, No. 1 (March 2002), pp. 1–14; and Frank Zagare, The Games of July: Explaining the
Great War (Ann Arbor: University of Michigan Press, 2011), p. x.
2. Norman Angell, The Great Illusion, 2d ed. (New York: G.P. Putnam’s Sons, 1933); Richard
Rosecrance, The Rise of the Trading State: Commerce and Conquest in the Modern World (New York:
Basic Books, 1985); and Patrick J. McDonald and Kevin Sweeney, “The Achilles’ Heel of Liberal IR
Theory? Globalization and Conºict in the Pre–World War I Era,” World Politics, Vol. 59, No. 3
(April 2007), pp. 370–403.
International Security, Vol. 36, No. 4 (Spring 2012), pp. 115–150
© 2012 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.
115
contributes to expectations about the future of international affairs. If interdependence
at the turn of the last century failed to inhibit the most general conºict
up to that time, then it may fail to perpetuate peace in the new century.
Second, World War I serves as a “critical case” for proponents and critics of liberal
theory. The liberal argument will be strengthened to the degree that economic
integration a century ago can be reconciled with the outbreak of general
war in Europe in 1914. Finally, World War I is an inherently important case, implying
(as others have already suggested) that explanations for its advent and
expansion deserve careful scrutiny.
Recasting the outbreak of World War I in a light more favorable to liberal
theory is actually less difªcult than the prevailing consensus would suggest.
At the very least, World War I is not a particularly compelling example of the
failure of liberal trade theory, and it may even indicate the strength of the liberal
perspective. The conventional wisdom about interdependence and World
War I seems peculiar in light of three historical facts. First, the turn of the century
saw a series of intense crises among the interdependent states of Western
Europe that nevertheless did not result in open warfare. Second, despite these
growing tensions among the Western powers, the ªghting in 1914 actually began
among the less interdependent powers of Austria-Hungary and Serbia.
Third, during the same period in which the highly interdependent European
powers were generally able to resolve their crises without resorting to war, the
less interdependent powers were typically unable to do so. These observations
indicate that there may indeed have been a relationship between interdependence
and the paciªc settlement of disputes during this era. These facts also
suggest that the role of interdependence in the outbreak of war in 1914 may be
better understood in the context of the relationship between interdependence
and conºict in the era leading up to the war.
We argue that the relationship between economic interdependence and the
outbreak of World War I has been broadly misinterpreted and that interdependence
did not really fail in 1914. World War I did happen, of course, but it is no
coincidence that it began among states that were less well integrated into the
global economy than their key alliance partners. There is much evidence to
suggest that economic linkages served an important role in averting escalation
to warfare in the series of crises that led up to the Great War. These crises, however,
created an incentive for more integrated countries, most importantly,
Germany and Russia, to show an increasing resolve to support their weaker,
less interdependent, allies, Austria-Hungary and Serbia. Tight alliance ties
then effectively handed the foreign policies of interdependent powers over to
countries that were less well integrated into the world economy. Economic integration
was incapable of forestalling conºict where integration had yet to ocInternational
Security 36:4 116
cur. Nor could interdependence prevent war, once started, from spreading
through competing networks of military alliance commitments that made it
much more costly for Western powers to take advantage of available economic
linkages to the greatest paciªc effect.
This article begins by explaining the various theories that argue that interdependence
reduces the likelihood of conºict, as well as theories opposed to this
view. Second, we discuss the shortcomings of treating the outbreak of World
War I as a failure of interdependence. Third, we describe interdependence in
the two economic subsystems of Europe prior to the war. In the fourth section,
we compare crisis behavior in the highly interdependent and less interdependent
portions of Europe during the period 1871 to 1913. Fifth, we discuss the
July crisis and the outbreak of war in 1914. We conclude by using the evidence
from our analysis to further reªne commercial liberal theory.
War and Interdependence
Theories of interdependence argue that economic relationships that span international
borders decrease the likelihood of states to engage in conºict. Nations
inclined to war can be deterred, informed, or transformed by the value of economic
linkages. Critics of interdependence argue that economic ties do not
have powerful pacifying effects on world politics, either because the stakes are
insufªcient to deter conºict or because some aspect of the causal logic offered
by liberals and others is incorrect.
The most traditional approach to the relationship between economic relations
and conºict focuses on explaining how economic ties linking nations
change the incentives of actors in the international system. Beginning in the
modern era, scholar-statesmen such as Richard Cobden and Norman Angell
argued that interdependence, primarily in the form of interstate trade, raises
the opportunity costs of war, thus making contests less likely. The logic of
these arguments is that a war between trading partners would likely disrupt
that trade, forcing states to seek other markets. This would require a shift to
different, less lucrative, trade partners.3
Others argue that, as trade increases,
states can achieve gains more efªciently through economic means than
through warfare. In other words, when states can grow their economies
through international commerce, there is a decreased incentive to attempt to
Trading on Preconceptions 117
3. Richard Cobden, Political Writings of Richard Cobden, Vol. 1 (London: T. Fisher Unwin, 1903
[1867]); Angell, The Great Illusion; Francis Delaisi, Political Myths and Economic Realities (London:
Noel Douglas, 1925), pp. 57–78; and Richard N. Rosecrance and Arthur Stein, “Interdependence:
Myth or Reality?” World Politics, Vol. 26, No. 1 (October 1973), pp. 1–27.
do so through territorial conºict.4
Open ªnancial and goods markets may also
create similar disincentives for states to ªght.5
The notion that international trade is associated with a decreased likelihood
of conºict has found signiªcant empirical support. A series of studies using
events data found that higher levels of trade interdependence (deªned as
trade/gross domestic product [GDP]) are associated with lower probabilities
of interstate war.6
Other studies have built on this research to demonstrate that
the ªnding is robust to alternate speciªcations of the temporal domain and
unit of analysis.7
The relationship between trade and conºict is likely more
complex than initially theorized. First, high levels of trade dependence may
embolden a state’s opponents.8
Responding to this critique, Patrick McDonald
argues that scholars must shift their focus from aggregate trade ºows to the extent
to which states pursue free trade policies.9
Han Dorussen offers another
important theoretical reªnement, noting that the opportunity costs of conºict
created by different types of trade vary signiªcantly. Two factors he points to
that raise the opportunity costs of conºict created by trade are lower factor
mobility and higher asset speciªcity.10
Finally, the effects of trade interdependence
may be contingent on the mediating effects of democracy.11
One strain of research focuses on ways that trade may make conºict less
likely through mechanisms other than raising the opportunity costs of war.
Etel Solingen, for example, argues that trade allows domestic actors to build
International Security 36:4 118
4. Rosecrance, The Rise of the Trading State.
5. Arthur A. Stein, “Governments, Economic Interdependence, and International Cooperation,” in
Philip E. Tetlock, Jo L. Husbands, Robert Jervis, Paul C. Stern, and Charles Tilly, eds., Behavior, Society,
and Nuclear War, Vol. 3 (New York: Oxford University Press, 1993); and Michael W. Doyle, Ways
of War and Peace: Realism, Liberalism, and Socialism (New York: W.W. Norton, 1997).
6. Solomon W. Polachek, “Conºict and Trade,” Journal of Conºict Resolution, Vol. 24, No. 1 (March
1980), pp. 55–78; Mark J. Gasiorowski, “Economic Interdependence and International Conºict:
Some Cross-National Evidence,” International Studies Quarterly, Vol. 30, No. 1 (March 1986), pp. 23–
38; and John R. Oneal and Bruce M. Russett, “The Classical Liberals Were Right: Democracy, Interdependence,
and Conºict, 1950–1985,” International Studies Quarterly, Vol. 41, No. 2 (June 1997),
pp. 267–293.
7. On speciªcation of the temporal domain, see William K. Domke, War and the Changing Global
System (New Haven, Conn.: Yale University Press, 1988). On the unit of analysis, see Edward D.
Mansªeld, Power, Trade, and War (Princeton, N.J.: Princeton University Press, 1994).
8. James D. Morrow, “How Could Trade Affect Conºict?” Journal of Peace Research, Vol. 36, No. 4
(July 1999), pp. 481–489; and Erik Gartzke, Quan Li, and Charles Boehmer, “Investing in the Peace:
Economic Interdependence and International Conºict,” International Organization, Vol. 5, No. 2
(Spring 2001), pp. 391–438.
9. Patrick J. McDonald, “Peace through Trade or Free Trade?” Journal of Conºict Resolution, Vol. 48,
No. 4 (August 2004), pp. 547–572.
10. Han Dorussen, “Heterogeneous Trade Interests and Conºict: What You Trade Matters,” Journal
of Conºict Resolution, Vol. 50, No. 1 (February 2006), pp. 87–107, at p. 90.
11. Papayoanou, Power Ties; and Christopher Gelpi and Joe Grieco, “Economic Interdependence,
the Democratic State, and the Liberal Peace,” in Edward D. Mansªeld and Brian M. Pollins, eds.,
Economic Interdependence and International Conºict (Ann Arbor: University of Michigan Press, 2003).
cross-national coalitions that both promote greater interdependence and cause
convergent transformations in state preferences.12
Along related lines, Paul
Papayoanou argues that economic relations create strong domestic-level interests
that delimit a leader’s ability to credibly counter external threats or chal-
lenges.13
Building on these arguments, several scholars have constructed what
is sometimes known as the “commercial peace” or “capitalist peace” view of
the relationship between economics and war. They argue that interdependence
molliªes the effects of states’ security dilemmas by creating common interests
and reducing uncertainty. Although the bulk of the work on interdependence
and conºict focuses on trade,14
other forms of transnational economic
relations are also crucial. Erik Gartzke, Quan Li, and Charles Boehmer argue
that, along with trade, interstate monetary policy cooperation and capital
ºows reduce the likelihood of conºict by allowing states in crisis situations
to send costly signals without needing to resort to violence or crisis escalation
that may precipitate violence.15
Building on this, Gartzke argues that
interdependence—deªned as including trade, development, open ªnancial
markets, and monetary policy coordination—reduces conºict by (1) aligning
states’ interests, which gives them less to ªght over; (2) providing a means
of peacefully securing resources; and (3) allowing states to foresee the costs of
ªghting, which facilitates bargaining and compromise.16
A signiªcant group of scholars, however, have posed important criticisms
of interdependence theory. At a theoretical level, several scholars have claimed
that interdependence and conºict are unrelated. Even if economic factors matter,
they argue, decisions to initiate conºict are primarily based on security and
military concerns.17
James Morrow provides two arguments against the causal
Trading on Preconceptions 119
12. Etel Solingen, Regional Orders at Century’s Dawn: Global and Domestic Inºuences on Grand Strategy
(Princeton, N.J.: Princeton University Press, 1998).
13. Papayoanou, Power Ties.
14. In addition to the work cited above, see John R. Oneal and James Lee Ray, “New Tests of the
Democratic Peace: Controlling for Economic Interdependence, 1950–85,” Political Research Quarterly,
Vol. 50, No. 4 (December 1997), pp. 751–775; Mark J.C. Crescenzi and Andrew J. Enterline,
“Time Remembered: A Dynamic Model of Interstate Interaction,” International Studies Quarterly,
Vol. 45, No. 3 (September 2001), pp. 409–431; and John R. Oneal, Bruce Russett, and Michael L.
