American Political Science Review Vol. 99, No.

3 August 2005

Three Pillars of the Liberal Peace

S
ebastian Rosato (2003) finds the logic of the “democratic peace” flawed in his “The Flawed Logic
of Democratic Peace Theory,” and he cites my work and other studies as examples of the flawed
logic. Some of the logic he describes is flawed, and it may characterize some of the literature in
the wide field of “democratic peace,” but it is not the logic underlying the core of liberal peace theory.
Indeed, the persuasive core of the logic underlying the theory of liberal democratic peace is missing from
Rosato’s account. Republican representation, an ideological commitment to fundamental human rights,
and transnational interdependence are the three pillars of the explanation. The logic underlying the peace
among liberal states rests on a simple and straightforward proposition that connects those three causal
mechanisms as they operate together and only together, and not separately as Sebastian Rosato claims.

I
explain the persuasive core of the logic underly- say that the three causes explained liberal peace and
ing the theory of liberal democratic peace logic liberal war when, and only when, combined. Rosato’s
in three places. The two-part essay “Kant, Liberal critique of the work, nonetheless, rests on treating each
Legacies and Foreign Affairs” published in Philoso- of these factors—–“normative,” “institutional,” etc.—–in
phy and Public Affairs (1983) showed how Immanuel isolation as if they were sufficient.
Kant’s (1970) 1795 essay, “Perpetual Peace,” could be This is important because, in my view, no one of
constructed as a coherent explanation of two impor- the factors alone is a sufficient explanation of the lib-
tant regularities in world politics—–the tendencies of eral peace or liberal war. First, as Rosato correctly
liberal states simultaneously to be peace-prone in their suggests, there is no reason for all direct or indi-
relations with each other and war-prone in their re- rectly majoritarian governments to be peaceful toward
lations with nonliberal states. Republican representa- other majoritarian governments. Clearly, a democ-
tion, an ideological commitment to fundamental hu- racy of xenophobes or hyper-nationalists would ex-
man rights, and transnational interdependence are the ternalize their preferences. Anticipating Rosato’s cri-
three causal mechanisms of the explanation. These are tique, I (1997, chaps. 4 and 9) pointed out in Ways
Kant’s three “definitive articles”—–the constitutional, of War and Peace—–in chapters that discuss Rousseau
international and cosmopolitan laws—–of the hypothe- and Marx—–that democratic institutions are completely
tical peace treaty he asks states to sign. The first part of compatible with Realist foreign policy when prefer-
the two-part essay focuses on the liberal peace and its ences are integrally and exclusively nationalist and
Kantian sources. The second part of the two-part essay with Socialist solidarity and international class war-
focuses on exposing the dangers of liberal imperialism, fare when strictly egalitarian (and societies lack in-
liberal aggression and liberal appeasement (Rosato dividual liberties and private property). Jean Jacques
1996 cites the reprints of the two articles in Debating the Rousseau’s classic account of democratic theory, for
Democratic Peace). I also addressed these themes in the example, anticipates that democracies will be locked,
American Political Science Review in December 1986 as any Realist would agree, in a generalized “state
and distinguished Kantian “liberal internationalism” of war” with all other states, whether democratic or
from “liberal pacifism” and “liberal imperialism.” In not (Rousseau 1756/1917). If information flows across
1997, in Ways of War and Peace, I distinguished liberal- borders are limited, subject to manipulation and na-
ism from the two other major traditions of international tionalist myth-making, and each democracy culti-
thought, Realism and Marxism. vates a normative commitment to complete auton-
All three have one consistent and key argument: “No omy and self-help, democracies will be likely to clash
one of these constitutional, international or cosmopoli- (Mearsheimer 1990; VanEvera 1990).
tan sources is alone sufficient, but together (and only Second, there should be no expectation that a pop-
where together) they plausibly connect the character- ulation widely sharing liberal values associated with
istics of liberal polities and economies with sustained human rights norm will shape policy unless they have
liberal peace” ([1983a, 1983b] 1996, 27). I repeat the democratic representation with the transparency and
same sentence as the summary of the argument—–“No accountability that can shape public decision-making.1
single constitutional, international . . . ” in the Amer- And third, there is no guarantee that commercial
ican Political Science Review (Doyle 1986, 1162); and and other forms of interdependence will alone provide
the identical sentence (this time in italics for empha- material foundations for cooperation among societies,
sis) in Ways of War and Peace (1997, 284). I explicitly rather than for sources of imperial rivalry and fuel to
balance of power competition, unless trade and invest-
ment are part of a relationship of trust and respect.2
Michael W. Doyle is Harold Brown Professor, School of Interna-
tional and Public Affairs and Columbia Law School, Columbia Uni-
1 Mueller (1989) stresses the norms of peace in his explanation, but
versity, 1314 IAB, 420 West 118th Street, New York, NY 10027.
(md@[email protected]). also links these norms to democratic institutions.
The author thanks Robert Jervis and Frank Wayman for advice 2 Cobden (1901) is a classic source on the pacifying effects of trade.
on an earlier draft of this note. Russett and O’Neal (2001) and Gartzke, Li, and Boehmer (2001)

