World
Presidential Operational Codes and Foreign Policy Conflicts in the Post-Cold War
®
Stephen G. Walker; Mark Schäfer; Michael D. Young
The Journal of Conflict Resolution, Vol. 43, No. 5. (Oct., 1999), pp. 610-625. Stable URL:
http://links.jstor.org/sici?sici=0022-0027%28199910%2943%3A5%3C610%3APOCAFP%3E2.0.CO%3B2-I The Journal of Conflict Resolution is currently published by Sage Publications, Inc..
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/sage.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org.
http: //www. j s tor. org
Tue Oct 16 06:44:59 2007
Presidential Operational Codes and Foreign Policy Conflicts in the Post-Cold War World
STEPHEN G. WALKER
Department of Political Science Arizona State University MARK SCHÄFER Department of Political Science Louisiana State University
MICHAEL D. YOUNG
Social Science Automation Inc. Milliard, Ohio
The authors investigate the linkage between presidential operational codes and the management of foreign policy conflicts during the period of strategic adjustment in American foreign policy following the cold war. Beliefs expressed in public speeches by Presidents George Bush and Bill Clinton are coded for self and other attributions that represent different forms of the exercise of political power. Bush's beliefs reveal a less cooperative, relatively inflexible approach to conflict management in the foreign policy domain, whereas Clinton's beliefs indicate a more flexible and cooperative approach. Their orientations interacted with contextual variables and the opponents' behavior to shape the selection of U.S. behavior in four post-cold war conflicts: Panama, Haiti, the Persian Gulf, and Bosnia. A favorable power position and the absence of vital or strategic U.S. interests enhanced the effect of presidential operational codes.
A s the United States emerges from the era of cold war confrontation with the Soviet Union, the strategic context for the use of force has changed in important ways for American leaders while remaining the same in other respects. The paramount position of the United States as the last superpower in the post-cold war world has prompted U.S. presidents to take the lead in managing conflicts such as the Iraqi invasion of Kuwait and the hostilities among the warring factions in Bosnia. These challenges come from less familiar sources and engage issues that are less clearly related to U.S. national interests than the ones associated with the old superpower rivalry. At the same time, a traditional hegemonic position in the Western Hemisphere continues to present the United States with opportunities to intervene in places such as Panama and Haiti.
AUTHORS' NOTE: This is a revised version of a paper presented at the 1996 annual meeting of the American Political Science Association, August, San Francisco. The data for this article are available from Mark Schafer, Department of Political Science, Louisiana State University, Baton Rouge, LA 70803. The authors gratefully acknowledge the research assistance of Brian Dille, Greg Marfleet, and Scott Crichlow.
JOURNAL OF CONFLICT RESOLUTION, Vol. 43 No. 5, October 1999 610-625 © 1999 Sage Publications, Inc.
610
Walker et al. / PRESIDENTIAL OPERATIONAL CODES    611
Presidents George Bush and Bill Clinton faced challenges during their respective administrations that led them to use force in the conduct of foreign policy conflicts between 1989 and 1995. In Latin America, Bush decided to land troops in Panama to oust Manuel Noriega, whereas Clinton used the threat of imminent military force to remove Raoul Cedras from power in Haiti (Buckley 1991; Donnelly, Roth, and Baker 1991; Woodward 1991; Flanagan 1993; Drew 1994). In addition to these smaller scale skirmishes, each chief executive responded with force to more serious threats in the Persian Gulf and the Balkans. The conflict with Iraq escalated into Operation Desert Storm under Bush's leadership before a cease-fire agreement was signed (Woodward 1991; Hybel 1993; Wayne 1993; Mervin 1996), and the hostilities between Serbs and Muslims in Bosnia finally resulted in military intervention by the North Atlantic Treaty Organization (NATO) under American leadership until the successful negotiation of the Dayton Accords (Drew 1994; Zimmerman 1996).
Even in each of these more serious confrontations, there was little doubt about the balance of power and the outcome of the conflict if war were to occur. As the last superpower, the United States was an ascendant state with low global vulnerability when the cold war ended. Local conditions at the onset of the Panama, Haiti, Persian Gulf, and Bosnian conflicts reinforce the assessment of low U.S. vulnerability. U.S. interests in all four cases were secondary (little threat to American lives, property, or territory), whereas the opponents' interests were vital (the survival of their respective regimes). The power distributions also clearly favored the United States, although the distributions of power and interests were less asymmetrical in the Persian Gulf and Bosnian cases than in the other two cases. The opponents' capacity to resist was greater, and the engagement of American interests took the form of a strategic threat to a U.S. ally (Saudi Arabia or Greece) while stopping short of direct harm to U.S. vital interests (territory, lives, or property).
The process of adjustment to a new strategic context experienced by the United States with the sudden end of the cold war is not altogether unique. As Kupchan (1994) points out, great powers often face a rapid shift in their security environment. In addition, Kupchan argues that states respond in different ways depending on their vulnerability, defined as "elite perceptions of the relative balance of power . . . [and] their assessment of the likely outcome of war" (pp. 15-16, n. 22). Under the condition of low vulnerability (a favorable balance of power), declining states tend to adapt with "timely balancing" by deterring adversaries at the core and accommodating them on the periphery of their spheres of influence. Ascendant states are likely to exhibit "paced imperial growth," deterring and compelling adversaries everywhere in response to threats. Under the condition of high vulnerability (an unfavorable balance of power), their adjustment tends to be flawed by overextension and either "strategic exposure" as a consequence of accommodating core adversaries and deterring peripheral foes in the case of a declining power or "self-encirclement" induced by overly competitive behavior toward all adversaries in the case of a rising power (pp. 16-17, 68-69).
In this model, the beliefs of elites play important roles in the process of strategic adjustment. When these beliefs become more or less fixed and relatively unresponsive to incoming information, they can autonomously affect the state's adjustment to new
612    JO URNAL OF CONFLICT RESOLUTION
strategic realities. Kupchan (1994) argues that the processes of cognitive closure and strategic maladjustment are more likely to occur under the condition of high vulnerability, that is, when the state's resources are not sufficient to deal with threats to its security and domestic political constraints or the impact of external events makes it difficult to adjust old beliefs to new realities. Conversely, under the condition of low vulnerability when the state's resources are sufficient to deal with security threats, beliefs should be less autonomous and shift readily in response to changes in the strategic environment (Kupchan 1994; see also Snyder 1991).
