Domestic politics and international relations are often somehow entangled,
but our theories have not yet sorted out the puzzling tangle. It is fruitless
to debate whether domestic politics really determine international relations,
or the reverse. The answer to that question is clearly "Both, sometimes."
The more interesting questions are "When?" and "How?" This article offers
a theoretical approach to this issue, but I begin with a story that illustrates
the puzzle.
One illuminating example of how diplomacy and domestic politics can
become entangled culminated at the Bonn summit conference of 1978.' In
the mid-1970s, a coordinated program of global reflation, led by the "locomotive"
economies of the United States, Germany, and Japan, had been
proposed to foster Western recovery from the first oil shock.2 This proposal
An earlier version of this article was delivered at the 1986 annual meeting of the American
Political Science Association. For criticisms and suggestions, I am indebted to Robert Axelrod,
Nicholas Bayne, Henry Brady, James A. Caporaso, Barbara Crane, Ernst B. Haas, Stephan
Haggard, C. Randall Henning, Peter B. Kenen, Robert 0. Keohane, Stephen D. Krasner, Jacek
Kugler, Lisa Martin, John Odell, Robert Powell, Kenneth A. Shepsle, Steven Stedman, Peter
Yu, members of research seminars at the Universities of Iowa, Michigan, and Harvard, and
two anonymous reviewers. I am grateful to the Rockefeller Foundation for enabling me to
complete this research.
1. The following account is drawn from Robert D. Putnam and C. Randall Henning, "The
Bonn Summit of 1978: How Does International Economic Policy Coordination Actually Work?"
Brookings Discussion Papers in International Economics, no. 53 (Washington, D.C.: Brookings
Institution, October 1986), and Robert D. Putnam and Nicholas Bayne, Hanging Together:
Cooperation and Conflict in the Seven-Power Summits, rev. ed. (Cambridge, Mass.: Harvard
University Press, 1987), pp. 62-94.
2. Among interdependent economies, most economists believe, policies can often be more
effective if they are internationally coordinated. For relevant citations, see Putnam and Bayne,
Hanging Together, p. 24.
International Organization 42, 3, Summer 1988
© 1988 by the World Peace Foundation and the Massachusetts Institute of Technology
428 International Organization
had received a powerful boost from the incoming Carter administration and
was warmly supported by the weaker countries, as well as the Organization
for Economic Co-operation and Development (OECD) and many private
economists, who argued that it would overcome international payments imbalances
and speed growth all around. On the other hand, the Germans and
the Japanese protested that prudent and successful economic managers should
not be asked to bail out spendthrifts. Meanwhile, Jimmy Carter's ambitious
National Energy Program remained deadlocked in Congress, while Helmut
Schmidt led a chorus of complaints about the Americans' uncontrolled appetite
for imported oil and their apparent unconcern about the falling dollar.
All sides conceded that the world economy was in serious trouble, but it
was not clear which was more to blame, tight-fisted German and Japanese
fiscal policies or slack-jawed U.S. energy and monetary policies.
At the Bonn summit, however, a comprehensive package deal was approved,
the clearest case yet of a summit that left all participants happier
than when they arrived. Helmut Schmidt agreed to additional fiscal stimulus,
amounting to 1 percent of GNP, Jimmy Carter committed himself to decontrol
domestic oil prices by the end of 1980, and Takeo Fukuda pledged new
efforts to reach a 7 percent growth rate. Secondary elements in the Bonn
accord included French and British acquiescence in the Tokyo Round trade
negotiations; Japanese undertakings to foster import growth and restrain
exports; and a generic American promise to fight inflation. All in all, the
Bonn summit produced a balanced agreement of unparalleled breadth and
specificity. More remarkably, virtually all parts of the package were actually
implemented.
Most observers at the time welcomed the policies agreed to at Bonn,
although in retrospect there has been much debate about the economic
wisdom of this package deal. However, my concern here is not whether the
deal was wise economically, but how it became possible politically. My
research suggests, first, that the key governments at Bonn adopted policies
different from those that they would have pursued in the absence of international
negotiations, but second, that agreement was possible only because
a powerful minority within each government actually favored on domestic
grounds the policy being demanded internationally.
Within Germany, a political process catalyzed by foreign pressures was
surreptitiously orchestrated by expansionists inside the Schmidt government.
Contrary to the public mythology, the Bonn deal was not forced on
a reluctant or "altruistic" Germany. In fact, officials in the Chancellor's
Office and the Economics Ministry, as well as in the Social Democratic party
and the trade unions, had argued privately in early 1978 that further stimulus
was domestically desirable, particularly in view of the approaching 1980
elections. However, they had little hope of overcoming the opposition of
the Finance Ministry, the Free Democratic party (part of the government
coalition), and the business and banking community, especially the leaderDiplomacy
and domestic politics 429
ship of the Bundesbank. Publicly, Helmut Schmidt posed as reluctant to the
end. Only his closest advisors suspected the truth: that the chancellor "let
himself be pushed" into a policy that he privately favored, but would have
found costly and perhaps impossible to enact without the summit's package
deal.
Analogously, in Japan a coalition of business interests, the Ministry of
Trade and Industry (MITI), the Economic Planning Agency, and some expansion-
minded politicians within the Liberal Democratic Party pushed for
additional domestic stimulus, using U.S. pressure as one of their prime
arguments against the stubborn resistance of the Ministry of Finance (MOF).
Without internal divisions in Tokyo, it is unlikely that the foreign demands
would have been met, but without the external pressure, it is even more
unlikely that the expansionists could have overridden the powerful MOF.
"Seventy percent foreign pressure, 30 percent internal politics," was the
disgruntled judgment of one MOF insider. "Fifty-fifty," guessed an official
from MITI.3
In the American case, too, internal politicking reinforced, and was reinforced
by, the international pressure. During the summit preparations American
negotiators occasionally invited their foreign counterparts to put more
pressure on the Americans to reduce oil imports. Key economic officials
within the administration favored a tougher energy policy, but they were
opposed by the president's closest political aides, even after the summit.
Moreover, congressional opponents continued to stymie oil price decontrol,
as they had under both Nixon and Ford. Finally, in April 1979, the president
decided on gradual administrative decontrol, bringing U.S. prices up to world
levels by October 1981. His domestic advisors thus won a postponement of
this politically costly move until after the 1980 presidential election, but in
the end, virtually every one of the pledges made at Bonn was fulfilled. Both
proponents and opponents of decontrol agree that the summit commitment
was at the center of the administration's heated intramural debate during
the winter of 1978-79 and instrumental in the final decision.4
In short, the Bonn accord represented genuine international policy coordination.
Significant policy changes were pledged and implemented by the
key participants. Moreover—although this counterfactual claim is necessarily
harder to establish—those policy changes would very probably not
have been pursued (certainly not the same scale and within the same time
frame) in the absence of the international agreement. Within each country,
one faction supported the policy shift being demanded of its country inter-
3. For a comprehensive account of the Japanese story, see I. M. Destler and Hisao Mitsuyu,
"Locomotives on Different Tracks: Macroeconomic Diplomacy, 1977-1979," in I. M. Destler
and Hideo Sato, eds., Coping with U.S.-Japanese Economic Conflicts (Lexington, Mass.:
Heath, 1982).
4. For an excellent account of U.S. energy policy during this period, see G. John Ikenberry,
"Market Solutions for State Problems: The International and Domestic Politics of American
Oil Decontrol," International Organization 42 (Winter 1988).
430 International Organization
nationally, but that faction was initially outnumbered. Thus, international
pressure was a necessary condition for these policy shifts. On the other
hand, without domestic resonance, international forces would not have sufficed
to produce the accord, no matter how balanced and intellectually persuasive
the overall package. In the end, each leader believed that what he
was doing was in his nation's interest—and probably in his own political
interest, too, even though not all his aides agreed.5 Yet without the summit
accord he probably would not (or could not) have changed policies so easily.
In that sense, the Bonn deal successfully meshed domestic and international
pressures.
Neither a purely domestic nor a purely international analysis could account
for this episode. Interpretations cast in terms either of domestic causes and
international effects ("Second Image"6) or of international causes and domestic
effects ("Second Image Reversed"7) would represent merely "partial
equilibrium" analyses and would miss an important part of the story, namely,
how the domestic politics of several countries became entangled via an
international negotiation. The events of 1978 illustrate that we must aim
instead for "general equilibrium" theories that account simultaneously for
the interaction of domestic and international factors. This article suggests a
conceptual framework for understanding how diplomacy and domestic
politics interact.
Domestic-international entanglements: the state of the art
Much of the existing literature on relations between domestic and international
affairs consists either of ad hoc lists of countless "domestic influences"
on foreign policy or of generic observations that national and international
affairs are somehow "linked."8 James Rosenau was one of the first scholars
to call attention to this area, but his elaborate taxonomy of "linkage politics"
generated little cumulative research, except for a flurry of work correlating
domestic and international "conflict behavior."9
A second stream of relevant theorizing began with the work by Karl
5. It is not clear whether Jimmy Carter fully understood the domestic implications of his
Bonn pledge at the time. See Putnam and Henning, "The Bonn Summit," and Ikenberry,
"Market Solutions for State Problems."
6. Kenneth N. Waltz, Man, the State, and War: A Theoretical Analysis (New York: Columbia
University Press, 1959).
7. Peter Gourevitch, "The Second Image Reversed: The International Sources of Domestic
Politics," International Organization 32 (Autumn 1978), pp. 881-911.
8. I am indebted to Stephan Haggard for enlightening discussions about domestic influences
on international relations.
9. James Rosenau, "Toward the Study of National-International Linkages," in his Linkage
Politics: Essays on the Convergence of National and International Systems (New York: Free
Press, 1969), as well as his "Theorizing Across Systems: Linkage Politics Revisited," in Jonathan
Wilkenfeld, ed., Conflict Behavior and Linkage Politics (New York: David McKay, 1973),
especially p. 49.
Diplomacy and domestic politics 431
Deutsch and Ernst Haas on regional integration.10 Haas, in particular, emphasized
the impact of parties and interest groups on the process of European
integration, and his notion of "spillover" recognized the feedback between
domestic and international developments. However, the central dependent
variable in this work was the hypothesized evolution of new supranational
institutions, rather than specific policy developments, and when European
integration stalled, so did this literature. The intellectual heirs of this tradition,
such as Joseph Nye and Robert Keohane, emphasized interdependence
and transnationalism, but the role of domestic factors slipped more and
more out of focus, particularly as the concept of international regimes came
to dominate the subfield."
The "bureaucratic politics" school of foreign policy analysis initiated
another promising attack on the problem of domestic-international interaction.
As Graham Allison noted, "Applied to relations between nations,
the bureaucratic politics model directs attention to intra-national games, the
overlap of which constitutes international relations."12 Nevertheless, the
nature of this "overlap" remained unclarified, and the theoretical contribution
of this literature did not evolve much beyond the principle that bureaucratic
interests matter in foreign policymaking.