463
Three Pillars of the Liberal Peace August 2005

Rosato (2003) is right to criticize the logic of each of pacific union removes the occasion of wars among lib-
these strands of the “democratic peace” standing alone. eral states and not wars between liberal and nonliberal
Once combined, however, the three sources do help states, we need to shift our attention from liberal repre-
explain why liberal states maintain peace with each sentation to liberal principles and liberal interests, the
other and are, nonetheless, and for explicable reasons other two elements in the liberal explanation of peace
prone to war and imperialism with nonliberal states. and war. These latter two elements account for the
Given the absence of this explanation in Mr. Rosato’s purposes that representative processes promote and
(2003) critique, it is worth briefly summarizing the three what credible signaling needs to signal.
hypotheses here. Each emphasizes one aspect of what Second, liberal principles add the prospect of inter-
characterizes a liberal republic. national respect. Liberal principles, or norms, involve
First, republican representative democratic govern- an appreciation of the legitimate rights of all individ-
ments tend to create an accountable relationship be- uals. Connecting these principles to public policy re-
tween the state and the voters, particularly median vot- quires publicity. Domestically, publicity helps ensure
ers. They preclude monarchs or dictators turning their that the officials of republics act according to the prin-
potentially aggressive interests into public policy while ciples they profess to be just and according to the
assuming that the costs will be borne by a subordinate interests of the electors they claim to represent. In-
public. Democratic representation introduces repub- ternationally, free speech and the effective communi-
lican caution, Kant’s (1970) “hesitation,” in place of cation of accurate conceptions of the political life of
autocratic caprice. Representative government allows foreign peoples are essential to establish and preserve
for a rotation of elites. This encourages a reversal of the understanding on which the guarantee of respect
disastrous policies as electorates punish the party in depends.
power with electoral defeat. Legislatures and public These principles begin the differentiation of policy
opinion further restrain executives from policies that toward liberal and nonliberal states, requiring trust of
clearly violate the obvious and fundamental interests and accommodation toward fellow liberals and pro-
of the public, as the public perceives those interests. ducing distrust of and opposition toward nonliberals.
As importantly, representation together with trans- Domestically just republics, which rest on the consent
parency (what Kant [1970] called “publicity”) may pro- of free individuals, presume foreign republics to be
vide for effective signaling, assuring foreign decision also consensual, just, and therefore deserving of the ac-
makers that democratic commitments are credible be- commodation that the individuals that compose them
cause rash acts and exposed bluffs will lead to electoral deserve. The experience of cooperation helps engender
defeat. Able to make more credible commitments than further cooperative behavior when the consequences
regimes with more narrow selectorates, democracies of state policy are unclear but (potentially) mutually
would thus be less likely to stumble into wars.3 beneficial. At the same time, liberal states assume that
We should not, however, overemphasize rational sig- nonliberal states, which do not rest on free consent,
naling. The division of powers and rotation of elites are not just. Because nonliberal governments are per-
characteristic of republican regimes can permit mixed ceived to be in a state of aggression with their own peo-
signals, allowing foreign powers to suspect that execu- ple, their foreign relations become, for liberal govern-
tive policies might be overturned by legislatures, courts, ments, deeply suspect. In short, fellow liberals benefit
or the next election. On the other hand, the shared from a presumption of amity; nonliberals suffer from
powers of republics should encourage better chances a presumption of enmity. Both presumptions may be
for deliberation. Most importantly, the combination of accurate. Each, however, may also be self-fulfilling.
representative institutions and purely rational material Democratic liberals do not need to assume either
interests do not control for the possibility that powerful that public opinion rules foreign policy or that the
states can have rational incentives to conquer wealthy entire governmental elite is liberal. They can assume
and exploit wealthy, weak democracies. If reputations that the elite typically manages public affairs but that
are short and differentiable and supposedly pacifying potentially nonliberal members of the elite have reason
long-run interests are indeterminate, as they often are, to doubt that antiliberal policies would be electorally
something more than rational material interest will be sustained and endorsed by the majority of the demo-
needed to explain liberal peace. cratic public.
Representation should, however, ensure that liberal Third and last, material incentives sustain interlib-
wars are only fought for popular, liberal purposes. This eral normative commitments. The “spirit of commerce”
does not produce peace. The historical liberal legacy spreads widely and creates incentives for states to pro-
is laden with popular wars fought to promote free- mote peace and to try to avert war. Liberal economic
dom, protect private property or support liberal allies theory holds that these cosmopolitan ties derive from
against nonliberal enemies.4 In order to see how the a cooperative international division of labor and free