Elite decisions, therefore, are not merely responses to material constraints and environmental stimuli. Beliefs interact with external conditions and events to provide an explanation of foreign policy behavior. We concur with Kupchan's (1994, 491) positive assessment of "the analytic utility of the notion of the operational code or belief
system____More research should focus on the content of elite beliefs and the historical
trajectory of the ideas and suppositions that shape policy." However, introducing beliefs into the analysis also raises a rival theoretical claim about the likely causal connection between autonomous beliefs and behavior.
A psychologically oriented approach to foreign policy emphasizes the pervasive influence of boundaries on decisions imposed by cognitive mechanisms and personality structure rather than the environmental constraints represented by the contextual variables of power, interests, and domestic opinion. Beliefs and behavior are consistent due to the internal stability of beliefs and the tendency of decision makers either to fit incoming information into already existing images or to ignore it as the basis for decisions and act according to previously held beliefs (Jervis 1976; Holsti 1976; Herrmann 1988; Vertzberger 1990,1998; Taber 1992,1998; Tetlock 1998). When environmental limits are weak, cognitive biases still are likely to be influential because beliefs provide comfortable anchors for decision making and act as expressions of social identities and personal idiosyncrasies that shape the definition of national interests (Hermann 1976; Holsti 1976; Walker 1983; Greenstein 1987; Wendt 1992; Goldgeier 1997). In short, beliefs are likely to be autonomous and matter even when states are not vulnerable.
The "unipolar moment" following the end of the cold war, when the international constraints of the bipolar cold war system disintegrated and left the United States less vulnerable with more freedom of action (Mastanduno 1997), provides an opportunity to test competing hypotheses from these different theoretical perspectives. By examining the operational codes of American chief executives during this time period, we can investigate the beliefs of U.S. elites to see whether they had a significant impact on U.S. foreign policy decisions in post-cold war conflicts. To assess the evidence on behalf of these rival claims, our analysis proceeds in two stages.
In stage 1, we ask whether the operational codes articulated by the two U.S. presidents are relatively stable over time and reflect the condition of low vulnerability with respect to their contents. If they remain stable in the face of different constraints and cues, then they are relatively autonomous and do not merely mirror changes in the international context. In stage 2, we ask whether U.S. behavior in foreign policy conflicts is consistent with the leaders' operational codes after allowing for stimulus effects from the opponents' behavior and realist effects from different power and inter-
Walker et al. / PRESIDENTIAL OPERA TIONAL CODES 613
est asymmetries. If beliefs remain stable over time, indicate low vulnerability, and are consistent with behavior, then these results will support a psychological explanation of foreign policy decisions for cases in which a more context-oriented explanation would discount autonomous beliefs as both unlikely and unimportant.
STAGE 1: ARE BELIEFS AUTONOMOUS?
We use the following research questions that traditionally have guided operational code analysis to identify a leader's beliefs and corresponding diagnostic, choice, and shift propensities (George 1969, 1979; see also Leites 1951, 1953, 1964). First, what are the leader's philosophical beliefs about the dynamics of world politics? Is the leader's image of the political universe a diagnosis marked by cooperation or conflict? What are the prospects for the realization of fundamental political values? What is the predictability of others, the degree of control over historical development, and the role of chance? Second, what are the leader's instrumental beliefs that indicate choice and shift propensities in the management of conflict? What is the leader's general approach to strategy and tactics and the utility of different means? How does the leader calculate, control, and manage the risks and timing of political action?
We look for evidence of these beliefs in the social construction of reality in speeches by the two presidents and recognize that their contents balance beliefs held by each leader with the expectations and norms of others in the political process. We make no attempt to differentiate these sources, and we assume that the contents of a speech are the product of the leader's own intellectual processes and the social processes of seeking advice from others. From these two assumptions, we draw the inference that the public operational code articulated by the chief executive is, in effect, the administration's operational code.
PROCEDURES
To assess the operational codes of political leaders, we use the Verbs in Context System (VICS) to score attribution patterns in random samples of public speeches by each leader. A previous application of this scoring system to the public speeches of President Jimmy Carter demonstrated reasonable support for the face, construct, and content validity of the operational code indexes (Walker, Schafer, and Young 1998; see also Rosati 1987; Sick 1986; Skidmore 1993). Although the assumptions behind the VICS vary somewhat by the index for each belief, the underlying substantive focus of operational code analysis is on the exercise of power in social relationships. Power here refers to the interplay of different types of control relations (e.g., authority, influence, reward, resistance, threat, punishment) between the self and others in the political universe (Dahl 1957; Cartwright 1959; French and Raven 1959; McClelland 1968, 1969; Baldwin 1971a, 1971b, 1978; Lukes 1986).
We employ these content analysis methods to examine the operational codes articulated by Bush and Clinton. We assess the stability of their operational codes and then describe the contents of their beliefs during their respective administrations. The
614    JOURNAL OF CONFLICT RESOLUTION
results are based on random samples from the Public Papers of the Presidents of the United States (Government Printing Office, 1989-96) of 12 speeches by Bush during his 4 years as president and 8 speeches by Clinton during the first 3 years of his presidency. Sampling frames of presidential speeches were developed and included only those speeches that contained at least 1,500 words to provide enough data per speech to construct the operational code indexes. Space does not permit an extensive discussion of the VICS coding procedures, but a brief description is appropriate here (see also Walker, Schafer, and Young 1998).