More recently, the most sophisticated work on the domestic determinants
of foreign policy has focused on "structural" factors, particularly "state
strength." The landmark works of Peter Katzenstein and Stephen Krasner,
for example, showed the importance of domestic factors in foreign economic
policy. Katzenstein captured the essence of the problem: "The main purpose
of all strategies of foreign economic policy is to make domestic policies
compatible with the international political economy."13 Both authors stressed
the crucial point that central decision-makers ("the state") must be concerned
simultaneously with domestic and international pressures.
10. Karl W. Deutsch et al., Political Community in the North Atlantic Area: International
Organization in the Light of Historical Experience (Princeton: Princeton University Press,
1957) and Ernst B. Haas, The Uniting of Europe: Political, Social, and Economic Forces,
1950-1957 (Stanford, Calif.: Stanford University Press, 1958).
11. Robert 0. Keohane and Joseph S. Nye, Power and Interdependence (Boston: Little,
Brown, 1977). On the regime literature, including its neglect of domestic factors, see Stephan
Haggard and Beth Simmons, "Theories of International Regimes," International Organization
41 (Summer 1987), pp. 491-517.
12. Graham T. Allison, Essence of Decision: Explaining the Cuban Missile Crisis (Boston:
Little, Brown, 1971), p. 149.
13. Peter J. Katzenstein, ed., Between Power and Plenty: Foreign Economic Policies of
Advanced Industrial States (Madison: University of Wisconsin Press, 1978), p. 4. See also
Katzenstein, "International Relations and Domestic Structures: Foreign Economic Policies of
Advanced Industrial States," International Organization 30 (Winter 1976), pp. 1-45; Stephen
D. Krasner, "United States Commercial and Monetary Policy: Unravelling the Paradox of
External Strength and Internal Weakness," in Katzenstein, Between Power and Plenty, pp.
51-87; and Krasner, Defending the National Interest: Raw Materials Investments and U.S.
Foreign Policy (Princeton: Princeton University Press, 1978).
432 International Organization
More debatable, however, is their identification of "state strength" as the
key variable of interest. Given the difficulties of measuring "state strength,"
this approach courts tautology,14 and efforts to locate individual countries
on this ambiguous continuum have proved problematic.15 "State strength,"
if reinterpreted as merely the opposite of governmental fragmentation, is no
doubt of some interest in the comparative study of foreign policy. However,
Gourevitch is quite correct to complain that "the strong state-weak state
argument suggests t h a t . . . the identity of the governing coalition does not
matter. This is a very apolitical argument."16 Moreover, because "state
structures" (as conceived in this literature) vary little from issue to issue or
from year to year, such explanations are ill-suited for explaining differences
across issues or across time (unless "time" is measured in decades or centuries).
A more adequate account of the domestic determinants of foreign
policy and international relations must stress politics: parties, social classes,
interest groups (both economic and noneconomic), legislators, and even
public opinion and elections, not simply executive officials and institutional
arrangements.17
Some work in the "state-centric" genre represents a unitary-actor model
run amok. "The central proposition of this paper," notes one recent study,
"is that the state derives its interests from and advocates policies consistent
with the international system at all times and under all circumstances."18 In
fact, on nearly all important issues "central decision-makers" disagree about
what the national interest and the international context demand. Even if we
arbitrarily exclude the legislature from "the state" (as much of this literature
does), it is wrong to assume that the executive is unified in its views. Certainly
this was true in none of the states involved in the 1978 negotiations. What
was "the" position of the German or Japanese state on macroeconomic
policy in 1978, or of the American state on energy policy? If the term "state"
is to be used to mean "central decision-makers," we should treat it as a
plural noun: not "the state, it . . ." but "the state, they . . ." Central executives
have a special role in mediating domestic and international pressures
precisely because they are directly exposed to both spheres, not because
14. For example, see Krasner, "United States Commercial and Money Policy," p. 55: "The
central analytic characteristic that determines the abiliiy of a state to overcome domestic
resistance is its strength in relation to its own society."
15. Helen Milner, "Resisting the Protectionist Temptation: Industry and the Making of Trade
Policy in France and the United States during the 1970s," International Organization 41
(Autumn 1987), pp. 639-65.
16. Gourevitch, "The Second Image Reversed," p. 903.
17. In their more descriptive work, "state-centric" scholars are often sensitive to the impact
of social and political conflicts, such as those between industry and finance, labor and business,
and export-oriented versus import-competing sectors. See Katzenstein, Between Power and
Plenty, pp. 333-36, for example.
18. David A. Lake, "The State as Conduit: The International Sources of National Political
Action," presented at the 1984 annual meeting of the American Political Science Association,
p. 13.
Diplomacy and domestic politics 433
they are united on all issues nor because they are insulated from domestic
politics.
Thus, the state-centric literature is an uncertain foundation for theorizing
about how domestic and international politics interact. More interesting are
recent works about the impact of the international economy on domestic
politics and domestic economic policy, such as those by Alt, Evans, Gourevitch,
and Katzenstein.19 These case studies, representing diverse methodological
approaches, display a theoretical sophistication on the international-
to-domestic causal connection far greater than is characteristic of
comparable studies on the domestic-to-international half of the loop. Nevertheless,
these works do not purport to account for instances of reciprocal
causation, nor do they examine cases in which the domestic politics of
several countries became entangled internationally.
In short, we need to move beyond the mere observation that domestic
factors influence international affairs and vice versa, and beyond simple
catalogs of instances of such influence, to seek theories that integrate both
spheres, accounting for the areas of entanglement between them.
Two-level games: a metaphor
for domestic-international interactions
Over two decades ago Richard E. Walton and Robert B. McKersie offered
a "behavioral theory" of social negotiations that is strikingly applicable to
international conflict and cooperation.20 They pointed out, as all experienced
negotiators know, that the unitary-actor assumption is often radically misleading.
As Robert Strauss said of the Tokyo Round trade negotiations:
"During my tenure as Special Trade Representative, I spent as much time
negotiating with domestic constituents (both industry and labor) and members
of the U.S. Congress as I did negotiating with our foreign trading
partners."2 1
19. James E. Alt, "Crude Politics: Oil and the Political Economy of Unemployment in Britain
and Norway, 1970-1985," British Journal of Political Science 17 (April 1987), pp. 149-99; Peter
B. Evans, Dependent Development: The Alliance of Multinational, State, and Local Capital
in Brazil (Princeton: Princeton University Press, 1979); Peter Gourevitch, Politics in Hard
Times: Comparative Responses to International Economic Crises (Ithaca, N.Y.: Cornell University
Press, 1986); Peter J. Katzenstein, Small States in World Markets: Industrial Policy in
Europe (Ithaca, N.Y.: Cornell University Press, 1985).
20. Richard E. Walton and Robert B. McKersie, A Behavioral Theory of Labor Negotiations:
An Analysis of a Social Interaction System (New York: McGraw-Hill, 1965).
21. Robert S. Strauss, "Foreword," in Joan E. Twiggs, The Tokyo Round of Multilateral
Trade Negotiations: A Case Study in Building Domestic Support for Diplomacy (Washington,
D.C.: Georgetown University Institute for the Study of Diplomacy, 1987), p. vii. Former
Secretary of Labor John Dunlop is said to have remarked that "bilateral negotiations usually
require three agreements—one across the table and one on each side of the table," as cited in
Howard Raiffa, The Art and Science of Negotiation (Cambridge, Mass.: Harvard University
Press, 1982), p. 166.
434 International Organization
The politics of many international negotiations can usefully be conceived
as a two-level game. At the national level, domestic groups pursue their
interests by pressuring the government to adopt favorable policies, and politicians
seek power by constructing coalitions among those groups. At the
international level, national governments seek to maximize their own ability
to satisfy domestic pressures, while minimizing the adverse consequences
of foreign developments. Neither of the two games can be ignored by central
decision-makers, so long as their countries remain interdependent, yet sovereign.
Each national political leader appears at both game boards. Across the
international table sit his foreign counterparts, and at his elbows sit diplomats
and other international advisors. Around the domestic table behind him sit
party and parliamentary figures, spokespersons for domestic agencies, representatives
of key interest groups, and the leader's own political advisors.
The unusual complexity of this two-level game is that moves that are rational
for a player at one board (such as raising energy prices, conceding territory,
or limiting auto imports) may be impolitic for that same player at the other
board. Nevertheless, there are powerful incentives for consistency between
the two games. Players (and kibitzers) will tolerate some differences in
rhetoric between the two games, but in the end either energy prices rise or
they don't.
The political complexities for the players in this two-level game are staggering.
Any key player at the international table who is dissatisfied with the
outcome may upset the game board, and conversely, any leader who fails
to satisfy his fellow players at the domestic table risks being evicted from
his seat. On occasion, however, clever players will spot a move on one board
that will trigger realignments on other boards, enabling them to achieve
otherwise unattainable objectives. This "two-table" metaphor captures the
dynamics of the 1978 negotiations better than any model based on unitary
national actors.
Other scholars have noted the multiple-game nature of international relations.
Like Walton and McKersie, Daniel Druckman has observed that a
negotiator "attempts to build a package that will be acceptable both to the
other side and to his bureaucracy." However, Druckman models the domestic
and international processes separately and concludes that "the interaction
between the processes . . . remains a topic for investigation."22
Robert Axelrod has proposed a "Gamma paradigm," in which the U.S.
president pursues policies vis-a-vis the Soviet Union with an eye towards
maximizing his popularity at home. However, this model disregards domestic
22. Daniel Druckman, "Boundary Role Conflict: Negotiation as Dual Responsiveness," in
I. William Zartman, ed., The Negotiation Process: Theories and Applications (Beverly Hills:
Sage, 1978), pp. 100-101, 109. For a review of the social-psychological literature on bargainers
as representatives, see Dean G. Pruitt, Negotiation Behavior (New York: Academic Press,
1981), pp. 41-43.
Diplomacy and domestic politics 435
cleavages, and it postulates that one of the international actors—the Soviet
leadership—cares only about international gains and faces no domestic constraint
while the other—the U.S. president—cares only about domestic gains,
except insofar as his public evaluates the international competition.23 Probably
the most interesting empirically based theorizing about the connection
between domestic and international bargaining is that of Glenn Snyder and
Paul Diesing. Though working in the neo-realist tradition with its conventional
assumption of unitary actors, they found that, in fully half of the crises
they studied, top decision-makers were not unified. They concluded that
prediction of international outcomes is significantly improved by understanding
internal bargaining, especially with respect to minimally acceptable compromises.
24
Metaphors are not theories, but I am comforted by Max Black's observation
that "perhaps every science must start with metaphor and end with
algebra; and perhaps without the metaphor there would never have been
any algebra."25 Formal analysis of any game requires well-defined rules,
choices, payoffs, players, and information, and even then, many simple twoperson,
mixed-motive games have no determinate solution. Deriving analytic
solutions for two-level games will be a difficult challenge. In what follows I
hope to motivate further work on that problem.