draw links between trade and peace, but only in the context of wider ing Kant in “Perpetual Peace,” regard these wars as unjust, and
relationships favoring accommodation. Kant warns liberals of their susceptibility to them (see 1970, 106).
3 For further discussion, see Fearon 1994, Gaubatz 1996; Schultz
At the same time, he argues that each nation “can and ought to”
1998, and Lipson 2003. demand that its neighboring nations enter into the pacific union of
4 This is the theme of Small and Singer (1976) and Doyle (1983b), liberal states (102) whose first requirement is domestically liberal
Chan (1984), and Weede (1984). Many liberal philosophers, includ- institutions.

464
American Political Science Review Vol. 99, No. 3

trade according to comparative advantage when the Thus, when Rosato (2003, 588, 593) criticizes the
parties can expect to be governed by a rule of law “norm externalization” argument or the “institutional
that respects property and that enforces legitimate ex- logic” explanations, he is in each case missing two thirds
changes. Each economy is said to be better off than it of the liberal argument. When he argues that economic
would have been under autarky; each thus acquires an interests and local strategic interests shaped liberal im-
incentive to avoid policies that would lead the other perial policy in the 19th century, he is not refuting—–he
to break these economic ties. But, because keeping is confirming—–the logic of the liberal peace. In these
open markets rests on an assumption that the next cases, principled liberal motives joined material in-
set of transactions will also be determined by prices terests in liberal imperialism. Campaigns against the
rather than coercion, a sense of mutual security is vi- slave trade destabilized commercial oligarchies, mak-
tal to avoid security-motivated searches for economic ing them prone to collapse. The mission civilatrice and
autarky. Thus, avoiding a challenge to another liberal the “dual mandate” imperial ideologies both included
state’s security or even enhancing each other’s secu- liberal principles, albeit ones that allowed for liberal
rity by means of alliance naturally follows economic imperial paternalism of the sort J. S. Mill (1859/1973)
interdependence. endorsed for societies he and his fellow liberals saw
In this same regard, a further cosmopolitan source as incapable of governing themselves. But commercial
of liberal peace is that the international market re- and property interests, which lacked institutionaliza-
moves difficult decisions of production and distribution tion in much of Africa and Asia, were even more im-
from the direct sphere of state policy. A foreign state portant. Lacking both legal recognition and the context
thus does not appear directly responsible for these of interliberal respect, commercial and property claims
outcomes; states can stand aside from, and to some fueled imperialism. (Doyle [1983a, 1983b] 1996, 37–
degree above, these contentious market rivalries and 9). Liberals were all too ready to enforce those prop-
be ready to step in to resolve crises. The interdepen- erty claims both as a matter of material interest and
dence of commerce and the international contacts of principled defense of rights. Interliberal peace rests on
state officials help create cross-cutting transnational the combined effect of the three pillars. Absent one
ties that serve as lobbies for mutual accommodation. of them, pacific policy is underdetermined and under-
According to modern liberal scholars, international fi- mined.
nanciers and transnational and transgovernmental or- During the Cold War, the United States did inter-