Berbaum, “Causes of Peace: Democracy, Interdependence, and International Organizations,” International
Studies Quarterly, Vol. 47, No. 3 (September 2003), pp. 371–393.
15. Gartzke, Li, and Boehmer, “Investing in the Peace.”
16. Erik Gartzke, “The Capitalist Peace,” American Journal of Political Science, Vol. 51, No. 1 (January
2007), pp. 166–191. See also William Reed, “Information and Economic Interdependence,” Journal
of Conºict Resolution, Vol. 47, No. 1 (February 2003), pp. 54–71; and Patrick J. McDonald, The
Invisible Hand of Peace: Capitalism, the War Machine, and International Relations Theory (Cambridge:
Cambridge University Press, 2009).
17. Barry Buzan, “Economic Structure and International Security: The Limits of the Liberal Case,”
International Organization, Vol. 38, No. 4 (August 1984), pp. 597–624; and Jack S. Levy, “The Causes
of War: A Review of Theories and Evidence,” in Philip E. Tetlock, Jo L. Husbands, Robert Jervis,
connection between trade and conºict. First, if trade reduces aggressive behavior
by one state, this would strategically have the effect of increasing the
likelihood of aggressiveness by another state, thus nullifying the effect of interdependence
in inhibiting conºict. Second, Morrow argues that ªrms should
rationally anticipate when conºict with other states is more likely and strategically
reduce their trade with ªrms in those states.18
As a result, trade may be
endogenous to conºict, or the direction of causality may be the reverse of that
theorized by commercial liberals.19
Omar Keshk, Brian Pollins, and Rafael
Reuveny investigate this problem, ªnding evidence that it is more likely that
conºict affects trade than vice versa.20
Similarly, Michael Ward, Randolph
Siverson, and Xun Cao argue that prior statistical ªndings are the result of geographical
proximity, dependence among militarized disputes with the same
initiator or target, and higher-order dependencies in dyadic data. Accounting
for these, they ªnd no relationship between trade levels and conºict.21
Others argue that trade increases, rather than decreases, the likelihood of
conºict, a result that would support the Marxist view that capitalism leads
to conºict. Generally speaking, capitalism and economic development may
generate demands for imperialistic territorial conquest. As commerce grows,
states compete over scarce resources and may use military means to obtain
them. States may also use trade and military force together to maintain their
colonial holdings. Economic ties may create rivalries as partners become more
concerned with dividing the gains from the trade than with mutual gains.22
Several political scientists have built on these arguments, noting that trade
may increase conºict by creating competition over relative gains.23
There is
also some empirical support for this proposition, although it is controversial.
International Security 36:4 120
Paul C. Stern, and Charles Tilly, eds., Behavior, Society, and Nuclear War, Vol. 1 (New York: Oxford
University Press, 1989).
18. Morrow, “How Could Trade Affect Conºict?”
19. Rafael Reuveny, “Bilateral Import, Export, and Conºict/Cooperation Simultaneity,” International
Studies Quarterly, Vol. 45, No. 1 (March 2001), pp. 131–158.
20. Omar M. Keshk, Brian M. Pollins, and Rafael Reuveny, “Trade Still Follows the Flag: The Primacy
of Politics in a Simultaneous Model of Interdependence and Armed Conºict,” Journal of Politics,
Vol. 66, No. 4 (November 2004), pp. 1155–1179.
21. Michael D. Ward, Randolph M. Siverson, and Xun Cao, “Disputes, Democracies, and Dependencies:
A Reexamination of the Kantian Peace,” American Journal of Political Science, Vol. 51, No. 3
(July 2007), pp. 583–601.
22. John A. Hobson, Imperialism (Ann Arbor: University of Michigan Press, 1938 [1905]); V.I. Lenin,
Imperialism: The Highest Stage of Capitalism (New York: International Publishers, 1970 [1916]); Albert
O. Hirschman, National Power and the Structure of Foreign Trade (Berkeley: University of
California Press, 1945); and Benjamin J. Cohen, The Question of Imperialism (New York: Basic Books,
1973).
23. Michael Mastanduno, “Do Relative Gains Matter? America’s Response to Japanese Industrial
Policy,” International Security, Vol. 16, No. 1 (Summer 1991), pp. 73–113; and Joanne Gowa, Allies,
Adversaries, and International Trade (Princeton, N.J.: Princeton University Press, 1994).
Using a measure of trade salience, Katherine Barbieri ªnds statistical support
for this proposition for the period 1870 to 1938.24
World War I as a Test of Liberal Theory
The causes of World War I have long been an important topic of debate and
conjecture among students of international politics.25
The case appears particularly
problematic for those claiming that interdependence reduces the likelihood
of war. In the four decades prior to the Great War, Europe experienced a
dramatic increase in the levels of interdependence, especially among several of
the major powers. Liberal theory, therefore, would appear to predict a low
probability of war among interdependent European states and a low probability
that wars would spread to countries with globalized economies. That interdependent
states engaged in a war of unprecedented intensity thus bolsters
the view that interdependence is not an important determinant of conºict.26
Several scholars have attempted to defend or explain away the onset of war
in 1914 in light of the apparent contradiction with liberal theory. John Oneal
and Bruce Russett argue that the effects of interdependence are conditional on
democracy. Thus, these effects did not function in the lead-up to the war because
Germany was not sufªciently democratic. They also argue that, although
economic interdependence was much greater during this period than
in prior eras, it was nonetheless not sufªciently deep as to prevent a major
conºict, a point that Richard Rosecrance also makes. Rosecrance, repeating an
argument posed by Angell, also claims that European leaders misperceived
both the value of interdependence and the costs of a potential war.27
Patrick
Trading on Preconceptions 121
24. Katherine Barbieri, “Economic Interdependence: A Path to Peace or a Source of International
Conºict?” Journal of Peace Research, Vol. 33, No. 1 (February 1996), pp. 29–49. For critiques, see John
R. Oneal and Bruce Russett, “Assessing the Liberal Peace with Alternative Speciªcations: Trade
Still Reduces Conºict,” Journal of Peace Research, Vol. 36, No. 4 (July 1999), pp. 423–442; and Erik
Gartzke and Quan Li, “Measure for Measure: Concept Operationalization and the Trade Interdependence-Conºict
Debate,” Journal of Peace Research, Vol. 40, No. 5 (September 2003), pp. 553–571.
25. Jack Snyder, “Civil-Military Relations and the Cult of the Offensive, 1914 and 1984,” International
Security, Vol. 9, No. 1 (Summer 1984), pp. 108–146; Scott D. Sagan, “1914 Revisited: Allies, Offense,
and Instability,” International Security, Vol. 11, No. 2 (Autumn 1986), pp. 151–175; Marc
Trachtenberg, “The Meaning of Mobilization in 1914,” International Security, Vol. 15, No. 3 (Winter
1990/91), pp. 120–150; Jack S. Levy, “Preferences, Constraints, and Choices in July 1914,” International
Security, Vol. 15, No. 3 (Winter 1990/91), pp. 151–186; and Keir Lieber, “The New History of
World War I and What It Means for International Relations Theory,” International Security, Vol. 32,
No. 2 (Fall 2007), pp. 155–191.
26. E.H. Carr, The Twenty Years’ Crisis: 1919–1939 (London: Macmillan, 1939); Waltz, Theory of International
Politics; John J. Mearsheimer, “Back to the Future: Instability in Europe after the Cold
War,” International Security, Vo. 15, No. 1 (Summer 1990), pp. 5–56; and Copeland, “Economic Interdependence
and War.”
27. John R. Oneal and Bruce Russett, Triangulating Peace: Democracy, Interdependence, and Interna-
McDonald and Kevin Sweeney provide a detailed analysis delineating the relationship
between interdependence and the outbreak of World War I. They argue
that most theories of commercial liberalism incorrectly focus on trade
ºows rather than on trade protection. They ªnd that during the period 1865 to
1914, lower levels of trade protection were associated with a reduced propensity
toward conºict, showing that interdependence (following their deªnition)
was associated with peace during this era. In addition, they note that economic
conºict between Germany and Russia, in the form of rising tariff rates, preceded
the outbreak of war.
We build on McDonald and Sweeney’s argument by asking to what extent
the outbreak of World War I can be viewed as a failure of interdependence. We
differ in how we deªne interdependence, adopting the more conventional association
between trade ºows and conºict, rather than examining protectionist
measures, which strike us as closely associated with the very forces propelling
states to war. We also differ from McDonald and Sweeney in our focus on the
competing interconnectedness of alliance ties and their role in overcoming
the pacifying effect of national economic interest that deterred conºict among
interdependent states prior to World War I.
On the surface, it would seem clear that interdependence failed to function
as predicted by liberal theory at the advent of World War I. Yet there are three
key ºaws in the connections drawn by both critics and advocates of liberal theory
that make such a conclusion empirically suspect. First, the occurrence of
war between interdependent states in one important case demonstrates only
that interdependence is not sufªcient by itself to guarantee interstate peace; interdependence
could still reduce the overall likelihood of war. Second, focusing
on the outbreak of World War I to test this theory is problematic because it
is a case chosen based on the dependent variable, the outbreak of war.
War may be made very unlikely and nonetheless occur. If we are to look back
afterward and conclude that the likelihood of war was not much affected by
mitigating factors, then we are committing the classic post hoc, ergo propter
hoc fallacy. We can gain greater explanatory power by examining the variance
in this dependent variable over the relevant period. Third, the conventional
wisdom—that Europe was interdependent but nonetheless went to war—
overlooks crucial variance in the independent variable. Europe in 1914 was not
uniformly interdependent, but consisted of several highly interdependent
powers, other powers that were signiªcantly less reliant on economic relations
with their neighbors, and still other states that were minimally integrated into
this system. We discuss all three issues below.
International Security 36:4 122
tional Organizations (New York: W.W. Norton, 2001); and Rosecrance, The Rise of the Trading State,
pp. 18–19, 88–99.
First, while a probabilistic theory cannot be disproved with a single case, a
reasonable observer might counter that some cases are particularly worth explaining.
Liberal theory is deeply ºawed if it cannot account for a contest as
catastrophic as World War I. We agree completely. Treating a war in which
more than 10 million people died as an outlying case in which a probabilistic
relationship failed would be intellectually unsatisfactory and misleading.
World War I was not just a blip.
The second ºaw with using 1914 to test theories of interdependence, however,
is more consequential. While many scholars have argued that it is not appropriate
to choose cases based on the dependent variable,28
others argue that
doing so is valid when one is particularly interested in explaining a speciªc
case or outcome.29
We take no issue with the study of World War I as a unique
historical phenomenon or with including interdependence as a possible explanatory
variable when attempting to explain what did or did not contribute
to its outbreak. Instead, our concern is that scholars have made erroneous inferences
in using the Great War to falsify liberal theory more generally. Doing
so uses the July 1914 crisis as a “hypothesis testing” case, which is among the
most difªcult and controversial types of case study.30
It is one thing to study a
case to understand what led to a particular outcome in that case; it is a wholly
different undertaking to use the case to test a hypothesis meant to represent
general tendencies of a variable across a great many cases. While Gary King,
Robert Keohane, and Sidney Verba doubt that valid inference can be drawn
from testing a hypothesis with a single case, others argue that doing so is valid
for crucial or critical cases, often referred to as “most likely” and “least likely”
cases.31
Conducting a hypothesis-testing case study of a probabilistic theory based
on a most/least likely logic requires analysts to deªne their case based on the
independent variable rather than the dependent variable. What makes the outbreak
of World War I relevant to theories of interdependence is not just that it
was a war, but that it was a war that involved several powers that were formerly
highly interdependent. Thus, to use this case to test the hypothesis that
Trading on Preconceptions 123
28. Gary King, Robert O. Keohane, and Sidney Verba, Designing Social Inquiry: Scientiªc Inference in
Qualitative Research (Princeton, N.J.: Princeton University Press, 1994).