Towards a theory of ratification:
the importance of "win-sets"
Consider the following stylized scenario that might apply to any two-level
game. Negotiators representing two organizations meet to reach an agreement
between them, subject to the constraint that any tentative agreement
must be ratified by their respective organizations. The negotiators might be
heads of government representing nations, for example, or labor and management
representatives, or party leaders in a multiparty coalition, or a
finance minister negotiating with an IMF team, or leaders of a House-Senate
conference committee, or ethnic-group leaders in a consociational democracy.
For the moment, we shall presume that each side is represented by a
single leader or "chief negotiator," and that this individual has no indepen-
23. Robert Axelrod, "The Gamma Paradigm for Studying the Domestic Influence on Foreign
Policy," prepared for delivery at the 1987 Annual Meeting of the International Studies Association.
24. Glenn H. Snyder and Paul Diesing, Conflict Among Nations: Bargaining, Decision Making,
and System Structure in International Crises (Princeton: Princeton University Press, 1977),
pp. 510-25.
25. Max Black, Models and Metaphors (Ithaca, N.Y.: Cornell University Press, 1962), p.
242, as cited in Duncan Snidal, "The Game Theory of International Politics," World Politics
38 (October 1985), p. 36n.
436 International Organization
dent policy preferences, but seeks simply to achieve an agreement that will
be attractive to his constituents.26
It is convenient analytically to decompose the process into two stages:
1. bargaining between the negotiators, leading to a tentative agreement;
call that Level I.
2. separate discussions within each group of constituents about whether
to ratify the agreement; call that Level II.
This sequential decomposition into a negotiation phase and a ratification
phase is useful for purposes of exposition, although it is not descriptively
accurate. In practice, expectational effects will be quite important. There
are likely to be prior consultations and bargaining at Level II to hammer
out an initial position for the Level I negotiations. Conversely, the need for
Level II ratification is certain to affect the Level I bargaining. In fact, expectations
of rejection at Level II may abort negotiations at Level I without
any formal action at Level II. For example, even though both the American
and Iranian governments seem to have favored an arms-for-hostages deal,
negotiations collapsed as soon as they became public and thus liable to de
facto "ratification." In many negotiations, the two-level process may be
iterative, as the negotiators try out possible agreements and probe their
constituents' views. In more complicated cases, as we shall see later, the
constituents' views may themselves evolve in the course of the negotiations.
Nevertheless, the requirement that any Level I agreement must, in the end,
be ratified at Level II imposes a crucial theoretical link between the two
levels.
"Ratification" may entail a formal voting procedure at Level II, such as
the constitutionally required two-thirds vote of the U.S. Senate for ratifying
treaties, but I use the term generically to refer to any decision-process at
Level II that is required to endorse or implement a Level I agreement,
whether formally or informally. It is sometimes convenient to think of ratification
as a parliamentary function, but that is not essential. The actors at
Level II may represent bureaucratic agencies, interest groups, social classes,
or even "public opinion." For example, if labor unions in a debtor country
withhold necessary cooperation from an austerity program that the government
has negotiated with the IMF, Level II ratification of the agreement
may be said to have failed; ex ante expectations about that prospect will
surely influence the Level I negotiations between the government and the
IMF.
Domestic ratification of international agreements might seem peculiar to
democracies. As the German Finance Minister recently observed, "The limit
of expanded cooperation lies in the fact that we are democracies, and we
26. To avoid unnecessary complexity, my argument throughout is phrased in terms of a single
chief negotiator, although in many cases some of his responsibilities may be delegated to aides.
Later in this article I relax the assumption that the negotiator has no independent preferences.
Diplomacy and domestic politics 437
need to secure electoral majorities at home."27 However, ratification need
not be "democratic" in any normal sense. For example, in 1930 the Meiji
Constitution was interpreted as giving a special role to the Japanese military
in the ratification of the London Naval Treaty;28 and during the ratification
of any agreement between Catholics and Protestants in Northern Ireland,
presumably the IRA would throw its power onto the scales. We need only
stipulate that, for purposes of counting "votes" in the ratification process,
different forms of political power can be reduced to some common denominator.
The only formal constraint on the ratification process is that since the
identical agreement must be ratified by both sides, a preliminary Level I
agreement cannot be amended at Level II without reopening the Level I
negotiations. In other words, final ratification must be simply "voted" up
or down; any modification to the Level I agreement counts as a rejection,
unless that modification is approved by all other parties to the agreement.29
Congresswoman Lynn Martin captured the logic of ratification when explaining
her support for the 1986 tax reform bill as it emerged from the
conference committee: "As worried as I am about what this bill does, I am
even more worried about the current code. The choice today is not between
this bill and a perfect bill; the choice is between this bill and the death of
tax reform."30
Given this set of arrangements, we may define the "win-set" for a given
Level II constituency as the set of all possible Level I agreements that would
"win"—that is, gain the necessary majority among the constituents—when
simply voted up or down.31 For two quite different reasons, the contours of
the Level II win-sets are very important for understanding Level I agreements.
First, larger win-sets make Level I agreement more likely, ceteris paribus.32
By definition, any successful agreement must fall within the Level II win-
27. Gerhardt Stoltenberg, Wall Street Journal Europe, 2 October 1986, as cited in C. Randall
Henning, Macroeconomic Diplomacy in the 1980s: Domestic Politics and International Conflict
Among the United States, Japan, and Europe, Atlantic Paper No. 65 (New York: Croom Helm,
for the Atlantic Institute for International Affairs, 1987), p. 1.
28. Ito Takashi, "Conflicts and Coalition in Japan, 1930: Political Groups and the London
Naval Disarmament Conference," in Sven Groennings et al., eds, The Study of Coalition
Behavior (New York: Holt, Rinehart, & Winston, 1970); Kobayashi Tatsuo, "The London
Naval Treaty, 1930," in James W. Morley, ed., Japan Erupts: The London Naval Conference
and the Manchurian Incident, 1928-1932 (New York: Columbia University Press, 1984), pp.
11-117. I am indebted to William Jarosz for this example.
29. This stipulation is, in fact, characteristic of most real-world ratification procedures, such
as House and Senate action on conference committee reports, although it is somewhat violated
by the occasional practice of appending "reservations" to the ratification of treaties.
30. New York Times, 26 September 1986.
31. For the conception of win-set, see Kenneth A. Shepsle and Barry R. Weingast, "The
Institutional Foundations of Committee Power," American Political Science Review 81 (March
1987), pp. 85-104. I am indebted to Professor Shepsle for much help on this topic.
32. To avoid tedium, I do not repeat the "other things being equal" proviso in each of the
propositions that follow. Under some circumstances an expanded win-set might actually make
practicable some outcome that could trigger a dilemma of collective action. See Vincent P.
Crawford, "A Theory of Disagreement in Bargaining," Econometrica 50 (May 1982), pp. 607-
37.
438 International Organization
sets of each of the parties to the accord. Thus, agreement is possible only
if those win-sets overlap, and the larger each win-set, the more likely they
are to overlap. Conversely, the smaller the win-sets, the greater the risk that
the negotiations will break down. For example, during the prolonged prewar
Anglo-Argentine negotiations over the Falklands/Malvinas, several tentative
agreements were rejected in one capital or the other for domestic
political reasons; when it became clear that the initial British and Argentine
win-sets did not overlap at all, war became virtually inevitable.33
A brief, but important digression: The possibility of failed ratification
suggests that game theoretical analyses should distinguish between voluntary
and involuntary defection. Voluntary defection refers to reneging by a rational
egoist in the absence of enforceable contracts—the much-analyzed
problem posed, for example, in the prisoner's dilemma and other dilemmas
of collective action. Involuntary defection instead reflects the behavior of
an agent who is unable to deliver on a promise because of failed ratification.
Even though these two types of behavior may be difficult to disentangle in
some instances, the underlying logic is quite different.
The prospects for international cooperation in an anarchic, "self-help"
world are often said to be poor because "unfortunately, policy makers generally
have an incentive to cheat."34 However, as Axelrod, Keohane, and
others have pointed out, the temptation to defect can be dramatically reduced
among players who expect to meet again.35 If policymakers in an anarchic
world were in fact constantly tempted to cheat, certain features of the 1978
story would be very anomalous. For example, even though the Bonn agreement
was negotiated with exquisite care, it contained no provisions for
temporal balance, sequencing, or partial conditionality that might have protected
the parties from unexpected defection. Moreover, the Germans and
the Japanese irretrievably enacted their parts of the bargain more than six
months before the president's action on oil price decontrol and nearly two
years before that decision was implemented. Once they had done so, the
temptation to the president to renege should have been overpowering, but
in fact virtually no one on either side of the decontrol debate within the
administration dismissed the Bonn pledge as irrelevant. In short, the Bonn
"promise" had political weight, because reneging would have had high political
and diplomatic costs.
33. The Sunday Times Insight Team, The Falklands War (London: Sphere, 1982); Max
Hastings and Simon Jenkins, The Battle for the Falklands (New York: Norton, 1984); Alejandro
Dabat and Luis Lorenzano, Argentina: The Malvinas and the End of Military Rule (London:
Verso, 1984). I am indebted to Louise Richardson for these citations.
34. Matthew E. Canzoneri and Jo Anna Gray, "Two Essays on Monetary Policy in an
Interdependent World," International Finance Discussion Paper 219 (Board of Governors of
the Federal Reserve System, February 1983).
35. Robert Axelrod, The Evolution of Cooperation (New York: Basic Books, 1984); Robert
0. Keohane, After Hegemony: Cooperation and Discord in the World Political Economy
(Princeton: Princeton University Press, 1984), esp. p. 116; and the special issue of World
Politics, "Cooperation Under Anarchy," Kenneth A. Oye, ed., vol. 38 (October 1985).
Diplomacy and domestic politics 439
On the other hand, in any two-level game, the credibility of an official
commitment may be low, even if the reputational costs of reneging are high,
for the negotiator may be unable to guarantee ratification. The failure of
Congress to ratify abolition of the "American Selling Price" as previously
agreed during the Kennedy Round trade negotiations is one classic instance;
another is the inability of Japanese Prime Minister Sato to deliver on a
promise made to President Nixon during the "Textile Wrangle."36 A key
obstacle to Western economic coordination in 1985-87 was the Germans'
fear that the Reagan administration would be politically unable to carry out
any commitment it might make to cut the U.S. budget deficit, no matter
how well-intentioned the president.
Unlike concerns about voluntary defection, concern about "deliver-ability"
was a prominent element in the Bonn negotiations. In the post-summit
press conference, President Carter stressed that "each of us has been careful
not to promise more than he can deliver." A major issue throughout the
negotiations was Carter's own ability to deliver on his energy commitments.
The Americans worked hard to convince the others, first, that the president
was under severe domestic political constraints on energy issues, which
limited what he could promise, but second, that he could deliver what he
was prepared to promise. The negotiators in 1978 seemed to follow this
presumption about one another: "He will do what he has promised, so long
as what he has promised is clear and within his power."