The recording unit is the verb-based attribution. Each verb is identified in context. The attribution is identified as a self-attribution (I, us, we, the United States, etc.) or an other-attribution (you, Israel, Hussein, etc.). The verb is identified as a transitive or an intransitive verb and as a positive/cooperation or a negative/conflict attribution. If the verb is transitive, then it is categorized as a word or deed and placed in the appropriate verb category—cooperation deeds in Reward, cooperation words in either Appeal/Support or Promise, conflict words in Oppose/Resist or Threaten, and conflict deeds in Punish. The verbs in these categories are assigned the following scale values: -~3 (Punish), -2 (Threaten), -1 (Oppose/Resist), +1 (Appeal/Support), +2 (Promise), +3 (Reward). Coders spent an average of 6 hours in training. They compared their coding judgments to precoded samples and then discussed errors and disagreements. This process continued until intercoder agreement reached 90%, at which time the coders were assigned speeches from the sample. Throughout the coding process, we randomly sampled sets of 20 attributions from each coder and conducted intercoder reliability tests. They generally averaged 90% agreement. Whenever agreement dropped noticeably below this level, we identified and corrected any problems.
INDEXES AND HYPOTHESES
The indexes for each element of a leader's operational code appear in Table 1, organized according to the diagnostic, choice, and shift propensities in a leader's operational code (George 1969,1979). The logic of these indexes rests on the assumption that the balance, intensity, dispersion, and relative frequency of verbs in the cooperation and conflict categories of the VICS scoring system indicate a speaker's diagnosis of the use of power by others in the political universe and the speaker's own propensities to exercise power in that same universe (Walker, Schafer, and Young 1998). The valences (+ or - signs) of each verb are used in some of the indexes, their numerical scale values are used for others, and both valences and scale values are employed to compute still others. From these assumptions, we can hypothesize that the resulting scores, which range between -1.0 and +1.0 for the balance indexes and between 0.0 and +1.0 for the intensity, frequency, and dispersion indexes, provide a summary of the speaker's view of the political universe and his or her orientation toward action in that universe.
For example, a hypothetical speaker with high negative scores for D-l and D-2 in Table 1 views the political universe as very hostile and is rather pessimistic about the prospects for realizing fundamental values. Low scores on D-3 and D-4 indicate uncertainty about the political future and low control over historical development,
TABLE 1
Indexes for Diagnostic, Choice, and Shift Propensities in a Leader's Operational Code
Element
Index
Interpretation
Diagnostic propensities
D-1. Nature of the political
universe (image of others)
D-2. Realization of political values (optimism/pessimism)
D-3. Political future (predictability of others' tactics)
D-4. Historical development (locus of control)
D-5. Role of chance
(absence of control)
Choice and shift propensities C-1. Approach to goals (direction of strategy)
C-2. Pursuit of goals
(intensity of tactics)
C-3. Utility of means (exercise of power)
a. Reward
b. Promise
c. Appeal/Support
d. Oppose/Resist
e. Threaten
f. Punish
Percentage positive other-attributions minus percentage negative other-attributions
Mean intensity of transitive other attributions divided by 3
1 minus Index of Qualitative Variation3 for other-attributions
Self-attributions divided by (self-attributions plus other-attributions)
1 minus (Political Future x Historical Development Index)
Percentage positive self-attributions minus percentage negative self-attributions
Mean intensity of transitive self-attributions divided by 3
Percentages for exercise of power categories
(a through f) a's Frequency divided by Total b's Frequency divided by Total c's Frequency divided by Total d's Frequency divided by Total e's Frequency divided by Total f's Frequency divided by Total
Ranges from +1.0 (friendly political universe) to -1.0 (hostile political universe)
Ranges from +1.0 (optimistic) to -1.0 (pessimistic)
Ranges from 1.0 (predictable) to 0.0 (uncertain)
Ranges from 1.0 (high self-control) to 0.0 (low self-control)
Ranges from 1.0 (high role of chance) to 0.0 (low role of chance)
Ranges from +1.0 (high cooperation) to -1.0 (high conflict)
Ranges from +1.0 (high cooperation) to -1.0 (high conflict)
Ranges from +1.0 (very frequent use) to 0.0 (infrequent use)
(continued)
TABLE 1 Continued		
Element	Index	Interpretation
S-l,  Risk orientation (predictability of tactics)	1 minus Index of Qualitative Variation for self-attributions'1	Ranges from 1.0 (risk acceptant) to 0.0 (risk averse)
S-2.  Timing of action (flexibility of tactics)	1 minus absolute value [percentage x minus percentage y self-attributions]	Ranges from 1.0 (high shift propensity) to 0.0 (low shift propensity)
a. Cooperation tactics versus conflict tactics b. Word tactics versus deed tactics	Where x - cooperation tactics and y = conflict tactics Where x = word tactics and y = deed tactics	
SOURCE: Adapted from Walker, Schafer, and Young (1998,178-82).
NOTE: All indexes vary between 0 and 1.0 except for D-1, D-2, C-1, and C-2, which vary between -1.0 and+1.0. D-2 and C-2 are divided by 3 to standardize the range, a. "The Index of Qualitative Variation is a ratio of the number of different pairs of observations in a distribution to the maximum possible number of different pairs for a distribution with the same N [number of cases] and the same number of variable classifications" (Watson and McGaw 1980, 88).
Walker et al. / PRESIDENTIAL OPERATIONAL CODES    617
making the index for the role of chance (D-5) very high. If the same speaker exhibits high positive scores for C-1 and C-2, then he or she has a choice propensity for a highly cooperative strategy accompanied by very cooperative tactics. The relative frequency of the verbs in the speaker's rhetoric would be concentrated more in the Reward and Promise categories for the exercise of power (C-3) with a relatively high risk orientation score (S-l), indicating acceptance of the risk of exploitation by others associated with a cooperative strategy and tactics. The control of this risk by shifting between cooperation and conflict (S-2a) would be low because of the concentration of verbs in the cooperation categories. However, if the dispersion of verbs is relatively high between the word and deed categories, leading to a high score for S-2b, then the speaker displays a propensity to manage this risk by shifting between promises and rewards.
Our analysis of the operational codes for Bush and Clinton uses their speeches as our units of analysis. This allows us to do trend analyses for each leader and make comparisons between them. We use analysis of variance (ANOVA) with year as the grouping variable and ordinary least squares regression with days in office as the independent variable to analyze the stability of each president's operational code. With these techniques, we assess the relative support for two hypotheses based on the rival claims about the autonomy of beliefs in presidential operational codes during the process of strategic adjustment:
Hypothesis 1: If internal cognitive mechanisms and personality structure matter, then beliefs are autonomous.