29. Henry E. Brady and David Collier, Rethinking Social Inquiry: Diverse Tools, Shared Standards
(Lanham, Md.: Rowman and Littleªeld, 2004).
30. Harry Eckstein, “Case Studies and Theory in Political Science,” in Fred I. Greenstein and Nelson
W. Polsby, eds., Handbook of Political Science, Vol. 7 (Reading, Mass.: Addison-Wesley, 1975);
and Jack S. Levy, “Case Studies: Types, Designs, and Logics of Inference,” Conºict Management and
Peace Science, Vol. 25, No. 1 (March 2008), pp. 1–18.
31. Eckstein, “Case Studies and Theory in Political Science.” As Jack Levy argues, “The inferential
logic of least likely case design is based on the ‘Sinatra inference’—if I can make it there, I can
make it anywhere. The logic of most likely case design is based on the inverse Sinatra inference—if
I cannot make it there, I cannot make it anywhere.” Levy, “Case Studies,” p. 12.
interdependence reduces the probability of conºict, we must deªne it in terms
of an era of high European interdependence and study conºict behavior in the
period and region as a whole. As David Stevenson argues, “[T]o concentrate
exclusively on the events of 1914 provides too narrow a basis for understanding
early-twentieth-century crisis behavior. We need to study the confrontations
that did not end in ªghting as well as the ones that did.”32
Put another
way, when deªning a case analysts must ask themselves: “What is this a case
of?”33
We argue that an answer of “This is a case of war” may allow us to understand
the causes of that war, or possibly the causes of war in general. Yet
only by deªning the case as “a case of high interdependence and war” can we
understand whether this factor makes war less likely. To do otherwise, to test
this hypothesis by deªning the case as the outbreak of war in July 1914, is to
commit a fallacy as signiªcant in turn as it would be to dismiss the war as an
“outlier.”
Deªning the case in this way brings us to the third ºaw in the conventional
wisdom: the role of the explanatory variable. The notion that World War I represents
a crucial failure of interdependence seems to rely on the view that it is
a most likely case for testing the hypothesis. World War I, as we noted above,
is often cited as evidence against commercial liberalism. On its face, the case
involves a period of signiªcant and increasing interdependence followed by a
war, and thus presents a problem for those who argue that interdependence
promotes peace. If instead even this case offers some support for liberal claims,
then this is powerful evidence in favor of the liberal perspective. Yet treating
1914 alone as a most likely case in the traditional sense overlooks several historical
facts that are inconsistent with framing the case in this way. First, not all
of Europe was economically interdependent during the pre–World War I period;
indeed, this factor varied in crucial ways across both space and time. Ian
Beckett explains this difference as follows: “The rapid pace of change wrought
by industrialization in western Europe sharpened the contrast with most of the
states of central, southern and eastern Europe. These remained in effect peasant
societies.”34
Second, the war did not begin among the most interdependent
states in Europe but instead among their less interdependent allies. A more
clear-cut most likely case to test liberal theory would have been a war that began
between Germany and Russia or France. Certainly, Austria-Hungary and
Serbia were supported and possibly encouraged by their allies, but the fact reInternational
Security 36:4 124
32. David Stevenson, “Militarization and Diplomacy in Europe before 1914,” International Security,
Vol. 22, No. 1 (Summer 1997), pp. 125–161, at p. 154.
33. Levy, “Case Studies.”
34. Ian F.W. Beckett, The Great War, 2d ed. (London: Pearson, 2007), p. 5.
mains that, despite rising tensions in the West, the war began on the banks of
the Danube, not the Rhine. Interdependence may have played a more important
and nuanced role than is widely recognized, as exhibited by the pattern of
initiation, joining, and nonjoining by European states.
Our research design is intended to provide a more comprehensive test
of the question of whether the outbreak of World War I constitutes a failure of
interdependence. To answer this question in a manner that overcomes the
shortcomings outlined above, we address two sub-questions. First, is there historical
evidence that interdependence made the war less likely than it would
have been otherwise? Second, if so, why did war break out in 1914 despite
these effects? We begin by analyzing crisis behavior in Europe in the era
prior to World War I. This allows us to observe multiple cases of crises that
vary both along the dependent variable of crisis outcome (e.g., escalation or
de-escalation) and the key independent variable of interdependence. Although
it is possible to deªne any previous point in history as beginning the lead-up
to the war, many historians begin their study of the war’s origins in 1871, following
Germany’s victory in the Franco-Prussian War. Crises among the
highly interdependent powers—most importantly, Britain, France, Germany
and Russia—were generally resolved peacefully during this era, despite important
differences among these powers. By contrast, crises among the less interdependent
powers of Europe in this period often led to war.
We continue our analysis by examining the July 1914 crisis. We do so for two
reasons. First, as mentioned above, regardless of the methodological arguments
one might make to the contrary, it is important to understand why
interdependence failed to prevent a wider war in 1914. Second, deviant case
analyses can be useful for the reªnement and extension of hypotheses.35
By examining
July 1914 within the broader context, we hope to provoke a reªnement
or extension of liberal theory.
While our research design has several advantages, it also has limitations.
These limitations are not unique to our study, but are characteristic of any attempt
to draw inferences in small samples. We cannot conclusively demonstrate
that interdependence “worked” in 1914, but rather that the widely
accepted view that interdependence failed to promote peace is poorly sup-
ported.36
In an ideal world, we could go further and “process trace” our
Trading on Preconceptions 125
35. Imre Lakatos, “Falsiªcation and the Methodology of Scientiªc Research Programmes,” in
Lakatos and Alan Musgrave, eds., Criticism and the Growth of Knowledge (New York: Cambridge
University Press, 1970), pp. 91–196; and Levy, “Case Studies.”
36. Liberal theory does not argue that noninterdependence is a cause of war (i.e., increasing the
likelihood of conºict), but rather that interdependence is a cause of peace, reducing motivations to
ªght below what they might be otherwise.
argument, providing detailed narratives showing that de-escalation among interdependent
states in crisis resulted (at least in part) from concern among key
decisionmakers over jeopardizing economic relations. Beyond the fact that we
lack space to conduct such analyses here, evidence of this type may not be
forthcoming. A lack of discussions about trade, for example, could be interpreted
as proof that leaders were unaffected by economic considerations. The
leaders most affected by economic ties, however, may refrain from entering
crises altogether, making assessments of their (non)actions far from transparent.
Few would ªnd it surprising if leaders of less interdependent states that
escalated to war failed to include in their reasons for ªghting their lack of economic
interdependence, but leaders who fail to ªght for economic reasons
may end up attributing their timidity to discretion or diplomacy. Leaders may
also adopt measures that mitigate the effects of interdependence, such as tight
alliance ties. In short, the absence of evidence cannot be interpreted as evidence
of absence. The conventional narrative on economic interdependence
and World War I relies on observable behavior (i.e., there was a war, thus economic
factors failed to deter war) to judge the inefªcacy of economic interdependence,
so we also focus on such behavior to support our argument, at
least initially.
One thing we can demonstrate is that a correlation exists between interdependence
levels and conºict in Europe during this era.37
The correlation is important
for two reasons. First, it casts prima facie doubt on the view of 1914 as
an adequate “most likely” test of commercial liberalism. If it is truly a most
likely case with which to test the theory, appropriately deªned in terms of the
independent variable, then a ªnding that crises among interdependent states
during this era tended to result in conºict would provide strong evidence
against liberal theory. If they did not, however, then this would suggest that
the use of World War I to falsify commercial liberalism is myopic and that the
role of interdependence in the origins of the war deserves greater attention.
Second, the correlation is useful for counterfactual analysis. By identifying the
most salient sources of tension during the era, we can predict ex ante where
war would have been most expected if interdependence and war are unrelated.
By then comparing crisis behavior in response to such tensions among
highly and weakly interdependent groups of states, we can use the groups as
counterfactuals for each other.
International Security 36:4 126
37. Patrick McDonald and Kevin Sweeney have already demonstrated that states with higher levels
of trade protection were more likely to engage in conºict in 1914. McDonald and Sweeney,
“The Achilles’ Heel of Liberal IR Theory?”
Interdependence in Pre–World War I Europe
We deªne “economic interdependence” broadly, as consisting of both trade
and capital ºows. Groups of interdependent states are those connected by
dense trade and capital ºows, whereas less interdependent states and groups
are closer to autarky. Although data are not available for every state we address
in each year, we possess enough information to develop key conclusions
regarding categories of states, their relative levels of interdependence, and the
effects of economic linkages on conºict in this period.
It is possible to think of Europe during this era as consisting of two distinct
economic subsystems. The ªrst is a highly interdependent group consisting
mostly of Western powers, most importantly, Germany, France, and the United
Kingdom, but also Russia. The second subsystem is a signiªcantly less interdependent
group of Eastern European states, including the Ottoman Empire, its
successor states, and again Russia. Scholars generally agree that, between 1871
and the beginning of World War I, many European powers became increasingly
economically interdependent, particularly in terms of increased trade
and capital mobility. The highly interdependent states in Europe were mostly
in the West, including Belgium, Britain, Denmark, France, Germany, Holland,
Norway, and Sweden. Other powers, such as Austria-Hungary, Italy, and
Russia were signiªcantly less developed economically and less integrated into
the global economy, although Russia was heavily dependent in economic
terms on the Western powers. The least economically integrated and developed
of the European powers at the time were the Ottoman Empire and its
various successor states, as well as the Iberian states. The key Western powers
of Britain, France, and Germany constituted a highly economically interdependent
group to which Russia was also attached. Austria-Hungary was connected
to this group, largely via its economic relations with Germany, but its
economic ties to other Western nations were far less signiªcant. By contrast, no
such economically interdependent group existed in the East. The key Eastern
powers—Austria-Hungary, the Ottoman Empire and its successor states,
and Russia—conducted relatively little of their economic relations with one
another.38
Trading on Preconceptions 127
38. Albert Fishlow, “Lessons from the Past: Capital Markets during the 19th Century and the
Interwar Period,” International Organization, Vol. 39, No. 3 (Summer 1985), pp. 383–439; Peter J.
Katzenstein, “International Interdependence: Some Long-Term Trends and Recent Changes,” International
Organization, Vol. 29, No. 4 (Autumn 1975), pp. 1021–1034; Ripsman and Blanchard,
“Commercial Liberalism Under Fire”; Jeffrey G. Williamson, “Globalization and the Great Divergence:
Terms of Trade Booms, Volatility, and the Poor Periphery, 1782–1913,” European Review of
We use the trade levels of the key powers during this era to illustrate the two
European subsystems. Several dimensions of monadic economic development
and dyadic economic relations are argued to reduce the risk of conºict, perhaps
the most analyzed of which is trade. As ªgures 1 and 2 illustrate using
data from the Correlates of War (COW) project,39
Britain, France, and Germany
were the key trade powers in Europe. In a second tier were Austria-Hungary,
Italy, and Russia, all with trade levels of more than $1 billion (2008 U.S. dolInternational
Security 36:4 128
Economic History, Vol. 12, No. 3 (December 2008), pp. 355–391; and Christopher Blattman, Jason
Hwang, and Jeffrey G. Williamson, “Winners and Losers in the Commodity Lottery: The Impact of
Terms of Trade Growth and Volatility in the Periphery 1870–1939,” Journal of Development Economics,
Vol. 82, No. 1 (January 2007), pp. 156–179.