Involuntary defection, and the fear of it, can be just as fatal to prospects
for cooperation as voluntary defection. Moreover, in some cases, it may be
difficult, both for the other side and for outside analysts, to distinguish
voluntary and involuntary defection, particularly since a strategic negotiator
might seek to misrepresent a voluntary defection as involuntary. Such behavior
is itself presumably subject to some reputational constraints, although
it is an important empirical question how far reputations generalize from
collectivities to negotiators and vice versa. Credibility (and thus the ability
to strike deals) at Level I is enhanced by a negotiator's (demonstrated) ability
to "deliver" at Level II; this was a major strength of Robert Strauss in the
Tokyo Round negotiations.37
Involuntary defection can only be understood within the framework of a
two-level game. Thus, to return to the issue of win-sets, the smaller the winsets,
the greater the risk of involuntary defection, and hence the more applicable
the literature about dilemmas of collective action.38
36.1. M. Destler, Haruhiro Fukui, and Hideo Sato, The Textile Wrangle: Conflict in Japanese-
American Relations, 1969-1971 (Ithaca, N.Y.: Cornell University Press, 1979), pp. 121-57.
37. Gilbert R. Winham, "Robert Strauss, the MTN, and the Control of Faction," Journal
of World Trade Law 14 (September-October 1980), pp. 377-97, and his International Trade
and the Tokyo Round (Princeton: Princeton University Press, 1986).
38. This discussion implicitly assumes uncertainty about the contours of the win-sets on the
part of the Level I negotiators, for if the win-sets were known with certainty, the negotiators
would never propose for ratification an agreement that would be rejected.
440 International Organization
The second reason why win-set size is important is that the relative size
of the respective Level II win-sets will affect the distribution of the joint gains
from the international bargain. The larger the perceived win-set of a negotiator,
the more he can be "pushed around" by the other Level I negotiators.
Conversely, a small domestic win-set can be a bargaining advantage: "I'd
like to accept your proposal, but I could never get it accepted at home."
Lamenting the domestic constraints under which one must operate is (in the
words of one experienced British diplomat) "the natural thing to say at the
beginning of a tough negotiation."39
This general principle was, of course, first noted by Thomas Schelling
nearly thirty years ago:
The power of a negotiator often rests on a manifest inability to make
concessions and meet demands. . . . When the United States Government
negotiates with other goverments . . . if the executive branch negotiates
under legislative authority, with its position constrained by
law, . . . then the executive branch has a firm position that is visible to
its negotiating partners. . . . [Of course, strategies such as this] run the
risk of establishing an immovable position that goes beyond the ability
of the other to concede, and thereby provoke the likelihood of stalemate
or breakdown.40
Writing from a strategist's point of view, Schelling stressed ways in which
win-sets may be manipulated, but even when the win-set itself is beyond
the negotiator's control, he may exploit its leverage. A Third World leader
whose domestic position is relatively weak (Argentina's Alfonsin?) should
be able to drive a better bargain with his international creditors, other things
being equal, than one whose domestic standing is more solid (Mexico's de
la Madrid?).41 The difficulties of winning congressional ratification are often
exploited by American negotiators. During the negotiation of the Panama
Canal Treaty, for example, "the Secretary of State warned the Panamanians
several times . . . that the new treaty would have to be acceptable to at least
sixty-seven senators," and "Carter, in a personal letter to Torrijos, warned
that further concessions by the United States would seriously threaten chances
for Senate ratification."42 Precisely to forestall such tactics, opponents may
demand that a negotiator ensure himself "negotiating room" at Level II
before opening the Level I negotiations.
The "sweet-and-sour" implications of win-set size are summarized in
Figure 1, representing a simple zero-sum game between X and Y. XM and
39. Geoffrey W. Harrison, in John C. Campbell, ed., Successful Negotiation: Trieste 1954
(Princeton: Princeton University Press, 1976), p. 62.
40. Thomas C. Schelling, The Strategy of Conflict (Cambridge, Mass.: Harvard University
Press, 1960), pp. 19-28.
41. I am grateful to Lara Putnam for this example. For supporting evidence, see Robert R.
Kaufman, "Democratic and Authoritarian Responses to the Debt Issue: Argentina, Brazil,
Mexico," International Organization 39 (Summer 1985), pp. 473-503.
42. W. Mark Habeeb and I. William Zartman, The Panama Canal Negotiations (Washington,
D.C.: Johns Hopkins Foreign Policy Institute, 1986), pp. 40, 42.
Diplomacy and domestic politics 441
XM|
Y, Y2 X, Y3
FIGURE l. Effects of reducing win-set size
YM represent the maximum outcomes for X and Y, respectively, while Xi
and Y, represent the minimal outcomes that could be ratified. At this stage
any. agreement in the range between X, and Yi could be ratified by both
parties. If the win-set of Y were contracted to, say, Y2 (perhaps by requiring
a larger majority for ratification), outcomes between Y, and Y2 would no
longer be feasible, and the range of feasible agreements would thus be truncated
in Y's favor. However, if Y, emboldened by this success, were to
reduce its win-set still further to Y3 (perhaps by requiring unanimity for
ratification), the negotiators would suddenly find themselves deadlocked,
for the win-sets no longer overlap at all.43
Determinants of the win-set
It is important to understand what circumstances affect win-set size. Three
sets of factors are especially important:
43. Several investigators in other fields have recently proposed models of linked games akin
to this "two-level" game. Kenneth A. Shepsle and his colleagues have used the notion of
"interconnected games" to analyze, for example, the strategy of a legislator simultaneously
embedded in two games, one in the legislative arena and the other in the electoral arena. In
this model, a given action is simultaneously a move in two different games, and one player
maximizes the sum of his payoffs from the two games. See Arthur Denzau, William Riker, and
Kenneth Shepsle, "Farquharson and Fenno: Sophisticated Voting and Home Style," American
Political Science Review 79 (December 1985), pp. 1117-34; and Kenneth Shepsle, "Cooperation
and Institutional Arrangements," unpublished manuscript, February 1986. This approach is
similar to models recently developed by economists working in the "rational expectations"
genre. In these models, a government contends simultaneously against other governments and
against domestic trade unions over monetary policy. See, for example, Kenneth Rogoff, "Can
International Monetary Policy Cooperation be Counterproductive," Journal of International
Economics 18 (May 1985), pp. 199-217, and Roland Vaubel, "A Public Choice Approach to
International Organization," Public Choice 51 (1986), pp. 39-57. George Tsebelis ("Nested
Games: The Cohesion of French Coalitions," British Journal of Political Science 18 [April
1988], pp. 145-70) has developed a theory of "nested games," in which two alliances play a
competitive game to determine total payoffs, while the individual players within each alliance
contend over their shares. Fritz Sharpf ("A Game-Theoretical Interpretation of Inflation and
Unemployment in Western Europe," Journal of Public Policy 7 [1988], pp. 227-257) interprets
macroeconomic policy as the joint outcome of two simultaneous games; in one, the government
plays against the unions, while in the other, it responds to the anticipated reactions of the
electorate. James E. Alt and Barry Eichengreen ("Parallel and Overlapping Games: Theory
and an Application to the European Gas Trade," unpublished manuscript, November 1987)
offer a broader typology of linked games, distinguishing between "parallel" games, in which
"the same opponents play against one another at the same time in more than one arena," and'
"overlapping" games, which arise "when a particular player is engaged at the same time in
games against distinct opponents, and when the strategy pursued in one game limits the strategies
available in the other." Detailed comparison of these various linked-game models is a
task for the future.
442 International Organization
• Level II preferences and coalitions
• Level II institutions
• Level I negotiators' strategies
Let us consider each in turn.
/. The size of the win-set depends on the distribution of
power, preferences, and possible coalitions among Level II
constituents.
Any testable two-level theory of international negotiation must be rooted
in a theory of domestic politics, that is, a theory about the power and
preferences of the major actors at Level II. This is not the occasion for even
a cursory evaluation of the relevant alternatives, except to note that the twolevel
conceptual framework could in principle be married to such diverse
perspectives as Marxism, interest group pluralism, bureaucratic politics, and
neo-corporatism. For example, arms negotiations might be interpreted in
terms of a bureaucratic politics model of Level II politicking, while class
analysis or neo-corporatism might be appropriate for analyzing international
macroeconomic coordination.
Abstracting from the details of Level II politics, however, it is possible
to sketch certain principles that govern the size of the win-sets. For example,
the lower the cost of "no-agreement" to constituents, the smaller the winset.
44 Recall that ratification pits the proposed agreement, not against an
array of other (possibly attractive) alternatives, but only against "no-agreement."
45 No-agreement often represents the status quo, although in some
cases no-agreement may in fact lead to a worsening situation; that might be
a reasonable description of the failed ratification of the Versailles Treaty.
Some constituents may face low costs from no-agreement, and others high
costs, and the former will be more skeptical of Level I agreements than the
latter. Members of two-wage-earner families should be readier to strike, for
example, than sole breadwinners, and small-town barbers should be more
isolationist than international bankers. In this sense, some constituents may
offer either generic opposition to, or generic support for, Level I agreements,
more or less independently of the specific content of the agreement, although
naturally other constituents' decisions about ratification will be closely conditioned
on the specifics. The size of the win-set (and thus the negotiating
44. Thomas Romer and Howard Rosenthal, "Political Resource Allocation, Controlled Agendas,
and the Status Quo," Public Choice 33 (no. 4, 1978), pp. 27-44.
45. In more formal treatments, the no-agreement outcome is called the "reversion point."
A given constituent's evaluation of no-agreement corresponds to what Raiffa terms a seller's
"walk-away price," that is, the price below which he would prefer "no-deal." (Raiffa, Art and
Science of Negotiation.) No-agreement is equivalent to what Snyder and Diesing term "breakdown,"
or the expected cost of war. (Snyder and Diesing, Conflict Among Nations.)
Diplomacy and domestic politics 443
room of the Level I negotiator) depends on the relative size of the "isolationist"
forces (who oppose international cooperation in general) and the
"internationalists" (who offer "all-purpose" support). All-purpose support
for international agreements is probably greater in smaller, more dependent
countries with more open economies, as compared to more self-sufficient
countries, like the United States, for most of whose citizens the costs of noagreement
are generally lower. Ceteris paribus, more self-sufficient states
with smaller win-sets should make fewer international agreements and drive
harder bargains in those that they do make.
In some cases, evaluation of no-agreement may be the only significant
disagreement among the Level II constituents, because their interests are
relatively homogeneous. For example, if oil imports are to be limited by an
agreement among the consuming nations—the sort of accord sought at the
Tokyo summit of 1979, for example—then presumably every constituent
would prefer to maximize his nation's share of the available supply, although
some constituents may be more reluctant than others to push too hard, for
fear of losing the agreement entirely. Similarly, in most wage negotiations,
the interests of constituents (either workers or shareholders) are relatively
homogeneous, and the most significant cleavage within the Level II constituencies
is likely to be between "hawks" and "doves," depending on their
willingness to risk a strike. (Walton and McKersie refer to these as "boundary"
conflicts, in which the negotiator is caught between his constituency
and the external organization.) Other international examples in which domestic
interests are relatively homogeneous except for the evaluation of
no-agreement might include the SALT talks, the Panama Canal Treaty negotiations,
and the Arab-Israeli conflict. A negotiator is unlikely to face
criticism at home that a proposed agreement reduces the opponents' arms
too much, offers too little compensation for foreign concessions, or contains
too few security guarantees for the other side, although in each case opinions
may differ on how much to risk a negotiating deadlock in order to achieve
these objectives.