ganizations create interests in favor of accommodation. vene against or take measures to undermine covertly
Moreover, their variety has ensured no single conflict numerous popular regimes in the Third World. In
sours an entire relationship by setting off a spiral of many cases the U.S. administration in office was con-
reciprocated retaliation. vinced that the regimes in question (Mossadegh in Iran,
Conversely, the suspicion that characterizes relations Arbenz in Guatemala, Jagan in Guyana, Allende in
between liberal and nonliberal governments can lead to Chile, and the Sandinistas in Nicaragua) were threats
restrictions on the range of contacts between societies. both to property and to the rule of law. The fact that
And this can increase the prospect that a single conflict these regimes were more progressive and popular than
will determine an entire relationship. As importantly, any previous regime in those countries (and, in some
in relations with weak societies, “protecting “native cases, since) did not make them well-established liberal
rights” from native oppressors, and protecting univer- democracies. Many U.S. officials doubted their stability
sal rights of property and settlement from local trans- as democracies. They were also seen as influenced by
gressions, introduced especially liberal motives for im- and allied with communist regimes. President Kennedy
perial rule” (Doyle [1983a, 1983b] 1996, p. 37). When articulated the logic clearly, referring to the assassina-
property lacks clear title and exchanges are subject tion of Trujillo in the Democratic Republic: “There
to manipulation and uncertain legal enforcement—–the are three possibilities in descending order of prefer-
typical environment of non-liberal states—–then eco- ence, a decent democratic regime, a continuation of
nomic contact generates strife. the Trujillo regime or a Castro regime. We ought to
No single constitutional, international, or cosmopoli- aim at the first, but we cannot really renounce the
tan source alone is sufficient. This variant of liberal second until we are sure that we can avoid the third”
theory is neither solely institutional, nor solely ideo- (Schlesinger 1965, 769, quoted in Doyle [1983] 1996,
logical, nor solely economic. But together (and only to- 41). As importantly, all of these interventions were
gether) the three specific strands of liberal institutions, covert; they lacked the mechanisms of publicity on
liberal ideas, and transnational ties plausibly connect which the liberal peace rests. The explanation underly-
the characteristics of liberal polities and economies ing the liberal peace makes no assumption that every
with sustained liberal peace. But in their relations with official, always and everywhere, is motivated by liberal
nonliberal states, liberal states have not escaped from principle and interest—–just that over the normal po-
the insecurity caused by anarchy in the world politi- litical cycle nonliberal principles and interests will not
cal system considered as a whole. Moreover, the very become the norm in the formation of liberal foreign
constitutional restraint, international respect for in- policy.
dividual rights, and shared commercial interests that A much more logical explanation comes with
establish grounds for peace among liberal states estab- methodological costs. Data sets on the liberal peace do
lish grounds for additional conflict in relations between not adequately code for these three pillars together and
liberal and nonliberal societies. separately. My own coding (1983a, 1983b, 1986, 1997)