39. Katherine Barbieri, Omar Keshk, and Brian Pollins, Correlates of War Project Trade Data Set
Codebook, ver. 2, 2008.
Figure 1. Trade Levels in Western Europe, 1870–1913
lars) by the outbreak of the war. The Ottoman Empire and the Balkan states,
however, traded relatively little.
Similar trends can be observed in the levels of dyadic trade dependence during
this era. Scholars have debated whether dependence should be measured
as dyadic trade relative to income or dyadic trade relative to total trade.40
We
need not weigh in on this debate, however, as GDP data are generally unavailable
for the period.41
We deªne “trade dependence” for country i on country j
as:
Trading on Preconceptions 129
40. Gartzke and Li, “Measure for Measure”; and Katherine Barbieri and Richard Alan Peters II,
“Measure for Mismeasure: A Response to Gartzke & Li,” Journal of Peace Research, Vol. 40, No. 6
(November 2003), pp. 713–719.
41. Peter Katzenstein reports the following ratios of monadic trade to GDP for the pre–World
Figure 2. Trade Levels in Eastern Europe, 1870–1913
Dependenceijt ϭ
x m
TotalTrade
ijt ijt
i t
+
.
Table 1 provides the dyadic levels of trade dependence among the key states
in 1913. It should be noted that there are missing observations in the COW
data for this period, so we are not able to calculate trade dependence for all of
the relevant dyads.42
A few patterns are evident. First, the largest dyadic trade
relationships by sheer volume generally involve trade within Western Europe,
between the West and Russia, and between Germany and Austria-Hungary.
The dyadic data show that these powers were also highly trade interdependent.
By 1913, for example, Russia depended on Germany for 45 percent of its
trade, while Germany conducted 13 percent of its trade with Russia.43
The extent
of economic interdependence between Germany and Russia was a key
factor in shaping their crisis bargaining incentives, as we explain below.
Among the four Western powers and Russia, the average trade ºow in 1913
was $290 million (2008 U.S. dollars), while the average dyadic trade dependence
was 10.3 percent. Put another way, this group of states was highly interdependent.
By contrast, the volume of trade ºows between the Eastern powers
remained minor throughout this era, reaching an average in 1913 of $17 million
(2008 U.S. dollars). Most Eastern dyads were much less trade dependent
on each other, the few exceptions being countries that traded heavily with
Austria-Hungary. The average dyadic trade dependence among the Eastern
powers was only 5 percent in 1913, most of which resulted from Ottoman and
Bulgarian dependence on Austria-Hungary. Although we cannot calculate the
trade dependence between Austria-Hungary and Serbia using these data,
International Security 36:4 130
War I period: France, 53.7 percent; Germany, 38.3 percent; Italy, 28.1 percent; and the United Kingdom,
43.5 percent. Katzenstein, “International Interdependence,” p. 1032. Relying on newer data,
Paul Papayoanou reports that “[f]or the three biggest European powers, total trade (exports and
imports) as a percentage of gross national product (GNP) was 38 percent for Germany, 52 percent
for Great Britain, and 54 percent for France in the years leading up to World War I, and much of
their trade was with one another.” Papayoanou, “Interdependence, Institutions, and the Balance of
Power,” p. 55. Finally, Michael Huberman and Wayne Lewchuk rank sixteen European countries
(not including the Balkans) in terms of monadic trade over GDP in 1913. Only Portugal (13 percent)
ranks lower than Austria-Hungary (16 percent). Huberman and Lewchuk, “European Economic
Integration and the Labor Compact, 1850–1913,” European Review of Economic History, Vol. 7,
No. 1 (April 2003), pp. 3–41.
42. COW provides no data on Serbian trade in 1913. Monadic trade in 1912 is reported at only
$37 million (2008 U.S. dollars). Barbieri, Keshk, and Pollins, Correlates of War Project Trade Data Set
Codebook.
43. According to Ian Beckett, “Between 1905 and 1914 Russo-German trade was immensely important,
with Russia sending 44 percent of her exports to Germany and receiving 47 percent of her
imports from Germany.” Beckett, The Great War, p. 28. See also William Jannen Jr., The Lions of July:
Prelude to War, 1914 (Novato, Calif.: Presidio, 1996), p. 83.
the consequences of the Pig War, discussed below, suggest that trade was low
between them by this period.
European Crises, 1871–1913
Wars often appear inevitable in retrospect, but few wars seemed more unlikely
to contemporaries in the preceding years than World War I. Journalists at
the time, and historians in later periods, emphasize that Europe, or at least the
West, had largely discounted the danger of a major conºagration by the summer
of 1914. Why should this be so? Certainly, European history did not suggest
that peace was inevitable, or even particularly durable. Warfare had been
such a regular feature of European politics that it is hard to escape the conviction
that war was overdue. One reason for optimism in the West about the durability
of peace was that stability had been tested several times and found to
Trading on Preconceptions 131
Table 1. Dyadic Trade Dependence in 1913
Country A Country B
Total Dyadic
Trade
Dependence
A-B
Dependence
B-A
Bulgaria Ottoman Empire 1.98 2.78403 0.69814
Italy Bulgaria 1.98 0.16915 2.78403
France Bulgaria 3.37 0.11408 4.73847
United Kingdom Bulgaria 4.67 0.08 6.56637
Bulgaria Russia 6.05 8.50675 0.41027
Germany Bulgaria 10.15 0.20499 14.27165
Austria-Hungary Bulgaria 12.93 1.04397 18.18054
Russia Ottoman Empire 27.62 1.87299 9.73873
Italy Ottoman Empire 28.94 2.47238 10.20415
France Austria-Hungary 31.96 1.08194 2.58046
France Ottoman Empire 34.14 1.15574 12.03766
Austria-Hungary Ottoman Empire 37.14 2.99869 13.09545
Germany Ottoman Empire 40.89 0.82582 14.41769
Austria-Hungary Russia 51.45 4.15408 3.48896
Italy Russia 53.68 4.58596 3.64019
United Kingdom Austria-Hungary 57.92 0.99221 4.67647
United Kingdom Ottoman Empire 60.85 1.04241 21.45552
Austria-Hungary Italy 92.35 7.45636 7.88959
France Italy 100.28 3.39478 8.56706
France Russia 117.42 3.97502 7.96257
United Kingdom Italy 148.7 2.54734 12.70365
Germany Italy 191.83 3.87423 16.3883
United Kingdom Russia 276.28 4.73288 18.73529
France Germany 344.71 11.66946 6.96181
United Kingdom France 416.4 7.13324 14.09638
Germany Austria-Hungary 458.5 9.25993 37.01939
United Kingdom Germany 583.31 9.99253 11.78061
Germany Russia 670.34 13.53828 45.45757
be robust. A series of crises involving Germany, France, Britain, and Russia repeatedly
seemed to take Europe to the brink of war, only to abate as the
beneªts of existing commercial relations asserted themselves. This reinforced
the view that mutual interest, diplomacy, or civilization made it clear to affected
leaders that war was not in their interest.
In part for these reasons, few conºicts have received more careful attention
from historians and students of international affairs than World War I. So
much has been said and written about the Great War that key aspects are substantially
understood. The lasting interest in World War I clearly stems in part
from its size. But this “war to end all wars” grew by increments, after the perennial,
if much smaller, conºict in the Balkans had already established itself.
Yet, it is still important to remember that the Great War did begin in the
Balkans. If one wants to know why World War I became so large, then it is
reasonable to explore joining by the major European powers. If instead one
wants to know how and why World War I began, as a war, then it makes
more sense to focus on Serbia and Austria-Hungary. The point is that war in
the Balkans was no surprise to anyone and requires no special explanation.
The same era in which disputes tended to be defused in the West, conºict on
the edges of Europe ºared up repeatedly. The rest of Europe had managed to
keep these contests localized, however. The expansion of a Balkan war is what
citizens in the West had discounted, and when the expansion came, it came as
a surprise.
the highly interdependent subsystem
The late nineteenth and early twentieth centuries in Europe were marked by
considerable tensions and a number of crises, including crises both between
states that were highly economically interdependent and between those that
were minimally so. Three major sources of tension were at the root of these crises,
and each was exacerbated by changes in the balance of power. The ªrst
source of tension was Germany’s occupation of Lorraine and (to a lesser degree)
Alsace, both captured from France during the Franco-Prussian War. To
many in France, relations with the Reich could not be normal so long as these
territories remained in German hands. A second source of tension involved the
competing colonial ambitions of the major powers. Few territories were left to
conquer, even as Germany increased its drive for colonial expansion after uniªcation.
Finally, after decades of decline, the Ottoman Empire no longer possessed
the strength to maintain dominance over its European possessions.
This, and the accompanying rise of Balkan nationalism, threatened not only
Turkey’s interests but also those of the other powers in the region, especially
Austria-Hungary and Russia.
International Security 36:4 132
Changes in the balance of power, and the competing ambitions associated
with them, led to frequent clashes among European states, but these disputes
generally did not escalate to war when conºicts involved states that were economically
interdependent. To begin with, despite the bitter rivalry between
France and Germany, the two powers remained at peace for more than forty
years after the Treaty of Frankfurt. What makes the era so remarkable is that
both nations spent these decades girded for war, having created, established,
and maintained elaborate plans for invading each other.
Signiªcantly, the competition for colonies among the European powers led
to a series of crises. It should be noted that it was the interdependent Western
powers that featured most prominently in the struggle for colonial empire. Yet,
none of these crises resulted in a European war. Britain and France, long-time
rivals both on the continent and in terms of their global possessions, experienced
two near-war crises during this era. In 1893 France sent gunboats to
Bangkok to demand that Siam cede a large portion of its territory (most of
which later became Laos). Siam, a British client state, requested support from
London. Rather than backing its protégé against its chief historical adversary,
Britain prevailed on the Siamese to cede the disputed region to the French. In
return, France promised to make no future demands for additional territory in
the region, relieving the danger of future disputes among the interdependent
European powers at the expense of English prestige and French ambition.
Again, in 1898 Britain and France nearly went to war, this time as a result of
the Fashoda crisis. A standoff between the two countries was caused by competing
attempts to extend their spheres of inºuence in East Africa. After several
months, the French withdrew their troops, and the governments agreed
on a division of the region among themselves, ending the last major source of
colonial tensions between them.44
In both crises, one of the major powers
backed down or declined to intervene, apparently deciding that war was not
worth the cost.