The distinctive nature of such "homogeneous" issues is thrown into sharp
relief by contrasting them to cases in which constituents' preferences are
more heterogeneous, so that any Level I agreement bears unevenly on them.
Thus, an internationally coordinated reflation may encounter domestic opposition
both from those who think it goes too far (bankers, for example)
and from those who think it does not go far enough (unions, for example).
In 1919, some Americans opposed the Versailles Treaty because it was too
harsh on the defeated powers and others because it was too lenient.46 Such
patterns are even more common, as we shall shortly see, where the negotiation
involves multiple issues, such as an arms agreement that involves
tradeoffs between seaborne and airborne weapons, or a labor agreement that
46. Thomas A. Bailey, Woodrow Wilson and the Great Betrayal (New York: Macmillan,
1945), pp. 16-37.
444 International Organization
involves tradeoffs between take-home pay and pensions. (Walton and McKersie
term these "factional" conflicts, because the negotiator is caught between
contending factions within his own organization.)
The problems facing Level I negotiators dealing with a homogeneous (or
"boundary") conflict are quite different from those facing negotiators dealing
with a heterogeneous (or "factional") conflict. In the former case, the more
the negotiator can win at Level I—the higher his national oil allocation, the
deeper the cuts in Soviet throw-weight, the lower the rent he promises for
the Canal, and so one—the better his odds of winning ratification. In such
cases, the negotiator may use the implicit threat from his own hawks to
maximize his gains (or minimize his losses) at Level I, as Carter and Vance
did in dealing with the Panamanians. Glancing over his shoulder at Level
II, the negotiator's main problem in a homogeneous conflict is to manage
the discrepancy between his constituents' expectations and the negotiable
outcome. Neither negotiator is likely to find much sympathy for the enemy's
demands among his own constituents, nor much support for his constituents'
positions in the enemy camp. The effect of domestic division, embodied in
hard-line opposition from hawks, is to raise the risk of involuntary defection
and thus to impede agreement at Level I. The common belief that domestic
politics is inimical to international cooperation no doubt derives from such
cases.
The task of a negotiator grappling instead with a heterogeneous conflict
is more complicated, but potentially more interesting. Seeking to maximize
the chances of ratification, he cannot follow a simple "the more, the better"
rule of thumb; imposing more severe reparations on the Germans in 1919
would have gained some votes at Level II but lost others, as would hastening
the decontrol of domestic oil prices in 1978. In some cases, these lines of
cleavage within the Level II constituencies will cut across the Level I division,
and the Level I negotiator may find silent allies at his opponent's
domestic table. German labor unions might welcome foreign pressure on
their own government to adopt a more expansive fiscal policy, and Italian
bankers might welcome international demands for a more austere Italian
monetary policy. Thus transnational alignments may emerge, tacit or explicit,
in which domestic interests pressure their respective governments to
adopt mutually supportive policies. This is, of course, my interpretation of
the 1978 Bonn summit accord.
In such cases, domestic divisions may actually improve the prospects for
international cooperation. For example, consider two different distributions
of constituents' preferences as between three alternatives: A, B, and noagreement.
If 45 percent of the constituents rank these A > no-agreement >
B, 45 percent rank them B > no-agreement > A, and 10 percent rank them
B > A > no-agreement, then both A and B are in the win-set, even though
B would win in a simple Level-II-only game. On the other hand, if 90 percent
rank the alternatives A > no-agreement > B, while 10 percent still rank
them B > A > no-agreement, then only A is in the win-set. In this sense,
Diplomacy and domestic politics 445
a government that is internally divided is more likely to be able to strike a
deal internationally than one that is firmly committed to a single policy.47
Conversely, to impose binding ex ante instructions on the negotiators in
such a case might exclude some Level I outcomes that would, in fact, be
ratifiable in both nations.48
Thus far we have implicitly assumed that all eligible constituents will
participate in the ratification process. In fact, however, participation rates
vary across groups and across issues, and this variation often has implications
for the size of the win-set. For example, when the costs and/or benefits
of a proposed agreement are relatively concentrated, it is reasonable to
expect that those constituents whose interests are most affected will exert
special influence on the ratification process.49 One reason why Level II games
are more important for trade negotiations than in monetary matters is that
the "abstention rate" is higher on international monetary issues than on
trade issues.50
The composition of the active Level II constituency (and hence the character
of the win-set) also varies with the politicization of the issue. Politicization
often activates groups who are less worried about the costs of noagreement,
thus reducing the effective win-set. For example, politicization
of the Panama Canal issue seems to have reduced the negotiating flexibility
on both sides of the diplomatic table.51 This is one reason why most professional
diplomats emphasize the value of secrecy to successful negotiations.
However, Woodrow Wilson's transcontinental tour in 1919 reflected the
opposite calculation, namely, that by expanding the active constituency he
could ensure ratification of the Versailles Treaty, although in the end this
strategy proved fruitless.52
Another important restriction of our discussion thus far has been the
47. Raiffa notes that "the more diffuse the positions are within each side, the easier it might
be to achieve external agreement." (Raiffa, Art and Science of Negotiation, p. 12.) For the
conventional view, by contrast, that domestic unity is generally a precondition for international
agreement, see Michael Artis and Sylvia Ostry, International Economic Policy Coordination,
Chatham House Papers: 30 (London: Routledge & Kegan Paul, 1986), pp. 75-76.
48. "Meaningful consultation with other nations becomes very difficult when the internal
process of decision-making already has some of the characteristics of compacts between quasisovereign
entities. There is an increasing reluctance to hazard a hard-won domestic consensus
in an international forum." Henry A. Kissinger, "Domestic Structure and Foreign Policy," in
James N. Rosenau, ed., International Politics and Foreign Policy (New York: Free Press,
1969), p. 266.
49. See James Q. Wilson, Political Organization (New York: Basic Books, 1975) on how the
politics of an issue are affected by whether the costs and the benefits are concentrated or
diffuse.
50. Another factor fostering abstention is the greater complexity and opacity of monetary
issues; as Gilbert R. Winham ("Complexity in International Negotiation," in Daniel Druckman,
ed., Negotiations: A Social-Psychological Perspective [Beverly Hills: Sage, 1977], p. 363)
observes, "complexity can strengthen the hand of a negotiator vis-a-vis the organization he
represents."
51. Habeeb and Zartman, Panama Canal Negotiations.
52. Bailey, Wilson and the Great Betrayal.
446 International Organization
assumption that the negotiations involve only one issue. Relaxing this assumption
has powerful consequences for the play at both levels.53 Various
groups at Level II are likely to have quite different preferences on the several
issues involved in a multi-issue negotiation. As a general rule, the group
with the greatest interest in a specific issue is also likely to hold the most
extreme position on that issue. In the Law of the Sea negotiations, for
example, the Defense Department felt most strongly about sea-lanes, the
Department of the Interior about sea-bed mining rights, and so on.54 If each
group is allowed to fix the Level I negotiating position for "its" issue, the
resulting package is almost sure to be "non-negotiable" (that is, non-ratifiable
in opposing capitals).55
Thus, the chief negotiator is faced with tradeoffs across different issues:
how much to yield on mining rights in order to get sea-lane protection, how
much to yield on citrus exports to get a better deal on beef, and so on. The
implication of these tradeoffs for the respective win-sets can be analyzed in
terms of iso-vote or "political indifference" curves. This technique is analogous
to conventional indifference curve analysis, except that the operational
measure is vote loss, not utility loss. Figure 2 provides an illustrative
Edgeworth box analysis.56 The most-preferred outcome for A (the outcome
which wins unanimous approval from both the beef industry and the citrus
industry) is the upper right-hand corner (AM), and each curve concave to
point AM represents the locus of all possible tradeoffs between the interests
of ranchers and farmers, such that the net vote in favor of ratification at A's
Level II is constant. The bold contour A,-A2 represents the minimal vote
necessary for ratification by A, and the wedge-shaped area northeast of
A,-A2 represents A's win-set. Similarly, B,-B2 represents the outcomes that
are minimally ratifiable by B, and the lens-shaped area between Ai-A2 and
Bi-B2 represents the set of feasible agreements. Although additional subtleties
(such as the nature of the "contract curve") might be extracted from
this sort of analysis, the central point is simple: the possibility of package
deals opens up a rich array of strategic alternatives for negotiators in a twolevel
game.
One kind of issue linkage is absolutely crucial to understanding how domestic
and international politics can become entangled.57 Suppose that a
majority of constituents at Level II oppose a given policy (say, oil price
53. I am grateful to Ernst B. Haas and Robert O. Keohane for helpful advice on this point.
54. Ann L. Hollick, U.S. Foreign Policy and the Law of the Sea (Princeton: Princeton
University Press, 1981), especially pp. 208-37, and James K. Sebenius, Negotiating the Law
of the Sea (Cambridge, Mass.: Harvard University Press, 1984), especially pp. 74-78.
55. Raiffa, Art and Science of Negotiation, p. 175.
56.1 am indebted to Lisa Martin and Kenneth Shepsle for suggesting this approach, although
they are not responsible for my application of it. Note that this construction assumes that each
issue, taken individually, is a "homogeneous" type, not a "heterogeneous" type. Constructing
iso-vote curves for heterogeneous-type issues is more complicated.
57. I am grateful to Henry Brady for clarifying this point for me.
Diplomacy and domestic politics 447
A,
FIGURE 2. Political indifference curves for two-issue negotiation
decontrol), but that some members of that majority would be willing to switch
their vote on that issue in return for more jobs (say, in export industries).
If bargaining is limited to Level II, that tradeoff is not technically feasible,
but if the chief negotiator can broker an international deal that delivers more
jobs (say, via faster growth abroad), he can, in effect, overturn the initial
outcome at the domestic table. Such a transnational issue linkage was a
crucial element in the 1978 Bonn accord.
Note that this strategy works not by changing the preferences of any
domestic constituents, but rather by creating a policy option (such as faster
export growth) that was previously beyond domestic control. Hence, I refer
to this type of issue linkage at Level I that alters the feasible outcomes at
Level II as synergistic linkage. For example, "in the Tokyo Round . . .
nations used negotiation to achieve internal reform in situations where constituency
pressures would otherwise prevent action without the pressure
448 International Organization
(and tradeoff benefits) that an external partner could provide."58 Economic
interdependence multiplies the opportunities for altering domestic coalitions
(and thus policy outcomes) by expanding the set of feasible alternatives in
this way—in effect, creating political entanglements across national boundaries.