Hypothesis 2: If external context matters, then under the condition of low vulnerability, beliefs are not autonomous.
RESULTS
Neither technique detected consequential changes in the operational codes of the two leaders. ANOVA analyses of each leader's choice and shift propensities reached significance for only one choice propensity (C-3e Threaten,/? = .08, two-tailed), when Bush showed a higher propensity to use this foreign policy instrument in 1990 and 1991. The results of the regression analyses were similar, reaching significance for only one choice propensity (C-3b Promise); both Bush (p = .10) and Clinton (p = .07) showed lower propensities to use this foreign policy instrument later in their terms of office. The ANOVA and regression analyses of diagnostic propensities revealed that Bush's view of the political universe (D-l) was more negative in 1991 (p = .03) than in the other years. Both Bush (p = .05) and Clinton (p = .02) saw the predictability of the political universe (D-3) as higher later in their terms of office. For Bush, there also was a corresponding decrease (p = .02) in the role of chance (D-5).
In general, these trends are relatively minor and do not show the scope of change reported for Carter's operational code (Walker, Schafer, and Young 1998). Although the Bush and Clinton administrations faced important international challenges, it appears from these results that their respective operational codes remained relatively stable over time. These results support the hypothesis that the operational code beliefs
618    JOURNAL OF CONFLICT RESOLUTION
of the two leaders were autonomous during their respective administrations. Computing means across the data in the entire panel of presidential speeches, therefore, is appropriate to describe each administration's operational code and to make statistical comparisons.
The mean profiles of the two leaders in Table 2 reveal that Bush and Clinton have remarkably similar operational codes; the contents also are consistent with a judgment that the vulnerability of the United States was low during both administrations. Whereas the condition of high vulnerability is characterized by a threatening international environment and a pessimistic assessment of the prospects for achieving fundamental goals (Kupchan 1994, 86), both leaders view the political universe (D-l) as fairly friendly (Bush = +.42, Clinton = +.32) and are modestly optimistic (Bush = +.29, Clinton = +.31) about realizing political values (D-2). Reflecting the complex and rapidly changing strategic environment at the end of the cold war, both administrations also make modest assessments (Bush = .50, Clinton = .58) about the predictability of others (D-3), display caution (Bush = .53, Clinton = .45) about their ability to control historical development (D-4), and believe that the role of chance (D-5) is fairly high (Bush = .73, Clinton = .74).
For the strategy, tactics, and risk orientation indexes indicating choice and shift propensities, the Bush (+.78) and Clinton (+.73) administrations share a very cooperative approach to goals (C-1) and have similar scores in four of the six utility of means (C-3) categories (Appeal/Support, Oppose/Resist, Threaten, and Punish). They both have modestly cooperative (Bush = +.45, Clinton = +.57) tactical propensities (C-2), have the same score (.63) for general risk orientation (S-l), and are virtually identical (Bush = .23, Clinton = .22) in their propensities to shift between cooperation and conflict (S-2a). This combination of very cooperative strategic propensities, modestly cooperative tactical propensities, moderation in the calculation and control of risks, and a low propensity to shift between cooperation and conflict is consistent with beliefs that reflect a condition of low vulnerability. By contrast, the condition of high vulnerability is associated with beliefs that specify more extreme, erratic, and risky behavior (Kupchan 1994, 84-86).
In spite of these similarities, the two presidential operational codes do exhibit some notable differences. Bush clearly relies more on each of the two word categories (Appeal/Support, .36; Promise, .33) than the deed category (Reward, .20), and he is as likely to choose conflict deeds (Punish, .06) as conflict words (Oppose/Resist, .03; Threaten, .03). The distributions among the six transitive verb categories for Clinton's speeches show a propensity to choose Reward (.43) over either of the other two cooperative behavior categories (Appeal/Support, .30; Promise, .16); however, he is almost equally likely to choose words (Appeal/Support, Promise) as to choose deeds (Reward). In the three conflict categories, he is just about equally likely to choose any one of the three (Oppose/Resist, .04; Threaten, .03; Punish, .04), although Clinton is more likely to choose words (Oppose/Resist, Threaten) than deeds (Punish).
The indexes for three choice and shift propensities showed statistically significant differences (p < .05) between Clinton and Bush: Clinton relies more on Rewards (C-3a), Bush uses more Promises (C-3b), and Bush tends to be less flexible in his use of words and deeds (S-2b). These differences lead us to characterize the contents of
Walker et al. /PRESIDENTIAL OPERATIONAL CODES 619
TABLE 2
Operational Codes of Presidents Bush and Clinton for the Foreign Policy Domain
	Bush	Clinton
Index	(n = 12)	(n = 8)
Diagnostic propensities		
D-l. Nature of the political universe (image of others)	+.42	+.32
D-2. Realization of political values (optimism/pessimism)	+.29	+.31
D-3. Political future (predictability of others' tactics)	.50	.58
D-4. Historical development (locus of control)	.53	.45
D-5. Role of chance (absence of control)	.73	.74
Choice propensities		
C-l. Approach to goals (direction of strategy)	+.78	+.73
C-2. Pursuit of goals (intensity of tactics)	+.45	+.57
C-3. Utility of means (exercise of power)		
a. Reward	.20*	.43*
b. Promise	.33*	.16*
c. Appeal/Support	.36	.30
d. Oppose/Resist	.03	.04
e. Threaten	.03	.03
f. Punish	.06	.04
Shift propensities		
S-l. Risk orientation (predictability of tactics)	.63	.63
S-2. Timing of action		
a. Flexibility of cooperation/conflict tactics	.23	.22
b. Flexibility of word/deed tactics	.50*	.74*
*Significantly different (p < .05, two-tailed).
Bush's operational code as less cooperative and less flexible in its choice and shift propensities than the contents of the operational code articulated by Clinton. We infer from these differences that if beliefs are consistent with behavior, then the Bush administration's behavior will be relatively less cooperative and less flexible than the Clinton administration's behavior in foreign policy conflict situations.