Tensions between France and Germany were also heightened by colonial
competition. First, in 1905, Kaiser Wilhelm I visited Tangier and gave a speech
promoting Moroccan independence, provoking the French authorities. France
moved troops to the German border while Germany called up its reservists,
but the crisis was resolved through an international conference before it became
highly militarized. The Germans reluctantly backed down in humiliation
when only Austria-Hungary supported their position. As David Herrmann
writes, “The Moroccan confrontation . . . marked the beginning of a series of
Trading on Preconceptions 133
44. James Lee Ray, Democracy and International Politics: An Evaluation of the Democratic Peace Proposition
(Columbia: University of South Carolina Press, 1995), pp. 175–200.
diplomatic trials of strength that continued until the outbreak of World War I,
each escalating from its predecessor and helping to drive the European powers
into two opposing and increasingly armed camps.”45
A second serious
Moroccan crisis erupted in 1911 when Germany sent the gunboat Panther to
Agadir in response to France’s attempt to take full control of Morocco by advancing
15,000 troops to Fez. Even if the war could be contained, its effects
on European commerce and capital ºows could be catastrophic, as many in
Europe recognized. Again, the European powers negotiated, and again the
Germans backed down. Neither the kaiser nor German Chancellor Theobald
von Bethmann-Hollweg was interested in war over Morocco. L.C.F. Turner
writes that three factors contributed to this decision: (1) the German navy was
relatively weak and could not take on the French, who were supported by the
British; (2) neither Austria-Hungary nor Russia had a signiªcant interest in
Morocco, allowing the conºict to be settled among the three Western powers;
and (3) “French withdrawal of short-term loans from the Berlin market led to
heavy falls on the stock exchange and demands from German ªnanciers for a
lessening of tension.”46
The latter factor provides a strong indication that close
economic relations between France and Germany contributed to preventing
war. As Stevenson notes, “[O]nce Wilhelm II had ruled that war must be
avoided, Germany again accepted a diplomatic setback in preference to provoking
hostilities or to blufªng.”47
Disputes between the other interdependent powers were also generally
resolved peacefully during this period. For example, in the Dogger Bank incident
during the Russo-Japanese War, the Russian navy ªred on several British
ªshing boats in the North Sea, mistaking them for Japanese men-of-war after
receiving false reports of a Japanese naval presence in the area. Although, as
Joachim Remak argues, “[I]f some sort of apparatus for the measurement of
hostility between nations were to exist, Anglo-Russian relations, in the half
century or so before 1904, might receive its highest score,”48
the British chose
not to retaliate. This decision is particularly striking given that the British were
formally in alliance with the Japanese (although Britain was not committed to
providing support unless Japan became involved in a conºict with two other
International Security 36:4 134
45. David G. Herrmann, The Arming of Europe and the Making of the First World War (Princeton, N.J.:
Princeton University Press, 1996), p. 37. See also Stevenson, “Militarization and Diplomacy in Europe
before 1914”; Jeffrey W. Taliaferro, Balancing Risks: Great Power Intervention in the Periphery
(Ithaca, N.Y.: Cornell University Press, 2004), pp. 55–93; and Richard F. Hamilton, “The European
Wars: 1815–1914,” in Hamilton and Holger H. Herwig, eds., The Origins of World War I (Cambridge:
Cambridge University Press, 2003), p. 85.
46. L.C.F. Turner, Origins of the First World War (New York: W.W. Norton, 1970), p. 19.
47. Stevenson, “Militarization and Diplomacy in Europe before 1914,” p. x.
48. Joachim Remak, The Origins of World War I: 1871–1914 (Hinsdale, Ill.: Dryden, 1967), p. 40.
powers). Had the British retaliated, leading to a war with Russia, we can easily
imagine many analysts calling such a war “inevitable” in retrospect. Similarly,
although the British-German rivalry grew in intensity during this period, fueled
partly by Germany’s post-uniªcation ambitions for colonial empire, this
rivalry did not result in war until 1914. On multiple other occasions—most notably,
the Boer War, in which Germany chose to remain neutral—disputes that
could well have led to a general war were resolved peacefully.
Why, then, did these powers, which had motives for war along with detailed
plans for ªghting, refrain from doing so for several decades? There were likely
many causes of the enduring peace among these powers. Economic relations
were a key factor in maintaining the peace. A.J.P. Taylor notes that “[t]he decisive
cause was, no doubt, economic. The secret that had made Great Britain
great was a secret no longer. Coal and steel offered prosperity to all Europe
and remade European civilization. The dream of Cobden seemed to have come
true. Men were too busy growing rich to have time for war.”49
Arguing against
the Marxist hypothesis that economic interests make war more likely, Niall
Ferguson notes that “there is scarcely any evidence that these interests made
businessmen want a major European war.”50
In addition, he points out that
“[i]f there was a war which imperialism should have caused it was the war between
Britain and Russia which failed to break out in the 1870s and 1880s;
or the war between Britain and France which failed to break out in the 1880s or
1890s.”51
Others go further, arguing that business interests actively opposed
war. Remak writes that “divergent national interests in the ªeld of commerce
did not make for armed conºict; in times of crisis, 1914 included, businessmen
on all sides were among the strongest advocates of peace.”52
Likewise, Richard
Hamilton argues that “[i]f business had been dominant, if ‘the bourgeoisie’
had been in power, the war would not have happened.”53
Relations between France and Germany provide a useful illustration. During
this era, ªnance and manufacturing became highly integrated; levels of
trade grew signiªcantly; and unprecedented amounts of capital ºowed across
their borders. A key example is the German ªnance of iron mining in the
French region of Longwy-Briey (a region the Germans later controlled during
Trading on Preconceptions 135
49. A.J.P. Taylor, The Struggle for Mastery in Europe, 1848–1918 (Oxford: Oxford History of Modern
Europe, 1954), p. 255.
50. Niall Ferguson, The Pity of War: Explaining World War I (New York: Basic Books, 1999), p. 32.
51. Ibid., p. 39.
52. Joachim Remak, “The Third Balkan War: Origins Reconsidered,” in Hannsjoachim W. Koch,
ed., The Origins of the First World War: Great Power Rivalry and German War Aims, 2d ed. (London:
Macmillan, 1984), p. 87.
53. Richard F. Hamilton, “On the Origins of the Catastrophe,” in Hamilton and Herwig, The Origins
of World War I, pp. 469–506, at p. 469.
the war to their strategic advantage). As Turner notes, “[S]o far from being bitter
rivals French and German capitalists were happy to cooperate.”54
Neither
nation was willing to concede key claims against the other; France wanted its
northeastern territories returned, and Germany was tired of being shut out of
colonial expansion. Still, neither would sacriªce mutual economic beneªts to
pursue these disputed objectives, at least not through direct confrontation and
not as isolated issues.
Other factors could have encouraged peace among the integrated powers
during this era. For example, it could be claimed that the integrated powers
had strategic incentives to capitalize on peace. As beneªciaries of the status
quo, perhaps Britain and France were simply seeking to perpetuate conditions
experienced in the Concert of Europe. Yet, if some Western nations were also
status quo powers, other integrated nations were revisionists. Germany was
unhappy with its limited allocation of colonies, for example, while Russia was
distressed with the treatment of its allies by Austria-Hungary. Even Britain
and France had differences in other regions, if not in Europe directly. Indeed,
Austria-Hungary demonstrates that a desire to maintain the conditions that
had brought peace in a previous epoch may well have exacerbated tensions in
the decades before World War I. The integrated powers of Europe responded
differently to the important tensions that did arise, whereas nonintegrated
powers in the same region reacted with force. The strategic argument thus
begs the question of why the highly integrated powers exhibited a preference
for peace, regardless of their status quo or revisionist preferences, but the less
interdependent powers, as shown below, did not. Maintaining the view that
strategic preferences accounted for peace before 1914 also requires explaining
why these preferences changed in 1914. Indeed, any theory of the prewar
peace must account for the changes that resulted in war in 1914. In the section
“The 1914 Crisis,” we explain what was different in 1914 in a manner that is
both consistent with the paciªc effects of economic relations, but also suggests
that the relationship between commerce and conºict is more nuanced than
many theories suggest.
the weakly interdependent subsystem
As the interdependent powers set a pattern of peace, other European states often
escalated their disputes to war, particularly in the Balkans. The underlying
reason for the tension in the Balkans during this era was the rapid decline of
the Ottoman Empire, which, accompanied by the rise of nationalism in the region,
led many Balkan nations to seek independence. In addition to Turkey’s
International Security 36:4 136
54. Turner, Origins of the First World War, p. 13.
interest in maintaining its European possessions, Russia and Austria-Hungary
also had key strategic interests in the region. Russia generally supported the
rise of Balkan nationalism because of its long-time rivalry with Turkey, its ambitions
for access to the Mediterranean, and the inºuence of pan-Slavism in
Moscow. The multinational Austria-Hungary, however, was threatened by rising
Balkan nationalism and faced a security dilemma, especially with respect
to Serbia.55
As L.L. Farrar writes, “[W]hat would preserve Austria-Hungary as
a great power would provoke Serbian nationalism, and what would satisfy
Serbian nationalism would threaten Austria-Hungary as a great power. Thus
coexistence seemed virtually impossible and conºict almost inevitable.”56
Again, however, what appears inevitable in retrospect may have been less so
in practice. If there was much to impel conºict in the Balkans, there was also a
considerable basis for clashes among the Western powers, as we have already
discussed. There were strong motives for war both in the West and in the
Balkans; what seems distinct about the Balkans was the general absence of a
positive motive for peace. Lacking the commercial incentives to cooperate that
saturated relations in the West, Eastern nations were prone to see the strategic
calculus as zero sum.
Other countries participated in driving or mediating festering Balkan disputes,
a process that increasingly enmeshed the foreign policies of European
powers with their Balkan protégés. Nations outside the region began to “tie
their hands” with alliance commitments to bolster the leverage of Balkan partners.
Ironically, the great German statesman Otto von Bismarck made it clear
that Germany’s interests in the region were decidedly limited: “For us, Balkan
questions can never be a motive for war.”57
Yet, the reputations of foreign
powers seemed increasingly to be tied to their actions in the region, even as
discretion over wider European foreign policy was increasingly handed over
to local ofªcials through tightening alliance commitments.
Several conºicts in the late nineteenth century exhibited the tendency of crises
to escalate to warfare in the region. Early conºicts included the Serbian and
Montenegrin war of independence against Turkey in 1876, the Russo-Turkish
War in 1878, the Serbo-Bulgarian War in 1885, and the Greco-Turkish War in
1897. In the direct lead-up to World War I, the ªrst crises of note were the Pig
War between Serbia and Austria-Hungary, followed by the Bosnian annexTrading
on Preconceptions 137
55. Benjamin Miller and Korina Kagan, “The Great Powers and Regional Conºicts: Eastern Europe
and the Balkans from the Post–Napoleonic Era to the Post–Cold War Era,” International
Studies Quarterly, Vol. 41, No. 1 (March 1997), pp. 51–85.
56. L.L. Farrar Jr., “The Limits of Choice: July 1914 Reconsidered,” Journal of Conºict Resolution,
Vol. 16, No. 1 (March 1972), pp. 1–23, at p. 10.
57. Quoted in Remak, The Origins of World War I, p. 14.
ation crisis. After Serbia increased French imports and created a customs
union with Bulgaria, Austria-Hungary responded in 1906 by refusing to import
Serbian livestock. Serbia found markets elsewhere, consequentially dramatically
reducing Austro-Serbian trade,58
which, in turn, meant that trade
was unlikely to prevent war between the two states.