Thus, we should expect synergistic linkage (which is, by definition,
explicable only in terms of two-level analysis) to become more frequent as
interdependence grows.
2. The size of the win-set depends on the Level II
political institutions.
Ratification procedures clearly affect the size of the win-set. For example,
if a two-thirds vote is required for ratification, the win-set will almost certainly
be smaller than if only a simple majority is required. As one experienced
observer has written: "Under the Constitution, thirty-four of the one
hundred senators can block ratification of any treaty. This is an unhappy
and unique feature of our democracy. Because of the effective veto power
of a small group, many worthy agreements have been rejected, and many
treaties are never considered for ratification."59 As noted earlier, the U.S.
separation of powers imposes a tighter constraint on the American win-set
than is true in many other countries. This increases the bargaining power
of American negotiators, but it also reduces the scope for international
cooperation. It raises the odds for involuntary defection and makes potential
partners warier about dealing with the Americans.
The Trade Expansion Act of 1974 modified U.S. ratification procedures
in an effort to reduce the likelihood of congressional tampering with the final
deal and hence to reassure America's negotiating partners. After the American
Selling Price fiasco, it was widely recognized that piecemeal congressional
ratification of any new agreement would inhibit international negotiation.
Hence, the 1974 Act guaranteed a straight up-or-down vote in Congress.
However, to satisfy congressional sensitivities, an elaborate system of private-
sector committees was established to improve communication between
the Level I negotiators and their Level II constituents, in effect coopting
the interest groups by exposing them directly to the implications of their
demands.60 Precisely this tactic is described in the labor-management case
by Walton and McKersie: "Instead of taking responsibility for directly persuading
the principals [Level II constituents] to reduce their expectations,
[the Level I negotiator] structures the situation so that they (or their more
immediate representatives) will persuade themselves."61
58. Gilbert R. Winham, "The Relevance of Clausewitz to a Theory of International Negotiation,"
prepared for delivery at the 1987 annual meeting of the American Political Science
Association.
59. Jimmy Carter, Keeping Faith: Memoirs of a President (New York: Bantam Books, 1982),
p. 225.
60. Winham (see note 37); Twiggs, The Tokyo Round.
61. Walton and McKersie, Behavioral Theory of Labor Organizations, p. 321.
Diplomacy and domestic politics 449
Not all significant ratification practices are formalized; for example, the
Japanese propensity for seeking the broadest possible domestic consensus
before acting constricts the Japanese win-set, as contrasted with majoritarian
political cultures. Other domestic political practices, too, can affect the size
of the win-set. Strong discipline within the governing party, for example,
increases the win-set by widening the range of agreements for which the
Level I negotiator can expect to receive backing. For example, in the 1986
House-Senate conference committee on tax reform, the final bill was closer
to the Senate version, despite (or rather, because of) Congressman Rostenkowski's
greater control of his delegation, which increased the House winset.
Conversely, a weakening of party discipline across the major Western
nations would, ceteris paribus, reduce the scope for international cooperation.
The recent discussion of "state strength" and "state autonomy" is relevant
here. The greater the autonomy of central decision-makers from their
Level II constituents, the larger their win-set and thus the greater the likelihood
of achieving international agreement. For example, central bank insulation
from domestic political pressures in effect increases the win-set and
thus the odds for international monetary cooperation; recent proposals for
an enhanced role for central bankers in international policy coordination
rest on this point.62 However, two-level analysis also implies that, ceteris
paribus, the stronger a state is in terms of autonomy from domestic pressures,
the weaker its relative bargaining position internationally. For example,
diplomats representing an entrenched dictatorship are less able than
representatives of a democracy to claim credibly that domestic pressures
preclude some disadvantageous deal.63 This is yet another facet of the disconcerting
ambiguity of the notion of "state strength."
For simplicity of exposition, my argument is phrased throughout in terms
of only two levels. However, many institutional arrangements require several
levels of ratification, thus multiplying the complexity (but perhaps also the
importance) of win-set analysis. Consider, for example, negotiations between
the United States and the European Community over agricultural trade.
According to the Treaty of Rome, modifications of the Common Agricultural
Policy require unanimous ratification by the Council of Ministers, representing
each of the member states. In turn, each of those governments must,
in effect, win ratification for its decision within its own national arena, and
in coalition governments, that process might also require ratification within
each of the parties. Similarly, on the American side, ratification would (informally,
at least) necessitate support from most, if not all, of the major
agricultural organizations, and within those organizations, further ratification
by key interests and regions might be required. At each stage, cleavage
patterns, issue linkages, ratification procedures, side-payments, negotiator
62. Artis and Ostry, International Economic Policy Coordination. Of course, whether this
is desirable in terms of democratic values is quite another matter.
63. Schelling, Strategy of Conflict, p. 28.
450 International Organization
strategies, and so on would need to be considered. At some point in this
analytic regress the complexity of further decomposition would outweigh
the advantages, but the example illustrates the need for careful thought about
the logic of multiple-level games.
3. The size of the win-set depends on the strategies of the
Level I negotiators.
Each Level I negotiator has an unequivocal interest in maximizing the
other side's win-set, but with respect to his own win-set, his motives are
mixed. The larger his win-set, the more easily he can conclude an agreement,
but also the weaker his bargaining position vis-a-vis the other negotiator.
This fact often poses a tactical dilemma. For example, one effective way to
demonstrate commitment to a given position in Level I bargaining is to rally
support from one's constituents (for example, holding a strike vote, talking
about a "missile gap," or denouncing "unfair trading practices" abroad).
On the other hand, such tactics may have irreversible effects on constituents'
attitudes, hampering subsequent ratification of a compromise agreement.64
Conversely, preliminary consultations at home, aimed at "softening up"
one's constituents in anticipation of a ratification struggle, can undercut a
negotiator's ability to project an implacable image abroad.
Nevertheless, disregarding these dilemmas for the moment and assuming
that a negotiator wishes to expand his win-set in order to encourage ratification
of an agreement, he may exploit both conventional side-payments
and generic "good will." The use of side-payments to attract marginal supporters
is, of course, quite familiar in game theory, as well as in practical
politics. For example, the Carter White House offered many inducements
(such as public works projects) to help persuade wavering Senators to ratify
the Panama Canal Treaty.65 In a two-level game the side-payments may
come from unrelated domestic sources, as in this case, or they may be
received as part of the international negotiation.
The role of side-payments in international negotiations is well known.
However, the two-level approach emphasizes that the value of an international
side-payment should be calculated in terms of its marginal contribution
to the likelihood of ratification, rather than in terms of its overall value to
the recipient nation. What counts at Level II is not total national costs and
benefits, but their incidence, relative to existing coalitions and protocoalitions.
An across-the-board trade concession (or still worse, a concession
on a product of interest to a committed free-trade congressman) is less
effective than a concession (even one of lesser intrinsic value) that tips the
balance with a swing voter. Conversely, trade retaliation should be targeted,
64. Walton and McKersie, Behavioral Theory of Labor Organizations, p. 345.
65. Carter, Keeping Faith, p. 172. See also Raiffa, Art and Science of Negotiation, p. 183.
Diplomacy and domestic politics 451
neither at free-traders nor at confirmed protectionists, but at the uncommitted.
An experienced negotiator familiar with the respective domestic tables
should be able to maximize the cost-effectiveness (to him and his constituents)
of the concessions that he must make to ensure ratification abroad,
as well as the cost-effectiveness of his own demands and threats, by targeting
his initiatives with an eye to their Level II incidence, both at home and
abroad. In this endeavor Level I negotiators are often in collusion, since
each has an interest in helping the other to get the final deal ratified. In
effect, they are moving jointly towards points of tangency between their
respective political indifference curves. The empirical frequency of such
targeting in trade negotiations and trade wars, as well as in other international
negotiations, would be a crucial test of the relative merits of conventional
unitary-actor analysis and the two-level approach proposed here.66
In addition to the use of specific side-payments, a chief negotiator whose
political standing at home is high can more easily win ratification of his
foreign initiatives. Although generic good will cannot guarantee ratification,
as Woodrow Wilson discovered, it is useful in expanding the win-set and
thus fostering Level I agreement, for it constitutes a kind of "all-purpose
glue" for his supporting coalition. Walton and McKersie cite members of
the United Auto Workers who, speaking of their revered leader, Walter
Reuther, said, "I don't understand or agree with this profit-sharing idea, but
if the Red Head wants it, I will go along."67 The Yugoslav negotiator in the
Trieste dispute later discounted the difficulty of persuading irredentist Slovenes
to accept the agreement, since "the government [i.e., Tito] can always
influence public opinion if it wants to."68
Note that each Level I negotiator has a strong interest in the popularity
of his opposite number, since Party A's popularity increases the size of his
win-set, and thus increases both the odds of success and the relative bargaining
leverage of Party B. Thus, negotiators should normally be expected
to try to reinforce one another's standing with their respective constituents.
66. The strategic significance of targeting at Level II is illustrated in John Conybeare, "Trade
Wars: A Comparative Study of Anglo-Hanse, Franco-Italian, and Hawley-Smoot Conflicts,"
World Politics 38 (October 1985), p. 157: Retaliation in the Anglo-Hanse trade wars did not
have the intended deterrent effect, because it was not (and perhaps could not have been) targeted
at the crucial members of the opposing Level II coalition. Compare Snyderand Diesing, Conflict
Among Nations, p. 552: "If one faces a coercive opponent, but the opponent's majority coalition
includes a few wavering members inclined to compromise, a compromise proposal that suits
their views may cause their defection and the formation of a different majority coalition. Or if
the opponent's strategy is accommodative, based on a tenuous soft-line coalition, one knows
that care is required in implementing one's own coercive stretegy to avoid the opposite kind
of shift in the other state."
67. Walton and McKersie, Behavioral Theory of Labor Negotiations, p. 319.
68. Vladimir Velebit, in Campbell, Trieste 1954, p. 97. As noted earlier, our discussion here
assumes that the Level I negotiator wishes to reach a ratifiable agreement; in cases (alluded
to later) when the negotiator's own preferences are more hard-line than his constituents, his
domestic popularity might allow him to resist Level I agreements.
452 International Organization
Partly for this reason and partly because of media attention, participation
on the world stage normally gives a head of government a special advantage
vis-a-vis his or her domestic opposition. Thus, although international policy
coordination is hampered by high transaction costs, heads of government
may also reap what we might term "transaction benefits." Indeed, the recent
evolution of Western summitry, which has placed greater emphasis on publicity
than on substance, seems designed to appropriate these "transaction
benefits" without actually seeking the sort of agreements that might entail
transaction costs.69
Higher status negotiators are likely to dispose of more side-payments and
more "good will" at home, and hence foreigners prefer to negotiate with a
head of government than with a lower official. In purely distributive terms,
a nation might have a bargaining advantage if its chief negotiator were a
mere clerk. Diplomats are acting rationally, not merely symbolically, when
they refuse to negotiate with a counterpart of inferior rank. America's negotiating
partners have reason for concern whenever the American president
is domestically weakened.