465
Three Pillars of the Liberal Peace August 2005

was approximate. The most thorough recent empirical Gartzke, Eric, Quan Li, and Charles Boehmer. 2001. “Investing in
test of Kantian propositions (Russett and O’Neal 2001) the Peace.” International Organization 55 (Spring), 391–438.
shows the separate positive effects of democratic insti- Gaubatz, Kurt. 1996. “Democratic States and Commitment in Inter-
national Politics.” International Organization 50:1 (Winter): 109–
tutions and trade (and membership in international 39.
organizations), but it doesn’t separately code for lib- Kant, Immanuel. 1970. “Perpetual Peace. In Kant’s Political Writings,
eral norms. The substantial statistical confirmation that ed. Hans Reiss; trans. H. B. Nisbet. Cambridge, UK: Cambridge
inter-democratic peace, (coding for democratic insti- University Press.
Lai, Brian and Dan Reiter. 2000. “Democracy, Political Similarity
tutions), does receive is thus probably a reflection of and International Alliances.” Journal of Conflict Resolution 44
the tendency for principles of liberal individualism and (April).
democratic institutions to evolve together.5 But we Lipson, Charles. 2003. Reliable Partners. Princeton: Princeton Uni-
cannot be sure of this. Compared to other testable in- versity Press.
Mearsheimer, John. 1990. “Back to the Future.” International Secu-
ternational theories of similar scope, the empirical con-
rity 15, 1 (Summer): 5–56.
firmation of the liberal peace is exceptionally strong, Mill, John Stuart. 1859/1973. “A Few Words on Nonintervention.”
but that does not mean that the theory does not need In Essays on Politics and Culture, ed. Gertrude Himmelfarb.
additional testing. Gloucester, MA: Peter Smith.
Mueller, John. 1989. Retreat from Doomsday. Basic Books.
Rosato, Sebastian. 2003. “The Flawed Logic of Democratic Peace
REFERENCES Theory.” American Political ScienceReview 97: 4 (November):
585–602.
Chan, Steve. 1984. “Mirror, Mirror on the Wall . . . Are the Freer Rousseau, Jean-Jacques. 1756/1917. “State of war” in C. E. Vaughan,
Countries More Pacific.” Journal of Conflict Resolution 28:4 ed. A Lasting Peace through the Federation of Europe. London:
(December): 617–48. Constable, 1917.
Cobden, Richard. 1901. Political Writings of Richard Cobden. Rummel, R. J. 1997. Power Kills. New Brusnwick, NJ: Transactions
London: Ridgway. Publishers.
Doyle, Michael. 1983a. “Kant, Liberal Legacies, and Foreign Affairs: Russett, Bruce, and John O’Neal. 2001. Triangulating Peace. New
Part I.” Philosophy and Public Affairs 12, 3 (June): 205–35. York: Norton.
Doyle, Michael. 1983b. “Kant, Liberal Legacies, and Foreign Affairs: Schlesinger, Arthur. 1965. A Thousand Days. Boston: Houghton
Part II.” Philosophy and Public Affairs 12, 4 (October): 323–53. Mifflin.
Doyle, Michael. 1986. “Liberalism and World Politics” American Schultz, Kenneth. 1998. “Domestic Opposition and Signaling in
Political Science Review 80: 4 (December): 1151–69. International Crises.” American Political Science Review 2: 4
Doyle, Michael. 1996. “Kant, Liberal Legacies, and Foreign Af- (December): 829–844.
fairs.” In Debating the Democratic Peace, ed. Michael Brown, Sean Small, Melvin, and J. David Singer. 1976. “The War Proneness of
Lynn-Jones, and Steven E. Miller. Cambridge: MIT Press, Democratic Regimes.” Jerusalem Journal of International Rela-
3–57. tions 1, 4 (Summer): 50–69.
Doyle, Michael. 1997. Ways of War and Peace. New York: Norton. Van Evera, Stephen. 1990. “Primed for Peace.” International Security
Fearon, James. 1994. “ Domestic Political Audiences and the Escala- 9, 1, 58–107.
tion of Political Disputes.” American Political Science Review 88, Weede, Eric. 1984. “Democracy and War Involvement,” Journal of
3 (September): 577–92. Conflict Resolution 28:4 (December): 649–64.

5 For thorough surveys of empirical literature see Rummel (1997)

466