The Pig War also led to Austria-Hungary’s decision in 1908 to annex Bosnia
and Herzegovina, both of which it had administered under the 1878 Treaty of
Berlin. Serbia had sought to use Bosnia and Herzegovina as an outlet to the
Adriatic in large part to overcome the Austro-Hungarian blockade of Serbian
goods. Austrian annexation understandably threatened both Russia, which
would have preferred that the territories remain independent, and Serbia,
which preferred them to be Serbian. Russia, still recovering from its war with
Japan, was in no position to ªght.59
Russian ofªcials actually took the extraordinary
step of informing the Austrians of their intention not to intervene even
if Austria-Hungary occupied Belgrade.60
Germany notiªed Russia that it
would declare war on Russia and France in the event of Russian mobilization.
As a result, Russia backed down, forcing Serbia to withdraw its protest of the
annexation. Annika Mombauer writes that “[g]iven the fact that Germany
gave unconditional support for Austria-Hungary over this Balkan matter, it
was primarily Russia’s mediating inºuence on Serbia that prevented war on
this occasion.”61
The Bosnian crisis is notable for at least three reasons. First, Russia suffered
a signiªcant humiliation by backing down, both in the eyes of the major powers
and with respect to Serbia, which began to doubt Russia’s support for its
ambitions.62
Second, this crisis, like that of Morocco in 1905, indicated to participants
that the side showing greater resolve got its way.63
Indeed, Germany
could reasonably draw the conclusion that, if it committed to war, Russia was
likely to back down, and conºict would be averted. Third, the crisis is an example
of conºict between the Triple Alliance and the Triple Entente being
courted by Austria-Hungary rather than by Germany. As Laurence Lafore arInternational
Security 36:4 138
58. Richard C. Hall notes that the Pig War “ended Serbian economic dependence on AustriaHungary.”
Hall, “Serbia,” in Hamilton and Herwig, The Origins of World War I, p. 92.
59. G.J. Meyer, A World Undone: The Story of the Great War, 1914 to 1918 (New York: Bantam Dell,
2006), pp. 21–22; and Herrmann, The Arming of Europe and the Making of the First World War,
pp. 117–122.
60. Stevenson, “Militarization and Diplomacy in Europe before 1914,” p. 135.
61. Annika Mombauer, The Origins of the First World War: Controversies and Consensus (London:
Pearson Education, 2002), p. 10
62. Remak, The Origins of World War I, p. 48; and Ronald Bobroff, “Behind the Balkan Wars:
Russian Policy toward Bulgaria and the Turkish Straits, 1912–13,” Russian Review, Vol. 59, No. 1
(January 2000), pp. 76–95.
63. Stevenson, “Militarization and Diplomacy in Europe before 1914,” p. 135.
gues, the Dual Alliance tied Germany “to a state whose future involved securities
of a very different kind from Germany’s.”64
This point is crucial as it
demonstrates that Germany’s foreign policy was becoming increasingly tied to
that of Austria-Hungary.65
The immediate consequence of rising tensions in the Balkans was a pair of
regional wars fought in 1912 and 1913. In 1912 a coalition made up of Bulgaria,
Greece, Montenegro, and Serbia declared war on Turkey, seeking to drive the
Turks out of Europe. Many observers in Europe predicted the conºict would
spread to the rest of the continent, especially because Austria-Hungary was expected
to intervene in an effort to repel Serbia from the Sanjak of Novibazar, a
region it had previously administered under the Treaty of Berlin and which
Serbia captured from the Ottomans during the ªghting. Similarly, Russia was
expected not to back down, as it had earlier in the Bosnian crisis.66
A general war was again averted, this time because Germany declined to
support Austria-Hungary’s ambitions and because Russia eventually also
backed down. At ªrst, Russia mobilized forces along its Western frontier (although
Russia claimed this was unrelated to the war). In Germany, however,
the leadership, apparently maintaining Bismarck’s disinterest in the Balkans,
decided not to support Austria-Hungary. Once the Germans made their decision,
the Russians also withdrew, perhaps in part because Germany’s abandonment
of Austria-Hungary allowed Russia to back down while still saving
face. Thus, despite an interest in avoiding further humiliation, the Russians
told the French that “even if Austria should attack Serbia, Russia will not
ªght.”67
A similar crisis arose in 1913 when the winners of the First Balkan War
fought over the spoils captured in the previous conºict. Speciªcally, Bulgaria,
Greece, and Serbia each fought for a greater share of Macedonia, with the war
later expanding to include Montenegro, Romania, and the Ottoman Empire.
Austria-Hungary, threatened again by the specter of Slavic nationalism, reacted
slowly to unfolding events, but eventually mobilized troops along the
Serb and Russian borders. Once again, the conºict was prevented from spreadTrading
on Preconceptions 139
64. Laurence Lafore, The Long Fuse: An Interpretation of the Origins of World War I (London:
Weidenfeld and Nicolson, 1966), p. 97.
65. On the Bosnian crisis generally, see ibid., pp. 151–159; Remak, The Origins of World War I,
pp. 46–48; and Stevenson, “Militarization and Diplomacy in Europe before 1914.”
66. Lafore, The Long Fuse, p. 170. James Joll notes that “failure to support Serbia again would
mean, the Russians thought, the end of Russian prestige in the Balkans and the beginning of a possible
new diplomatic alignment there.” Joll, The Origins of the First World War, p. 55.
67. Quoted in William C. Wohlforth, “The Perception of Power: Russia in the Pre-1914 Balance,”
World Politics, Vol. 39, No. 3 (April 1987), pp. 353–381, at p. 359. Turner also argues, “It therefore
appears that if Russia had resorted to military measures in support of Serbia’s claim to an Adriatic
port then a world war would have broken out in 1912.” Turner, Origins of the First World War, p. 43.
ing through the diplomacy of the interdependent states. James Joll argues that
Britain was successful in negotiating the end of the conºict and the independence
of Albania because, “to the annoyance of the Austrians, the Germans decided
that they would not put their whole weight behind the Austrian efforts
to limit Serbia’s gains.”68
The two Balkan Wars are signiªcant for several reasons relevant to the outbreak
of war in 1914. First, building on the earlier conºicts in the region, they
showed that Balkan disputes were generally settled by ªghting, a pattern far
different from that of the interdependent powers. Indeed, the interdependent
powers were often critical in mediating disputes and in helping to terminate
recurrent regional wars. Second, the cycle of Balkan conºicts showed the
potential for small, less interdependent states to exploit security relationships
with more interdependent allies. Alliances pose the risk that a state
may be entrapped into war.69
Balkan states used to their advantage the rivalry
between Austria-Hungary and Russia, as well as the desire of Britain,
France, and Germany to avoid a major war.70
Third, the crises showed that
Germany and Russia were the pivotal states in the alliance system; their decisions
on whether or not to support their smaller allies were the key to limiting
or expanding the size of wars.71
Joll notes that “the Balkan quarrels had not
escalated into a European war because the Germans were not prepared to
give their ally a free hand against Serbia.”72
Yet the crises did not escalate because
the Russians also appeared willing to back down in certain situations.
Fourth, the Balkan Wars continued a trend of increasingly militarized conºicts
in which the more militarized side was regularly the winner. This encouraged
International Security 36:4 140
68. Joll, The Origins of the First World War, p. 52. See also Sean M. Lynn-Jones, “Détente and Deterrence:
Anglo–German Relations 1911–1914,” International Security, Vol. 11, No. 2 (Fall 1986),
pp. 121–150; and Meyer, A World Undone, p. 25.
69. See, for example, Thomas J. Christensen and Jack Snyder, “Chain Gangs and Passed Bucks:
Predicting Alliance Patterns in Multipolarity,” International Organization, Vol. 44, No. 2 (Spring
1990), pp. 137–168; and David A. Lake, Entangling Relations: American Foreign Policy in Its Century
(Princeton, N.J.: Princeton University Press, 1999). For a similar argument, see Glenn H. Snyder,
“The Security Dilemma in Alliance Politics,” World Politics, Vol. 36, No. 4 (July 1984), pp. 461–495.
70. Samuel R. Williamson Jr. has argued that Austria-Hungary used the assassination of Archduke
Franz Ferdinand as a pretext for pulling Germany into a war with Serbia. Williamson, “Inºuence,
Power, and the Policy Process: The Case of Franz Ferdinand,” Historical Journal, Vol. 17, No. 4 (June
1974), pp. 17–34; and Williamson, “The Origins of World War I,” Journal of Interdisciplinary History,
Vol. 18, No. 7 (Spring 1988), pp. 795–818. Graydon A. Tunstall Jr. provides evidence that AustriaHungary
wanted a war, although a limited one, with Serbia. Turnstall, “Austria,” in Hamilton and
Herwig, The Origins of World War I, pp. 112–149. See also Remak, The Origins of World War I, p. 63.
71. Holger H. Herwig, The Marne, 1914: The Opening of World War I and the Battle That Changed the
World (New York: Random House, 2009), pp. 3–73; Joll, The Origins of the First World War, p. 53; and
Annika Mombauer, “German War Plans,” in Richard F. Hamilton and Holger H. Herwig, War
Planning 1914 (Cambridge: Cambridge University Press, 2010), pp. 48–79.
72. Joll, The Origins of the First World War, p. 53.
leaders on all sides to use brinkmanship tactics and to adopt commitment
mechanisms, such as ever tighter alliance ties, in the hope that their opponents
would be the ones to back down.73
Taken together, these effects of the Balkan Wars created increasing incentives
for the great powers to solidify their alliance relationships in the region. The
European alliance system was arguably created by the great powers as a deterrent
to reduce the likelihood of war. This was the case beginning with the
Bismarckian system created in the 1870s and 1880s, which was designed to
protect the newly uniªed Germany from its neighbors, but also with the Triple
Alliance and Triple Entente, designed in part to deter each other. The dilemma
facing the great powers in the early twentieth century was how to effectively
signal their resolve to ªght as a means to deter the other side.
Winning at the game of Chicken involves commitment. To increase their resolve,
great powers such as Germany and Russia were increasingly forced to
limit their options in crisis situations.74
As Thomas Schelling explains, decisionmakers
in a crisis can improve expected payoffs by constraining their ability
to act with discretion.75
German and Russian leaders tied their hands in
two ways during this era. First, they increasingly ceded decisionmaking authority
to the military by putting in place processes that, once an initial decision
was made to begin hostilities, made it costly for civilian leaders to back
down. As a result, Joll writes, “[T]he general staffs were taking decisions
which often committed them to irreversible military actions if war threatened:
and consequently in a crisis the freedom of action of the civilian ministers was
often more circumscribed than they themselves realized.”76
The second way in
which German and Russian leaders limited their options was by tightening alliance
commitments. For example, after failing to support its ally in the Second
Balkan War, G.J. Meyer argues that German policy was that “[n]ever again
must Vienna have reason to doubt the value of its alliance with Germany.”77
The effect of tightening alliances, however, was that of “chain-ganging.”
Tighter alliance commitments allowed the weaker partner to draw on the
power of the stronger partner in crisis bargaining, thereby increasing the prospects
for diplomatic success.
This leverage, however, also increased the hazard that a great power would
Trading on Preconceptions 141
73. Many historians and political scientists have referred to crisis bargaining during this period as
a game of brinkmanship. See, for example, Sagan, “1914 Revisited”; Remak, “The Third Balkan
War,” p. 95; and Stevenson, “Militarization and Diplomacy in Europe before 1914,” p. 155.