STAGE 2: IS BEHAVIOR CONSISTENT WITH BELIEFS? PROCEDURES
We use event data in an effort to see whether aspects of each presidential operational code match aspects of U.S. behavior in conflicts with Panama, Haiti, Iraq, and Bosnia. The dates for data collection in each conflict include January 4, 1989 (at the beginning of the Bush administration) through January 4,1990 (when Noriega surrendered) for the U.S.-Panama dyad; January 9, 1993 (at the beginning of the Clinton administration) through October 16,1994 (when the U.N. Security Council lifted eco-
620    JOURNAL OF CONFLICT RESOLUTION
nomic sanctions) for the U.S.-Haiti dyad; August 3,1990 (when the United States condemned Iraq's invasion of Kuwait) through February 28,1991 (when the United States declared the liberation of Kuwait and the defeat of Iraq) for the U.S.-Iraq dyad; and January 22, 1993 (at the beginning of the Clinton administration) through November 22,1995 (when warring leaders signed an agreement to end fighting in Bosnia) for the U.S.-Bosnian Serb dyad. For each conflict, only words and deeds exchanged between the U.S. government and one target were retrieved. When the members of each dyad acted toward one another in concert with others, those words and deeds also were included in the data set.
There are some well-known problems with the use of event data from just one or two sources in depicting foreign policy; however, they are less relevant when the source is the main newspaper for the state under analysis. We collected events from The New York Times and The Washington Post in chronological order and coded them into the six VICS categories. The Times and the Post are more valid as sources of information about U.S. relations with other countries than about non-U.S. dyads. There also are coding and scaling problems with event data that interact with retrieval difficulties (Beer, Ringer, Sinclair, Healy, and Bourne 1992). Depending on the coding scheme employed and the sources of data, frequency counts of events merely dichotomized into the categories of conflict and cooperation by year differ significantly (Howell 1983; Vincent 1983; see also McClelland 1983).
The six-position scale used in our analysis from the VICS is consistent with the distinctions made in the World Event Interaction Survey's coding scheme between cooperation and conflict and between words and deeds (McClelland and Hoggard 1969; see also Goldstein and Freeman 1990, 1991; Goldstein 1992). Two different individuals coded each event, reaching an intercoder percentage agreement level of .94 for the data sets. Coders read the leads for each story pertaining to these conflicts and coded events. For each event, they first identified the actor (subject), next coded the valence of the word or deed (verb) as conflict or cooperation, and finally assigned the verb to one of the six transitive verb categories in the VICS system: Reward (+3), Promise (+2), Appeal/Support (+1), Oppose/Resist (-1), Threaten (-2), Punish (-3). Collectively, these categories constitute a six-position intensity scale for the use of positive and negative sanctions with a range from +3 to -3. The scale resembles the six-position scale designed by Rubin and Hill (1973) for use with World Event Interaction Survey (WEIS) data and employed by Leng (1993) to analyze patterns of interaction during acute international crises, weighting actions from most hostile to most cooperative.
INDEXES AND HYPOTHESES
To index behavior exchanged between states in a conflict situation as a series of moves, we first arranged the behaviors in chronological order and defined the elements of a move as the words and deeds by one state bounded by the other's immediately preceding and succeeding words and deeds. The direction and intensity of each move is the net sum of VICS cooperation and conflict scale values for the set of words and deeds that constitute its elements. The rival hypotheses regarding the relationship between moves and beliefs are as follows:
Walker et al. / PRESIDENTIAL OPERA TIONAL CODES    621
Hypothesis 3: According to a psychological theory of strategic adjustment, autonomous beliefs in each leader's operational code will influence the moves of his administration in response to the opponent's moves.
Hypothesis 4: According to a context-oriented theory of strategic adjustment, under the condition of low vulnerability, the distributions of power and interests (rather than the leader's beliefs) will influence the moves of an ascendant power in response to the opponent's moves.
To test these rival hypotheses, we now turn to a three-factor multivariate analysis of variance (MANOVA), President x Conflict Type x Opponent's Move, with each factor having two levels. The following analysis evaluates the impact on the intensity of U.S. moves of three different independent variables: presidential operational code (1 = less cooperative/flexible Bush, 2 = more cooperative/flexible Clinton), asymmetry of the conflict situation (1 = more asymmetrical Panama or Haiti, 2 = less asymmetrical Persian Gulf or Bosnia), and the direction of the opponent's move (+ = cooperation, - = conflict). This multivariate analysis of behavior allows us to explicitly consider the interaction among beliefs, context, and stimulus as well as their independent effects in determining U.S. behavior.
RESULTS
Mean scores and effects are presented in Table 3. Using p = .10 (two-tailed) as a threshold of significance for our hypotheses, only one main effect is significant: opponent's move (p = .08). This initial finding suggests that the United States is responding to the stimulus from the opponent. Two different interactions also are significant, however, making it appropriate to conduct post hoc analyses that may qualify this finding in important ways. The first significant interaction in Table 3 is President x Opponent's Move. The two-way analysis shows a dramatic difference in the responsiveness of each presidential administration to the opponent's move.
On one hand, the magnitude of the Bush administration's conflict response is virtually identical whether the opponent's move is positive or negative. On the other hand, the Clinton administration's conflict response is more intense in response to the opponent's conflict moves and less intense in response to the opponent's cooperation moves. The post hoc statistical analyses of the means in Table 3 confirm this pattern. The effect of the opponent's move on U.S. behavior under Bush's leadership is insignificant, F(l, 155) = 0.00, p = .973, but the effect under Clinton's leadership is significant, F(l, 155) = 4.76, p = .031.