Uncertainty and bargaining tactics
Level I negotiators are often badly misinformed about Level II politics,
particularly on the opposing side. In 1978, the Bonn negotiators were usually
wrong in their assessments of domestic politics abroad; for example, most
American officials did not appreciate the complex domestic game that Chancellor
Schmidt was playing over the issue of German reflation. Similarly,
Snyder and Diesing report that "decision makers in our cases only occasionally
attempted such assessments, and when they tried they did pretty
miserably. . . . Governments generally do not do well in analyzing each
other's internal politics in crises [and, I would add, in normal times], and
indeed it is inherently difficult."70 Relaxing the assumption of perfect information
to allow for uncertainty has many implications for our understanding
of two-level games. Let me illustrate a few of these implications.
Uncertainty about the size of a win-set can be both a bargaining device
and a stumbling block in two-level negotiation. In purely distributive Level
I bargaining, negotiators have an incentive to understate their own win-sets.
Since each negotiator is likely to know more about his own Level II than
his opponent does, the claim has some plausibility. This is akin to a tactic
69. Transaction benefits may be enhanced if a substantive agreement is reached, although
sometimes leaders can benefit domestically by loudly rejecting a proffered international deal.
70. Snyder and Diesing, Conflict Among Nations, pp. 516,522-23. Analogous misperceptions
in Anglo-American diplomacy are the focus of Richard E. Neustadt, Alliance Politics (New
York: Columbia University Press, 1970).
Diplomacy and domestic politics 453
that Snyder and Diesing describe, when negotiators seek to exploit divisions
within their own government by saying, in effect, "You'd better make a
deal with me, because the alternative to me is even worse."71
On the other hand, uncertainty about the opponent's win-set increases
one's concern about the risk of involuntary defection. Deals can only be
struck if each negotiator is convinced that the proposed deal lies within his
opposite number's win-set and thus will be ratified. Uncertainty about party
A's ratification lowers the expected value of the agreement to party B, and
thus party B will demand more generous side-payments from party A than
would be needed under conditions of certainty. In fact, party B has an
incentive to feign doubt about party A's ability to deliver, precisely in order
to extract a more generous offer.72
Thus, a utility-maximizing negotiator must seek to convince his opposite
number that his own win-set is "kinky," that is, that the proposed deal is
certain to be ratified, but that a deal slightly more favorable to the opponent
is unlikely to be ratified. For example, on the energy issue in 1978, by sending
Senator Byrd on a personal mission to Bonn before the summit and then by
discussing his political problems in a length tete-a-tete with the chancellor,
Carter sought successfully to convince Schmidt that immediate decontrol
was politically impossible, but that decontrol by 1981 was politically doable.
Kinky win-sets may be more credible if they pivot on what Schelling
calls a "prominent" solution, such as a 50-50 split, for such outcomes may
be distinctly more "saleable" at home. Another relevant tactic is for the
negotiator actually to submit a trial agreement for ratification, in order to
demonstrate that it is not in his win-set.
Uncertainty about the contours of the respective "political indifference
curves" thus has strategic uses. On the other hand, when the negotiators
are seeking novel packages that might improve both sides' positions, misrepresentation
of one's win-set can be counterproductive. Creative solutions
that expand the scope for joint gain and improve the odds of ratification are
likely to require fairly accurate information about constituents' preferences
and points of special neuralgia. The analysis of two-level games offers many
illustrations of Zartman's observation that all negotiation involves "the controlled
exchange of partial information."73
71. Synder and Diesing, Conflict Among Nations, p. 517.
72. I am grateful to Robert 0. Keohane for pointing out the impact of uncertainty on the
expected value of proposals.
73. I. William Zartman, The 50% Solution (Garden City, N.J.: Anchor Books, 1976), p. 14.
The present analysis assumes that constituents are myopic about the other side's Level II, an
assumption that is not unrealistic empirically. However, a fully informed constituent would
consider the preferences of key players on the other side, for if the current proposal lies well
within the other side's win-set, then it would be rational for the constituent to vote against it,
hoping for a second-round proposal that was more favorable to him and still ratifiable abroad;
this might be a reasonable interpretation of Senator Lodge's position in 1919 (Bailey, Wilson
and the Great Betrayal). Consideration of such strategic voting at Level II is beyond the scope
of this article.
454 International Organization
Restructuring and reverberation
Formally speaking, game-theoretic analysis requires that the structure of
issues and payoffs be specified in advance. In reality, however, much of
what happens in any bargaining situation involves attempts by the players
to restructure the game and to alter one another's perceptions of the costs
of no-agreement and the benefits of proposed agreements. Such tactics are
more difficult in two-level games than in conventional negotiations, because
it is harder to reach constituents on the other side with persuasive messages.
Nevertheless, governments do seek to expand one another's win-sets. Much
ambassadorial activity—wooing opinion leaders, establishing contact with
opposition parties, offering foreign aid to a friendly, but unstable government,
and so on—has precisely this function. When Japanese officials visit
Capitol Hill, or British diplomats lobby Irish-American leaders, they are
seeking to relax domestic constraints that might otherwise prevent the administration
from cooperating with their governments.
Another illuminating example of actions by a negotiator at the opposing
Level II to improve the odds of ratification occurred during the 1977 negotiations
between the International Monetary Fund and the Italian government.
Initial IMF demands for austerity triggered strong opposition from
the unions and left-wing parties. Although the IMF's bargaining position at
Level I appeared strong, the Fund's negotiator sought to achieve a broader
consensus within Italy in support of an agreement, in order to forestall
involuntary defection. Accordingly, after direct consultations with the unions
and leftist leaders, the IMF restructured its proposal to focus on long-term
investment and economic recovery (incidentally, an interesting example of
targeting), without backing off from its short-term demands. Ironically, the
initial Communist support for this revised agreement subsequently collapsed
because of conflicts between moderate and doctrinaire factions within the
party, illustrating the importance of multilevel analysis.74
In some instances, perhaps even unintentionally, international pressures
"reverberate" within domestic politics, tipping the domestic balance and
thus influencing the international negotiations. Exactly this kind of reverberation
characterized the 1978 summit negotiations. Dieter Hiss, the German
sherpa and one of those who believed that a stimulus program was in
Germany's own interest, later wrote that summits change national policy
only insofar as they mobilize and/or change public opinion and the attitude
of political groups. . . . Often that is enough, if the balance of
74. John R. Hillman, "The Mutual Influence of Italian Domestic Politics and the International
Monetary Fund," The Fletcher Forum 4 (Winter 1980), pp. 1-22. Luigi Spaventa, "Two Letters
of Intent: External Crises and Stabilization Policy, Italy, 1973-77," in John Williamson, ed.,
IMF Conditionality (Washington, D.C.: Institute for International Economics, 1983), pp. 441-73,
argues that the unions and the Communists actually favored the austerity measures, but found
the IMF demands helpful in dealing with their own internal Level II constituents.
Diplomacy and domestic politics 455
opinion is shifted, providing a bare majority for the previously stymied
actions of a strong minority. . . . No country violates its own interests,
but certainly the definition of its interests can change through a summit
with its possible tradeoffs and give-and-take.75
From the point of view of orthodox social-choice theory, reverberation is
problematic, for it implies a certain interconnectedness among the utility
functions of independent actors, albeit across different levels of the game.
Two rationales may be offered to explain reverberation among utilitymaximizing
egoists. First, in a complex, interdependent, but often unfriendly
world, offending foreigners may be costly in the long run. "To get along,
go along" may be a rational maxim. This rationale is likely to be more
common the more dependent (or interdependent) a nation, and it is likely
to be more persuasive to Level II actors who are more exposed internationally,
such as multinational corporations and international banks.
A second rationale takes into account cognitive factors and uncertainty.
It would be a mistake for political scientists to mimic most economists'
disregard for the suasive element in negotiations.76 Given the pervasive
uncertainty that surrounds many international issues, messages from abroad
can change minds, move the undecided, and hearten those in the domestic
minority. As one reluctant German latecomer to the "locomotive" cause in
1978 explained his conversion, "In the end, even the Bank for International
Settlements [the cautious Basle organization of central bankers] supported
the idea of coordinated relation." Similarly, an enthusiastic advocate of the
program welcomed the international pressure as providing a useful "tailwind"
in German domestic politics.
Suasive reverberation is more likely among countries with close relations
and is probably more frequent in economic than in political-military negotiations.
Communique's from the Western summits are often cited by participants
to domestic audiences as a way of legitimizing their policies. After
one such statement by Chancellor Schmidt, one of his aides privately characterized
the argument as "not intellectually valid, but politically useful."
Conversely, it is widely believed by summit participants that a declaration
contrary to a government's current policy could be used profitably by its
opponents. Recent congressional proposals to ensure greater domestic publicity
for international commentary on national economic policies (including
hitherto confidential IMF recommendations) turn on the idea that reverberation
might increase international cooperation.77
75. Dieter Hiss, "Weltwirtschaftsgipfel: Betrachtungen eines Insiders [World Economic Summit:
Observations of an Insider]," in Joachim Frohn and Reiner Staeglin, eds., Empirische
Wirtschaftsforschung (Berlin: Duncker and Humblot, 1980), pp. 286-87.
76. On cognitive and communications explanations of international cooperation, see, for
example, Ernst B. Haas, "Why Collaborate? Issue-Linkage and International Regimes," World
Politics 32 (April 1980), pp 357-405; Richard N. Cooper, "International Cooperation in Public
Health as a Prologue to Macroeconomic Cooperation," Brookings Discussion Papers in International
Economics 44 (Washington, D.C.: Brookings Institution, 1986); and Zartman, 50%
Solution, especially Part 4.
77. Henning, Macroeconomic Diplomacy in the 1980s, pp. 62-63.
456 International Organization
Reverberation as discussed thus far implies that international pressure
expands the domestic win-set and facilitates agreement. However, reverberation
can also be negative, in the sense that foreign pressure may create
a domestic backlash. Negative reverberation is probably less common empirically
than positive reverberation, simply because foreigners are likely to
forgo public pressure if it is recognized to be counterproductive. Cognitive
balance theory suggests that international pressure is more likely to reverberate
negatively if its source is generally viewed by domestic audiences as
an adversary rather than an ally. Nevertheless, predicing the precise effect
of foreign pressure is admittedly difficult, although empirically, reverberation
seems to occur frequently in two-level games.
The phenomenon of reverberation (along with synergistic issue linkage of
the sort described earlier) precludes one attractive short-cut to modeling
two-level games. If national preferences were exogenous from the point of
view of international relations, then the domestic political game could be
molded separately, and the "outputs" from that game could be used as the
"inputs" to the international game.78 The division of labor between comparative
politics and international relations could continue, though a few
curious observers might wish to keep track of the play on both tables. But
if international pressures reverberate within domestic politics, or if issues
can be linked synergistically, then domestic outcomes are not exogenous,
and the two levels cannot be modeled independently.