74. For a similar argument, see Farrar, “The Limits of Choice.”
75. Thomas C. Schelling, The Strategy of Conºict (Cambridge, Mass.: Harvard University Press,
1960).
76. Joll, The Origins of the First World War, p. 56.
77. Meyer, A World Undone, p. 25.
be drawn into a larger dispute. John Maurer notes, for example, that “[i]t was
clear to Russian decision makers that a confrontation with Austria-Hungary
entailed running the high risk of conºict with Germany.”78
Attempting to increase
the leverage of their allies, the interdependent powers gradually signed
over their foreign policies to their allies in the Balkans, which lacked both economic
ties and interests in maintaining a wider peace in Europe. As a result,
German and Russian leaders faced an increasingly difªcult choice between
honoring alliance commitments and losing credibility in European affairs. The
former put in jeopardy the trade and capital networks that now fueled their increasingly
destructive military machines, while the latter threatened to nullify
their ability to achieve gains diplomatically, without the need to use force.
Credibility and support for their partners required increasingly binding commitments,
but these in turn created increasingly diametric alternatives for the
integrated powers. Backing down meant suffering increasing reputational
costs, whereas failing to do so required that their counterparts would suffer a
similar fate. Ironically, interdependence accentuated this logic, as the high
stakes for a contest and previous experience convinced both sides that their
opponents would blink ªrst. Again and again, Balkan crises failed to spread as
either Germany or Russia chose to back down. This reassured the interdependent
powers that the European system was robust to such crises and that, in
future crises, someone (else) would act with greater discretion.
We thus offer three main points about the period between the FrancoPrussian
War and World War I. First, disputes among the highly interdependent
powers were generally resolved peacefully in the pre–World War I era.
Second, disputes among the less interdependent powers generally escalated to
wars, except where the (interdependent) great powers saw it in their interests
to intervene to mediate disputes and prevent them from drawing in other
powers. Finally, as a result of a series of crises, leaders in Germany and Russia,
the pivotal states in the alliance system, increasingly found the need to bind
their fates, and the fate of Europe, to that of their allies Austria-Hungary and
Serbia. That a general war did not break out before 1914 is, in many ways, attributable
to German and Russian decisions to back down rather than to support
regional allies. This in turn led to excessive conªdence that wider war
would be averted, in no small part because all involved recognized the mutual
economic beneªt of avoiding a wider contest.
International Security 36:4 142
78. John H. Maurer, The Outbreak of the First World War: Strategic Planning, Crisis Decision Making,
and Deterrence Failure (Westport, Conn.: Praeger, 1995), p. 72.
The 1914 Crisis
In 1912 and 1913, Germany and Russia averted war. Why did 1914 end differently?
Although the conºict-reducing effects of interdependence may have
subsided with growing tensions on the continent, this in itself is a tacit acknowledgment
that interdependence inhibits conºict, at least for a while. Both
Germany and Russia honored their alliance commitments to the point of starting
a general war. German support for Austria-Hungary enabled the latter’s
action against Serbia, and the war would not have spread were it not for
Russian intervention.79
The question then becomes why Germany and Russia
made these decisions in 1914. Scholars have spent the better part of a century
analyzing the decisions that turned the crisis into a world war. Our aim here is
not to conduct such an analysis, but to interpret the sequence of events in
terms of the logic of escalation and crisis bargaining.
Two points are worth making. First, the series of crises leading to the war
created incentives for the key players, Germany and Russia, to appear willing
to escalate to a wider war, if necessary, in support of their Balkan allies. This, in
turn, increased uncertainty; both German and Russian ofªcials had reason to
suspect that their counterparts might be blufªng. Second, the paciªc effects of
interdependence had important consequences for crisis bargaining. Germany
and Russia each attempted to convince the other that it must back down by
showing increasing resolve to support its less interdependent ally. AustriaHungary
and Serbia had weaker economic disincentives to go to war, which is
precisely what made these tightening alliance commitments stronger signals of
resolve. Indeed, the need to forge ever tighter alliance ties may be explained in
part by the disincentives that interdependence posed for both threatening and
carrying out major war.80
Trading on Preconceptions 143
79. Gary Goertz and Jack S. Levy, “Causal Explanation, Necessary Conditions, and Case Studies,”
in Goertz and Levy, eds., Explaining War and Peace: Case Studies and Necessary Condition
Counterfactuals (New York: Routledge, 2007); Luigi Albertini, The Origins of the War of 1914, Vol. 2
(London: Oxford University Press, 1957), p. 162; David Stevenson, Cataclysm: The First World War
as Political Tragedy (New York: Basic Books, 2004), pp. 13–14; Hall, “Serbia,” pp. 92–111; Hew
Strachan, The First World War (London: Penguin, 2005), p. 12; Maurer, The Outbreak of the First
World War, p. 115; Meyer, A World Undone, p. 33; and Richard Ned Lebow, “Contingency, Catalysts,
and Nonlinear Change,” in Goertz and Levy, Explaining War and Peace, p. 102.
80. On the relationship among trade, alliances, and conºict, see Brian M. Pollins, “Does Trade Still
Follow the Flag?” American Political Science Review, Vol. 83, No. 2 (June 1989), pp. 465–480; Joanne
Gowa and Edward D. Mansªeld, “Power Politics and International Trade,” American Political Science
Review, Vol. 87, No. 2 (June 1993), pp. 408–420; James D. Morrow, Randolph M. Siverson, and
Tressa E. Tabare, “The Political Determinants of International Trade: The Major Powers, 1907–90,”
American Political Science Review, Vol. 92, No. 3 (September 1998), pp. 649–661; and Quan Li and
Europe in 1914 thus reveals a nuanced relationship between interdependence
and the likelihood of war. The logic of commercial peace ªnds signiªcant
support in the crisis bargaining witnessed in Europe leading up to the
war. At the same time, economic disincentives for war also led to military integration
designed in part to overcome these very same disincentives. Alliance
ties formed and strengthened between highly and weakly interdependent
powers led to decisionmaking that appeared to overlook economic ties, as
weakly interdependent powers were given increasing inºuence over the foreign
policies of nations that, by themselves, preferred commerce to conºict.
A few aspects of the game of Chicken that characterizes the crises leading
up to World War I are worthy of additional attention.81
There are two purestrategy
Nash equilibria in a standard two-player game of Chicken: one in
which player A swerves and one in which player B swerves.82
Yet, these equilibria
assume that players are fully informed about what the other will do. The
assumption of full information is inconsistent with the larger logic of brinkmanship
and tying-hands commitments through alliances. If the players are
uncertain about each other’s intentions, then a more complex set of outcomes
is possible. If, for example, A overestimates B’s willingness to swerve, then the
worst possible outcome can ensue.83
With incomplete information, equilibria
can occur in the game of Chicken in which neither player swerves, as it seems
happened with Germany and Russia in 1914.
More generally, the iterated game of Chicken played by Germany and
Russia leading up to the war created increased incentives to convince the other
to back down by signaling a resolve to ªght. Having learned in previous crises
that the other could be brought to heel, especially in the face of ªrm resolve,
each state was intent on pushing the other to its limit.84
Austrian leaders believed
that giving Serbia an ultimatum supported by Germany would reduce
the risk of Russian intervention.85
The kaiser, in a change in policy from the
Balkan Wars, decided to support Austria-Hungary unconditionally, recognizInternational
Security 36:4 144
David Sacko, “The (Ir)Relevance of Militarized Interstate Disputes for International Trade,” International
Studies Quarterly, Vol. 46, No. 1 (March 2002), pp. 11–43.
81. For a contrary view, see Stephen Van Evera, “Why Cooperation Failed in 1914,” World Politics,
Vol. 38, No. 1 (October 1985), pp. 80–117.
82. Anatol Rapoport and Albert M. Chammah, “The Game of Chicken,” American Behavioral Scientist,
Vol. 10, No. 10 (November 1966), pp. 10–28. A mixed-strategy Nash equilibrium exists where
each player swerves with some probability.
83. Glenn H. Snyder, “‘Prisoner’s Dilemma’ and ‘Chicken’ Models in International Politics,” International
Studies Quarterly, Vol. 15, No. 1 (March 1971), pp. 66–103.
84. Stevenson, “Militarization and Diplomacy in Europe before 1914.” Advances in military mobilization
logistics also enabled leaders to generate ever greater risks using brinkmanship tactics.
Beckett, The Great War, p. 29.
85. Williamson, “The Origins of World War I,” p. 810; and Joll, The Origins of the First World War,
p. 21.
ing that failure to do so risked the destruction of the alliance.86
As Jack Levy
notes, “German decision-makers hoped and expected that an Austrian fait accompli
against Serbia in the immediate aftermath of the royal assassination,
backed by German warnings to Russia, would minimize the likelihood of
Russian intervention.”87
German Foreign Minister Gottlieb von Jagow famously
remarked that “the more determined Austria shows herself, the more
energetically we support her, so much the more quiet Russia will remain.”88
Prior to the assassination of Archduke Franz Ferdinand, the kaiser met with
and urged Austria-Hungary to go to war with Serbia, believing that Russia
would stay out of the ªght.89
Russian ofªcials were convinced that Germany and Austria would relent if
pushed far enough. Having already experienced defeat in both the RussoJapanese
War and the 1908–09 Bosnian crisis, Russia did not want to appear
weak again.90
Vigorous Russian support, in turn, strengthened Serbian re-
solve.91
As Joll puts it, “[T]he Austrians had believed that vigorous actions
against Serbia and a promise of German support would deter Russia: the
Russians had believed that a show of strength against Austria would both
check the Austrians and deter Germany. In both cases the bluff had been
called, and the three countries were faced with the military consequences of
their actions.”92
Russia mobilized ªrst, and Germany responded by declaring war. Historians
have debated the reasoning for Germany’s decision for decades, mostly
famously following Joschka Fischer’s argument that Germany courted the
war and was largely responsible for it.93
We will certainly not settle this complex
and enduring debate here, but we can perhaps contribute two points.
Trading on Preconceptions 145
86. Farrar, “The Limits of Choice,” p. 11.
87. Levy, “Preferences, Constraints, and Choices in July 1914,” p. 160; and Stephen Van Evera,
“The Cult of the Offensive and the Origins of the First World War,” International Security, Vol. 9,
No. 1 (Summer 1984), pp. 58–107.
88. Quoted in Imanuel Geiss, The Outbreak of the First World War: Selected Documents (New York:
W.W. Norton, 1967), p. 123. Holger H. Herwig refers to German actions in 1914 as a “game of
bluff.” Herwig, “Germany,” in Hamilton and Herwig, The Origins of World War I, p. 177.
89. David Stevenson, The First World War and International Politics (Oxford: Oxford University
Press, 1988), pp. 25–26. See also Strachan, The First World War, p. 14.
90. D.C.B. Lieven, Russia and the Origins of the First World War (New York: St. Martin’s, 1983); Joll,
The Origins of the First World War, p. 55; William R. Thompson, “A Streetcar Named Sarajevo: Catalysts,
Multiple Causation Chains, and Rivalry Structures” International Studies Quarterly, Vol. 47,
No. 3 (September 2003), pp. 453–474; and Remak, The Origins of World War I, p. 136.