The second statistically significant interaction in Table 3, President x Conflict Type x Opponent's Move, requires taking the analysis one step further and conducting post hoc analyses of all eight mean scores. This type of analysis is similar to the one just conducted except that we now look for response patterns by type of conflict as well. Four post hoc analyses are appropriate—two for each president—that examine the effect of the opponent's move (+ or -) on U.S. behavior in the two different types of conflict. Neither of the post hoc analyses of means in Table 3 for Bush is significant. The opponent's move does not influence the Bush administration's moves in either a more asymmetrical conflict, F{ 1,151) = 0.09, p = .763, or a less asymmetrical conflict,
622    JOURNAL OF CONFLICT RESOLUTION
F(l, 151) = 0.06, p = .804. By contrast, the Clinton administration was responsive to the opponent's moves in the more asymmetrical conflict, F(l, 151) = 14.17,/? < .001, but not in the less asymmetrical conflict, F(l, 151) = 0.38,/? = .538.
These results qualify the main effects relationship in Table 3 between stimulus and response rather significantly, suggesting that it holds only for some leaders and only under certain conditions. The pattern of moves by the United States under Bush's leadership is less cooperative and less flexible—choosing a course of action, sticking with it, and disregarding the opponent's machinations to alter the process no matter what. On the other hand, the Clinton administration is more cooperative and more flexible—responding more to both friendly and hostile moves by the opponent. These differences between administrations are sharper in more asymmetrical conflict situations.
It is possible that these findings are confounded by the cases selected for this study. Key differences in the "balance of threat" among the four cases rather than differences in presidential operational codes might account for variations in U.S. responses (Mastanduno 1997; Walt 1987). But we find this argument less compelling than our interpretation on both theoretical and empirical grounds. Theoretically, our analysis has been guided by propositions that, in fact, specify the relationship of U.S. responses to threats to national interests as well as to power asymmetries and the beliefs of leaders. Empirically, we do not find significant differences in the mean level of hostility from U.S. opponents, F(3,155) = 0.10,/? = .96. Bush faced essentially the same level of hostile moves from Iraq (-0.74) as Clinton did from Bosnian Serbs (-0.79), and the differences between the Panama (-1.08) and Haiti (-0.79) hostility levels also are statistically insignificant. We do not have data that address the level of hostility between the dyad members prior to the beginning of each administration. Thus, we cannot test the historical hypothesis that our results are due to a more hostile record of previous U.S. relations with the governments of Panama and Iraq than with the Haitian military regime and the Bosnian Serbs.
CONCLUSION
Neither the leader's beliefs and propensities for action nor environmental constraints and incentives account by themselves for the pattern of moves taken by the United States in the four post-cold war conflicts. However, the autonomous beliefs of leaders do matter in the causal analysis of foreign policy decisions, even when the context indicates that the vulnerability of the state is low. When the stakes are lowest and the balance of power is most favorable, autonomous and idiosyncratic differences in leadership are indispensable in accounting for behavior (Greenstein 1987). The pattern of U.S. moves in managing the Panama, Haiti, Bosnia, and Persian Gulf conflicts indicates that autonomous beliefs in conjunction with power asymmetries account for underreactions and overresponses by the United States to the stimuli from the opponents' behavior. In the Panama and Haiti conflicts in which they were least vulnerable,
Walker et al. /PRESIDENTIAL OPERATIONAL CODES 623
TABLE 3
U.S. Moves in a Three-Factor Multivariate Analysis of Variance Design (President X Conflict Type x Opponent's Move)
		Main and Interaction Effects (N - 159)	
Source		F(l, 150)	p Value (two-tailed)
President		0.00	.480
Conflict Type		0.22	.322
Opponent's Move		3.09	.082
President x Conflict Type		0.10	.747
President x Opponent's Move		4.51	.035
Conflict Type X Opponent's Move		1.61	.206
President X Conflict Type x Opponent's Move		3.51	.063
		Post hoc Analysis of Two-Way Interaction	
		(mean scores)	
Stimulus		Bush	Clinton
Opponent's Positive Move		-4.05	-2.56
Opponent's Negative Move		-4.12	-5.68
		Post hoc Analysis of Three-Way	
		Interaction (mean scores)	
		Bush	Clinton
	More	Less	More Less
Stimulus	Asymmetry Asymmetry		Asymmetry Asymmetry
Opponent's Positive Move	-5.00	-3.89	0.40 -3.11
Opponent's Negative Move	-3.56	-4.24	-10.07 -4.21
a less cooperative, inflexible president and a more cooperative, flexible president exercised power quite differently but in ways consistent with the stable choice and shift propensities in their respective operational codes.
The conflicts facing U.S. foreign policy managers during the post-cold war era are likely to include more situations in which balance of power constraints and vital national interest incentives are less likely to limit the choices of these leaders. So long as the vulnerability of the United States remains low, elites are relatively free from external constraints under these circumstances to respond to such cues in a way that is consistent with autonomous beliefs. This condition leads us to assess the potential importance of presidential leadership in post-cold war conflicts as relatively high. Presidential operational codes may be decisive in how much blood and treasure the last superpower expends in managing foreign policy conflicts at the close of the 20th century.
624    JOURNAL OF CONFLICT RESOLUTION
REFERENCES
Baldwin, David. 1971a. The power of positive sanctions. World Politics 24:19-38.
-. 1971b. Thinking about threats. Journal of Conflict Resolution 15:71-78.
-. 1978. Power and social exchange. American Political Science Review 72:1229-42.
Beer, Francis, Jeffrey Ringer, Grant Sinclair, Alice Healy, and Lyle Bourne, Jr. 1992. Ranking cooperation
and conflict events. International Interactions 17:321-48. Buckley, Kevin. 1991. Panama: The whole story. New York: Simon & Schuster. Cartwright, David. 1959. Studies in social power. Ann Arbor: University of Michigan Press. Dahl, Robert. 1957. The concept of power. Behavioral Science 2:201-15.
Donnelly, Thomas, Margaret Roth, and Caleb Baker. 1991. Operation Just Cause: The storming of Panama.
Lexington, MA: Lexington Books. Drew, Elizabeth. 1994. On the edge: The Clinton presidency. New York: Simon & Schuster. Flanagan, Edward. 1993. Battle for Panama: Inside operation Just Cause. Washington, DC: Brassey's. French, John, and Bert Raven. 1959. The bases of social power. In Studies in social power, edited by Dorwin
Cartwright, 150-67. Ann Arbor: University of Michigan Press. George, Alexander. 1969. The operational code: A neglected approach to the study of political leaders and
decision making. International Studies Quarterly 13:190-222. -•. 1979. The causal nexus between cognitive beliefs and decision-making behavior: The operational
code belief system. In Psychological models in international politics, edited by Lawrence Falkowski,
95-121. Boulder, CO: Westview. Goldgeier, James. 1997. Psychology and security. Security Studies 6.T37-66.