The role of the chief negotiator
In the stylized model of two-level negotiations outlined here, the chief negotiator
is the only formal link between Level I and Level II. Thus far, I
have assumed that the chief negotiator has no independent policy views, but
acts merely as an honest broker, or rather as an agent on behalf of his
constituents. That assumption powerfully simplifies the analysis of two-level
games. However, as principal-agent theory reminds us, this assumption is
unrealistic.79 Empirically, the preferences of the chief negotiator may well
diverge from those of his constituents. Two-level negotiations are costly and
78. This is the approach used to analyze the Anglo-Chinese negotiations over Hong Kong
in Bruce Bueno de Mesquita, David Newman, and Alvin Rabushka, Forecasting Political
Events: The Future of Hong Kong (New Haven: Yale University Press, 1985).
79. For overviews of this literature, see Terry M. Moe, "The New Economics of Organization,"
American Journal of Political Science 28 (November 1984), pp. 739-77; John W. Pratt
and Richard J. Zeckhauser, eds., Principals and Agents: The Structure of Business (Boston,
Mass.: Harvard Business School Press, 1985); and Barry M. Mitnick, "The Theory of Agency
and Organizational Analysis," prepared for delivery at the 1986 annual meeting of the American
Political Science Association. This literature is only indirectly relevant to our concerns here,
for it has not yet adequately addressed the problems posed by multiple principals (or constituents,
in our terms). For one highly formal approach to the problem of multiple principals, see
R. Douglas Bernheim and Michael D. Whinston, "Common Agency," Econometrica 54 (July
1986), pp. 923-42.
Diplomacy and domestic politics 457
risky for the chief negotiator, and they often interfere with his other priorities,
so it is reasonable to ask what is in it for him.
The motives of the chief negotiator include:
1. Enhancing his standing in the Level II game by increasing his political
resources or by minimizing potential losses. For example, a head
of government may seek the popularity that he expects to accrue to
him if he concludes a successful international agreement, or he may
anticipate that the results of the agreement (for example, faster
growth or lower defense spending) will be politically rewarding.
2. Shifting the balance of power at Level II in favor of domestic policies
that he prefers for exogenous reasons. International negotiations
sometimes enable government leaders to do what they privately wish
to do, but are powerless to do domestically. Beyond the now-familiar
1978 case, this pattern characterizes many stabilization programs that
are (misleadingly) said to be "imposed" by the IMF. For example, in
the 1974 and 1977 negotiations between Italy and the IMF, domestic
conservative forces exploited IMF pressure to facilitate policy moves
that were otherwise infeasible internally.80
3. To pursue his own conception of the national interest in the international
context. This seems the best explanation of Jimmy Carter's
prodigious efforts on behalf of the Panama Canal Treaty, as well as
of Woodrow Wilson's ultimately fatal commitment to the Versailles
Treaty.
It is reasonable to presume, at least in the international case of two-level
bargaining, that the chief negotiator will normally give primacy to his domestic
calculus, if a choice must be made, not least because his own incumbency
often depends on his standing at Level II. Hence, he is more likely
to present an international agreement for ratification, the less of his own
political capital he expects to have to invest to win approval, and the greater
the likely political returns from a ratified agreement.
This expanded conception of the role of the chief negotiator implies that
he has, in effect, a veto over possible agreements. Even if a proposed deal
lies within his Level II win-set, that deal is unlikely to be struck if he opposes
it.81 Since this proviso applies on both sides of the Level I table, the actual
international bargaining set may be narrower—perhaps much narrower—
than the overlap between the Level II win-sets. Empirically, this additional
constraint is often crucial to the outcome of two-level games. One momentous
example is the fate of the Versailles Treaty. The best evidence suggests,
first, that perhaps 80 percent of the American public and of the Senate in
1919 favored ratification of the treaty, if certain reservations were attached,
and second, that those reservations were acceptable to the other key sig-
80. Hillman, "Mutual Influence," and Spaventa, "Two Letters of Intent."
81. This power of the chief negotiator is analogous to what Shepsle and Weingast term the
"penultimate" or "ex post veto" power of the members of a Senate-House conference committee.
(Shepsle and Weingast, "Institutional Foundations of Committee Power.")
458 International Organization
natories, especially Britain and France. In effect, it was Wilson himself who
vetoed this otherwise ratifiable package, telling the dismayed French Ambassador,
"I shall consent to nothing."82
Yet another constraint on successful two-level negotiation derives from
the leader's existing domestic coalition. Any political entrepreneur has a
fixed investment in a particular pattern of policy positions and a particular
supporting coalition. If a proposed international deal threatens that investment,
or if ratification would require him to construct a different coalition,
the chief negotiator will be reluctant to endorse it, even if (judged abstractly)
it could be ratified. Politicians may be willing to risk a few of their normal
supporters in the cause of ratifying an international agreement, but the greater
the potential loss, the greater their reluctance.
In effect, the fixed costs of coalition-building thus imply this constraint
on the win-set: How great a realignment of prevailing coalitions at Level II
would be required to ratify a particular proposal? For example, a trade deal
may expand export opportunities for Silicon Valley, but harm Aliquippa.
This is fine for a chief negotiator (for example, Reagan?) who can easily add
Northern California yuppies to his support coalition and who has no hope
of winning Aliquippa steelworkers anyhow. But a different chief negotiator
with a different support coalition (for example, Mondale?) might find it costly
or even impossible to convert the gains from the same agreement into politically
usable form. Similarly, in the 1978 "neutron bomb" negotiations
between Bonn and Washington, "asking the United States to deploy [these
weapons] in West Germany might have been possible for a Christian Democratic
Government; for a Social Democratic government, it was nearly
impossible."83 Under such circumstances, simple "median-voter" models
of domestic influences on foreign policy may be quite misleading.
Relaxing the assumption that the chief negotiator is merely an honest
broker, negotiating on behalf of his constituents, opens the possibility that
the constituents may be more eager for an agreement (or more worried about
"no-agreement") than he is. Empirical instances are not hard to find: in
early 1987, European publics were readier to accept Gorbachev's "doublezero"
arms control proposal than European leaders, just as in the early 1970s
the American public (or at least the politically active public) was more eager
for a negotiated end to the Vietnam War than was the Nixon administration.
As a rule, the negotiator retains a veto over any proposed agreement in such
cases. However, if the negotiator's own domestic standing (or indeed, his
incumbency) would be threatened if he were to reject an agreement that falls
within his Level II win-set, and if this is known to all parties, then the other
side at Level I gains considerable leverage. Domestic U.S. discontent about
82. Bailey, Wilson and the Great Betrayal, quotation at p. 15.
83. Robert A. Strong and Marshal Zeringue, "The Neutron Bomb and the Atlantic Alliance,"
presented at the 1986 annual meeting of the American Political Science Association, p. 9.
Diplomacy and domestic politics 459
the Vietnam War clearly affected the agreement reached at the Paris talks.84
Conversely, if the constituents are (believed to be) hard-line, then a leader's
domestic weakness becomes a diplomatic asset. In 1977, for example, the
Americans calculated that "a delay in negotiating a treaty . . . endangered
[Panamanian President Omar] Torrijos' position; and Panama without Torrijos
most likely would have been an impossible negotiating partner."85 Similarly,
in the 1954 Trieste negotiations, the weak Italian government claimed
that '"Unless something is done in our favor in Trieste, we can lose the
election.' That card was played two or three times [reported the British
negotiator later], and it almost always took a trick."86
My emphasis on the special responsibility of central executives is a point
of affinity between the two-level game model and the "state-centric" literature,
even though the underlying logic is different. In this "Janus" model
of domestic-international interactions, transnational politics are less prominent
than in some theories of interdependence.87 However, to disregard
"cross-table" alliances at Level II is a considerable simplification, and it is
more misleading, the lower the political visibility of the issue, and the more
frequent the negotiations between the governments involved.88 Empirically,
for example, two-level games in the European Community are influenced
by many direct ties among Level II participants, such as national agricultural
spokesmen. In some cases, the same multinational actor may actually appear
at more than one Level II table. In negotiations over mining concessions in
some less-developed countries, for example, the same multinational corporation
may be consulted privately by both the home and host governments.
In subsequent work on the two-level model, the strategic implications of
direct communication between Level II players should be explored.
Conclusion
The most portentous development in the fields of comparative politics and
international relations in recent years is the dawning recognition among
practitioners in each field of the need to take into account entanglements
between the two. Empirical illustrations of reciprocal influence between
domestic and international affairs abound. What we need now are concepts
84. I. William Zartman, "Reality, Image, and Detail: The Paris Negotiations, 1969-1973,"
in Zartman, 50% Solution, pp. 372-98.
85. Zbigniew Brzezinski, Power and Principle (New York: Farrar, Straus and Giroux, 1983),
p. 136, as quoted in Habeeb and Zartman, Panama Canal Negotiations, pp. 39-40.
86. Harrison in Campbell, Trieste 1954, p. 67.
87. Samuel P. Huntington, "Transnational Organizations in World Politics," World Politics
25 (April 1973), pp. 333-68; Keohane and Nye, Power and Interdependence; Neustadt, Alliance
Politics.
88. Barbara Crane, "Policy Coordination by Major Western Powers in Bargaining with the
Third World: Debt Relief and the Common Fund," International Organization 38 (Summer
1984), pp. 399-428.
460 International Organization
and theories that will help us organize and extend our empirical observations.
Analysis in terms of two-level games offers a promising response to this
challenge. Unlike state-centric theories, the two-level approach recognizes
the inevitability of domestic conflict about what the "national interest"
requires. Unlike the "Second Image" or the "Second Image Reversed,"
the two-level approach recognizes that central decision-makers strive to
reconcile domestic and international imperatives simultaneously. As we have
seen, statesmen in this predicament face distinctive strategic opportunities
and strategic dilemmas.
This theoretical approach highlights several significant features of the links
between diplomacy and domestic politics, including:
• the important distinction between voluntary and involuntary defection
from international agreements;
• the contrast between issues on which domestic interests are homogeneous,
simply pitting hawks against doves, and issues on which domestic
interests are more heterogeneous, so that domestic cleavage may actually
foster international cooperation;
• the possibility of synergistic issue linkage, in which strategic moves at
one game-table facilitate unexpected coalitions at the second table;
• the paradoxical fact that institutional arrangements which strengthen
decision-makers at home may weaken their international bargaining position,
and vice versa;
• the importance of targeting international threats, offers, and sidepayments
with an eye towards their domestic incidence at home and
abroad;
• the strategic uses of uncertainty about domestic politics, and the special
utility of "kinky win-sets";
• the potential reverberation of international pressures within the domestic
arena;
• the divergences of interest between a national leader and those on
whose behalf he is negotiating, and in particular, the international implications
of his fixed investments in domestic politics.
Two-level games seem a ubiquitous feature of social life, from Western
economic summitry to diplomacy in the Balkans and from coalition politics
in Sri Lanka to legislative maneuvering on Capitol Hill. Far-ranging empirical
research is needed now to test and deepen our understanding of how such
games are played.