91. Stevenson, “Militarization and Diplomacy in Europe before 1914.”
92. Joll, The Origins of the First World War, p. 21. See also Lafore, The Long Fuse, p. 233.
93. On the Fischer controversy, see, generally, Koch, The Origins of the First World War; Mombauer,
The Origins of the First World War; and Matthew Stibbe, “The Fischer Controversy over German
War Aims in the First World War and Its Reception by East German Historians, 1961–1989,” Historical
Journal, Vol. 46, No. 3 (September 2003), pp. 649–668.
First, to the extent that Germany appeared to be pushing Austria-Hungary toward
war, our argument suggests that Germany may have done so not because
it sought war but because it sought to convince Russia to back down.
Second, once Russia had mobilized, Germany’s options were perhaps quite
limited. If a major power war were to be fought, Germany preferred that it be
fought immediately rather than waiting for the French and Russians to have
more time to mobilize.94
As Meyer points out, “An open-ended postponement
of hostilities . . . would have destroyed Germany’s chances of defeating
France before having to ªght Russia.”95
Indeed, as James Fearon argues,
“[P]art of what made the Russian mobilization in 1914 an informative signal of
Russia’s willingness to ªght was that it was under-taken in the knowledge that
it would increase Germany’s incentive to choose preemptive war.”96
Thus,
that the kaiser opted for war in July 1914 may not indicate that Germany had a
preference for war (which would be damaging to commercial liberal theory).
Rather, it may indicate that Germany opted for a war that seemed the best of
limited options. Third, the notion that the war happened primarily because
Germany wanted it is difªcult to reconcile with theory. If wars result from
asymmetric information, then it may have been the case that if Germany had
better information, it would not have wanted to go to war. In addition, to the
extent that the war was the result of commitment problems, it could be argued
that a Germany in decline would have a preference for war because its rivals
could not commit to refrain from attacking Germany later. Yet Germany during
this era was a rising, rather than declining, power.
To be clear, it was far from inevitable that both Germany and Russia would
back their allies in 1914. It was entirely possible that at least one patron
would back down again, as had happened previously. Yet, each succeeding crisis
increased the incentive for Germany and Russia to heighten their commitments,
adding to the cost of backing down and increasing the danger of war.
Interdependence could have reduced the likelihood of a major contest. Arguably
it did, which paradoxically caused the interdependent powers to seek
more credible ways to backstop the demands of their Balkan allies. Less interInternational
Security 36:4 146
94. Ferguson, The Pity of War, p. 153; David Stevenson, Armaments and the Coming of War
(Oxford: Clarendon, 1996), p. 418; Herrmann, The Arming of Europe and the Making of the First World
War, pp. 227–228; and Annika Mombauer, Helmuth von Moltke and the Origins of the First World War
(New York: Cambridge University Press, 2001), pp. 203–204. Keir Lieber provides a different argument:
“The newest evidence about decision making during the July crisis, however, suggests that
German leaders did not lose control of events on the eve of war, but rather capitalized on what
they viewed as a golden opportunity to start the war they wanted.” Lieber, “The New History of
World War I and What It Means for International Relations Theory,” p. 184.
95. Meyer, A World Undone, p. 93.
96. James D. Fearon, “Domestic Political Audiences and the Escalation of International Disputes,”
American Political Science Review, Vol. 88, No. 3 (September 1994), pp. 577–592, at p. 580.
dependent than their patrons, Austria-Hungary and Serbia had weaker disincentives
for going to war. Given rising attachment to globalization, adding
credibility to Russia’s and Germany’s claims required delegating signiªcant
decisionmaking authority to Serbia and Austria. As a result of these incentives,
the highly interdependent European subsystem became increasingly entangled
with the less interdependent European subsystem through alliance ties.
A wider war may still have been avoided, even with Germany and Russia
committed to an expanded Balkans conºict. Most important, Britain and
France had not yet joined the war. The question remained whether they, too,
would honor their alliance commitments.97
At least one determinant of the
conºict may have been the ambiguity of the British position on intervention
and the German assumption that Britain would remain neutral, in part to protect
its commerce.98
There is even evidence that the relationship between interdependence and
conºict remained in effect during the war. Consider the European states that
were convinced to enter the war after August 1914: Bulgaria, Greece, Italy,
Portugal, Romania, and Turkey. Although each joined the war for different reasons,
with the possible exception of Italy none of these states was signiªcantly
interdependent before the war began. Italy entered the war because of signiªcant
territories offered it in Libya, Eritrea, and Somaliland, suggesting that
a large offer such as this may have been required to outweigh the importance
of other economic concerns. Furthermore, although less attention is paid to
the European states that remained neutral, it is worth noting that among these
were Denmark, the Netherlands, Norway, and Sweden—all highly economically
interdependent states. Denmark deserves particular attention because it
had fought wars with Germany in the nineteenth century and the sovereignty
of Southern Jutland was unresolved, so a salient dispute existed between the
two countries. Yet, as Bent Bludnikow notes, Denmark maintained its neutrality
“to be able to enjoy the advantages of its extensive international trade and
shipping.”99
The Netherlands, whose neutrality was not a foregone conclusion,
also chose to remain neutral for economic reasons.100
Thus, a pattern
Trading on Preconceptions 147
97. Remak notes that “if, for instance, the Austrians had been certain beyond a doubt that an attack
on Serbia would be an attack on Russia, or if the Germans had been altogether convinced that
to invade France would, undoubtedly, involve them in war with England, might not history have
taken a happier course?” Remak, The Origins of World War I, p. 89.
98. Jack S. Levy, “The Role of Necessary Conditions in the Outbreak of World War I,” in Goertz
and Levy, Explaining War and Peace; but for a counterargument, see Trachtenberg, “The Meaning of
Mobilization in 1914,” pp. 135–136.
99. Bent Bludnikow, “Denmark during the First World War,” Journal of Contemporary History,
Vol. 24, No. 4 (October 1989), pp. 683–703, at p. 683.
100. Irmtraud N. Gallhofer and Willem E. Saris, “Strategy Choices of Foreign Policy Decision
emerged in which less interdependent states chose to enter the war, generally
seeking territorial gain, whereas interdependent states preferred to remain
neutral, enjoying lucrative commerce.101
No exception to the trend, the United
States attempted similarly to pursue trade and avoid the conºict for much of
the war, until this became impossible under the combined impact of Allied political
machinations and German submarine warfare.
Conclusion
One’s understanding of the relationship between interdependence and World
War I hinges on the following counterfactual question: If interdependence did
not reduce the likelihood of conºict, where would we have expected ªghting
during the era leading up to the war? As the analysis above suggests, rivalries
involving Britain, France, Germany, and Russia were among the most intense
of the period and could well have resulted in major contests, as they had in the
past and as was expected by many at the time. Yet, war did not begin among
the interdependent powers. During the period of greatest commercial expansion,
interdependent powers were less likely to go to war with each other,
despite having the means and the motives to do so—motives so strong that
few students of history would have been surprised if war had in fact broken
out on several occasions. Among weakly interdependent powers, in contrast,
wars took place frequently, in patterns that seem both timeless and familiar. If
interdependence is ineffective, therefore, we must ask ourselves why World
War I began among the less interdependent powers and not in the interdependent
West.
Just as interdependence is said to increase the incentives for peace by
making war more expensive, widening the difference in payoffs between
cooperating and not cooperating, so too alliances operate by making war less
costly, discouraging aggression or increasing leverage by making it more expensive
for allies to fail to intervene in contests involving security partners. It
is not clear which effect—the conºict-inhibiting impact of commerce or the
conºict-inducing effect of alliance ties—was more intense for the nations of
Europe in 1914. We can, however, make the following comparisons. First, the
beneªts of interdependence varied among countries, ranging from the nearly
International Security 36:4 148
Makers: The Netherlands, 1914,” Journal of Conºict Resolution, Vol. 23, No. 3 (September 1979),
pp. 425–445.
101. A signiªcant exception is Spain, which, although poor and minimally interdependent, remained
neutral. Its decision may have resulted from the fact that it had little to gain from the
conºict, being geographically isolated from the main fronts and having already lost all of its colonial
possessions (unlike Portugal, which fought to hold on to them).
autarchic relationships among Balkan powers to the signiªcantly integrated
commercial powers. Second, the costs of backing down for European powers
almost certainly increased over time, as leaders observed their opponents
backing down in previous crises and as alliance ties became increasingly leveraged
through tighter commitments. Russia and Germany repeatedly showed a
preference for discretion rather than mutual destruction. Yet, tightening alliance
ties increasingly vied with crosscutting economic dependencies for the attention
and loyalty of national leaders. Only as alliance ties solidiªed and
policymaking increasingly shifted to the less interdependent powers of the
Balkans did the interdependent powers appear to become less attentive to the
economic consequences of a wider war. Even this risk was discounted by key
actors who expected their opponents to crumble under pressure, in part because
interdependence increased the cost of war.
The analysis provided here suggests that, in contrast to conventional wisdom,
the 1914 crisis is a particularly weak case against commercial liberalism.
Economic interdependence signiªcantly affected crisis bargaining during the
lead-up to the war. The system of alliances, created to deter opponents and reduce
the likelihood of great power war, had an important and unintended consequence
that manifested itself through a series of growing crises: it created an
incentive for the leaders of the interdependent powers to shift foreign policy
discretion away from themselves and toward powers less closely integrated
into the economic system. Importantly, this logic implies that under certain
conditions states can take actions to avoid conºict that, paradoxically, may
make conºict more likely. States that are highly economically integrated
may thus have an incentive to integrate militarily with states that are weakly
integrated into the global economy. This in turn can produce unintended results.
We hope to explore this potential reªnement to commercial liberalism in
future research.
Our analysis further suggests that conventional methods of inquiry may
tend to ignore the kinds of relationships responsible for the outbreak of World
War I. Only by looking at the network of ties between highly and weakly interdependent
powers can scholars appreciate the role of this structure in the era’s
crisis behavior.102
Focusing solely on economic interdependence paints an inTrading
on Preconceptions 149
102. Paul W. Schroeder makes a similar point about the war: “The disaster in 1914 did not derive
therefore from a failure by industrialists to understand the political logic and requirements of economic
integration or even the failure or refusal of politicians, military men, various interest
groups, and broad publics to appreciate the long-range advantages of peaceful international cooperation
over unrestrained competition and conºict. It lay rather in the structure of international
politics.” Schroeder, “Economic Integration and the European International System in the Era of
World War I,” American Historical Review, Vol. 98, No. 4 (October 1993), pp. 1130–37, at p. 1134.
complete and misleading picture, omitting important ways in which the economically
developed major powers were connected to less interdependent
states through alliances. This level of complexity makes it difªcult for scholars
to create parsimonious theories, although several scholars have begun to use
network theory and methods to do just this.103
We hope scholars will build on
our argument to reªne claims about how economic interdependence interacts
with other ties to affect war and peace.
International Security 36:4 150
103. Ward, Siverson, and Cao, “Disputes, Democracies, and Dependencies”; Han Dorussen and
Hugh Ward, “Trade Networks and the Kantian Peace,” Journal of Peace Research, Vol. 47, No. 1
(January 2010), pp. 29–42; and Yonatan Lupu and Vincent A. Traag, “Trading Communities, the
Networked Structure of International Relations, and the Kantian Peace,” University of California,
2011.