Goldstein, Joshua. 1992. A conflict-cooperation scale for WEIS event data. Journal of Conflict Resolution 36:369-85.
Goldstein, Joshua, and John Freeman. 1990. Three-way street: Strategic reciprocity in world politics. Chicago: University of Chicago Press.
-. 1991. U.S.-Soviet-Chinese relations: Routine, reciprocity, or rational expectations? American
Political Science Review 85:17-35.
Government Printing Office. 1989-96. Public papers of the presidents of the United States. Washington, DC: Government Printing Office.
Greenstein, Fred. 1987. Personality and politics. Princeton, NJ: Princeton University Press.
Hermann, Margaret. 1976. When leader personality will affect foreign policy. In In search of global patterns, edited by James Rosenau, 326-33. New York: Free Press.
Herrmann, Richard. 1988. The empirical challenge of the cognitive revolution. International Studies Quarterly 32:175-203.
Holsti, Ole. 1976. Foreign policy viewed cognitively. In The structure of decision, edited by Robert Axelrod,
18-54. Princeton, NJ: Princeton University Press. Howell, Llewellyn. 1983. A comparative study of the WEIS and COPDAB data sets. International Studies
Quarterly 27:149-59.
Hybel, Alex. 1993. Power over rationality: The Bush administration and the Gulf crisis. Albany: State University of New York Press.
Jervis, Robert. 1976. Perception andmisperception in international relations. Princeton, NJ: Princeton University Press.
Kupchan, Charles. 1994. The vulnerability of empire. Ithaca, NY: Cornell University Press. Leites, Nathan. 1951. The operational code of the Politburo. New York: McGraw-Hill.
-. 1953. A study of Bolshevism. New York: Free Press.
-. 1964. Kremlin moods. Santa Monica, CA: RAND.
Leng, Russell. 1993. Interstate crisis behavior, 1815-1980: Realism versus reciprocity. New York: Cambridge University Press. Lukes, Steven. 1986. Power. New York: New York University Press.
Mastanduno, Michael. 1997. Preserving the unipolar moment: Realist theories and U.S. grand strategy after the cold war. International Security 21:49-88.
Walker et al. / PRESIDENTIAL OPERA TIONAL CODES 625
McClelland, Charles. 1968. Access to Berlin: The quantity and variety of events. In Quantitative international politics, edited by J. David Singer, 159-86. New York: Free Press.
--. 1969. Theory and the international system. New York: Macmillan.
-. 1983. Let the user beware. International Studies Quarterly 27:169-78.
McClelland, Charles, and Gary Hoggard. 1969. Conflict patterns in the interactions among nations. In International politics and foreign policy, edited by James Rosenau, 711-24. New York: Free Press.
Mervin, David. 1996. George Bush and the guardianship presidency. New York: St. Martin's.
Rosati, Jerel. 1987. The Carter administration's quest for global community: Beliefs and their impact on behavior. Columbia: University of South Carolina Press.
Rubin, T, and G. Hill. 1973. Experiments in the scaling and weighting of international events data. Arlington, VA: Consolidated Analysis Centers.
Sick, Gary. 1986. All fall down: America's tragic encounter with Iran. New York: Penguin.
Skidmore, David. 1993. Carter and the future of foreign policy reform. Political Science Quarterly 108:699-729.
Snyder, Jack. 1991. Myths of empire. Ithaca, NY: Cornell University Press.
Taber, Charles. 1992. POLI: An expert system model of U.S. foreign policy belief systems. American Political Science Review 86:888-904.
-. 1998. The interpretation of foreign policy events: A cognitive process theory. In Problem representation in foreign policy decision making, edited by Donald Sylvan and James Voss, 29-52. Cambridge, UK: Cambridge University Press.
Tetlock, Phillip. 1998. Social psychology and world politics. In Handbook of social psychology, 4th ed., edited by D. Gilbert, S. Fiske, and G. Lindsay, 868-912. New York: McGraw-Hill.
Vertzberger, Yaacov. 1990. The world in their minds. Stanford, CA: Stanford University Press.
-. 1998. Risk taking and decision making. Stanford, CA: Stanford University Press.
Vincent, Jack. 1983. WEIS v. COPDAB: Correspondence problems. International Studies Quarterly 27:160-68.
Walker, Stephen. 1983. The motivational foundations of political belief systems. International Studies Quarterly 21:177'-201.
Walker, Stephen, Mark Schäfer, and Michael Young. 1998. Systematic procedures for operational code analysis: Measuring and modeling Jimmy Carter's operational code. International Studies Quarterly 42:175-90.
Walt, Stephen. 1987. The origins of alliances. Ithaca, NY: Cornell University Press.
Watson, George, and Dickinson McGaw. 1980. Statistical inquiry. New York: John Wiley.
Wayne, Stephen. 1993. President Bush goes to war: A psychological interpretation from a distance. In The
political psychology of the Gulf War: Leaders, publics, and the process of conflict, edited by Stanley A.
Renshon, 29-48. Pittsburgh, PA: University of Pittsburgh Press. Wendt, Alex. 1992. Anarchy is what states make of it. International Organization 46:391-425. Woodward, Bob. 1991. The commanders. New York: Simon & Schuster.
Zimmermann, Tim. 1996. A chilly peace in Bosnia. In World politics, edited by Helen Purkitt, 99-101. Guilford, CT: Dushkin Publishing. (Reprinted from U.S. News & WorldReport [December4,1995], 26-28,30)
http ://w ww .j stor.org
LINKED CITATIONS
- Page 1 of 3 -
You have printed the following article:
Presidential Operational Codes and Foreign Policy Conflicts in the Post-Cold War World
Stephen G. Walker; Mark Schäfer; Michael D. Young
The Journal of Conflict Resolution, Vol. 43, No. 5. (Oct., 1999), pp. 610-625. Stable URL:
http://links.istor.org/sici?sici=0022-0027%28199910%2943%3A5%3C610%3APOCAFP%3E2.0.CO%3B2-I
This article references the following linked citations. If you are trying to access articles from an
off-campus location, you may be required to first logon via your library web site to access JSTOR. Please
visit your library's website or contact a librarian to learn about options for remote access to JSTOR.
References
The Power of Positive Sanctions
David A. Baldwin
World Politics, Vol. 24, No. 1. (Oct., 1971), pp. 19-38. Stable URL:
http://links.jstor.org/sici?sici=0043-8871%28197110%2924%3Al%3C19%3ATPOPS%3E2.0.CO%3B2-X
Thinking about Threats
David A. Baldwin
The Journal of Conflict Resolution, Vol. 15, No. 1. (Mar., 1971), pp. 71-78. Stable URL:
http://links.istor.org/sici?sici=0022-0027%28197103%2915%3Al%3C71%3ATAT%3E2.0.CO%3B2-F
Power and Social Exchange
David A. Baldwin
The American Political Science Review, Vol. 72, No. 4. (Dec, 1978), pp. 1229-1242. Stable URL:
http://links.jstor.org/sici?sici=0003-0554%28197812%2972%3A4%3C1229%3APASE%3E2.0.CO%3B2-Z
The "Operational Code": A Neglected Approach to the Study of Political Leaders and Decision-Making
Alexander L. George
International Studies Quarterly, Vol. 13, No. 2. (Tun., 1969), pp. 190-222. Stable URL:
http://links.istor.org/sici?sici=0020-8833%28196906%2913%3A2%3C190%3AT%22CANA%3E2.0.CO%3B2-6
http ://w ww .j stor.org
LINKED CITATIONS
- Page 2 of 3 -
A Conflict-Cooperation Scale for WEIS Events Data
Joshua S. Goldstein
The Journal of Conflict Resolution, Vol. 36, No. 2. (Jun., 1992), pp. 369-385. Stable URL:
http://links.istor.org/sici?sici=0022-0027%28199206%2936%3A2%3C369%3AACSFWE%3E2.0.CO%3B2-3
U.S.-Soviet-Chinese Relations: Routine, Reciprocity, or Rational Expectations?
Joshua S. Goldstein; John R. Freeman
The American Political Science Review, Vol. 85, No. 1. (Mar., 1991), pp. 17-35. Stable URL:
http://lirJcs.jstor.org/sici?sici=0003-0554%28199103%2985%3Al%3C17%3AURRROR%3E2.0.CO%3B2-X
The Empirical Challenge of the Cognitive Revolution: A Strategy for Drawing Inferences about Perceptions
Richard Herrmann
International Studies Quarterly, Vol. 32, No. 2. (Jun., 1988), pp. 175-203. Stable URL:
http://links.istor.org/sici?sici=0020-8833%28198806%2932%3A2%3C175%3ATECOTC%3E2.0.CO%3B2-F
A Comparative Study of the WEIS and COPDAB Data Sets
Llewellyn D. Howell
International Studies Quarterly, Vol. 27, No. 2. (Jun., 1983), pp. 149-159. Stable URL:
http://links.jstor.org/sici?sici=0020-8833%28198306%2927%3A2%3C149%3AACSOTW%3E2.0.CO%3B2-J
Preserving the Unipolar Moment: Realist Theories and U.S. Grand Strategy after the Cold War
Michael Mastanduno
International Security, Vol. 21, No. 4. (Spring, 1997), pp. 49-88. Stable URL:
http://links.jstor.org/sici?sici=0162-2889%28199721 %2921%3A4%3C49%3APTUMRT%3E2.0.CO%3B2-D
Carter and the Failure of Foreign Policy Reform
David Skidmore
Political Science Quarterly, Vol. 108, No. 4. (Winter, 1993-1994), pp. 699-729. Stable URL:
http://linksjstor.org/sici?sici=0032-3195%28199324%2F199424%29108%3A4%3C699%3ACATFOF%3E2.0.CO%3B2-Z
http ://w ww .j stor.org
LINKED CITATIONS
- Page 3 of 3 -
POLI: An Expert System Model of U.S. Foreign Policy Belief Systems
Charles S. Taber
The American Political Science Review, Vol. 86, No. 4. (Dec, 1992), pp. 888-904. Stable URL:
http://links.istor.org/sici?sici=0003-0554%28199212%2986%3A4%3C888%3APAESMO%3E2.0.CO%3B2-W
WEIS vs. COPDAB: Correspondence Problems
Jack E. Vincent
International Studies Quarterly, Vol. 27, No. 2. (Jun., 1983), pp. 161-168. Stable URL:
http://links.jstor.org/sici?sici=0020-8833%28198306%2927%3A2%3C161%3AWVCCP%3E2.0.CO%3B2-9
The Motivational Foundations of Political Belief Systems: A Re-Analysis of the Operational Code Construct
Stephen G. Walker
International Studies Quarterly, Vol. 27, No. 2. (Jun., 1983), pp. 179-202. Stable URL:
http://links.istor.org/sici?sici=0020-8833%28198306%2927%3A2%3C179%3ATMFOPB%3E2.0.CO%3B2-4
Systematic Procedures for Operational Code Analysis: Measuring and Modeling Jimmy Carter's Operational Code
Stephen G. Walker; Mark Schäfer; Michael D. Young
International Studies Quarterly, Vol. 42, No. 1. (Mar., 1998), pp. 175-189.
Stable URL:
http://links.jstor.org/sici?sici=0020-8833%28199803%2942%3Al%3C175%3ASPFOCA%3E2.0.CO%3B2-9
Anarchy is what States Make of it: The Social Construction of Power Politics
Alexander Wendt
International Organization, Vol. 46, No. 2. (Spring, 1992), pp. 391-425. Stable URL:
http://links.istor.org/sici?sici=0020-8183%28199221%2946%3A2%3C391%3AAIWSMO%3E2.0.CO%3B2-9