Defining and Operationalizing Context Through a Structural Political Geography for International Relations
Colin Flint Political Science Department Utah State University Logan UT 84322-0725
[email protected]
Raymond Dezzani Geography Department University of Idaho Moscow, Idaho 83844-3021
[email protected]
Summary: A structural understanding of the contextualized behavior of states is introduced and operationalized. Context is a central theme of the discipline of geography and identifies context specific, rather than universal, social behavior. Social behavior is both defined by and creates contexts in a constant recursive interaction. Context is defined through a geographic perspective on world-systems analysis and we focus on the behavior of states. States are central actors because, through territorial sovereignty, are able to define key social relations and economic flows. The idea of context is developed in a way that extends the key IR concepts of milieu and opportunity and willingness. The recursive interaction between agency and context is conceptualized in a relational way as maneuver, the process by which the aggregate behavior of elites define state-level choices and behaviors that are made by considering the contextual position relative to all other states in the capitalist world-economy. In turn, the decision by any one state changes the behavior of other states so that context and state-level decisions interact and are constantly in flux. The elements of context include the position of a state in the hierarchy of the capitalist world-economy as well as regional and local inter-state relations, some of which may display path dependency. The operationalization of maneuver requires an understanding of states as signaling and learning entities and a set of modeling techniques that identify 1) the degree of change within the system as a whole – or the degree of stability in the number and identity of states within particular positions in the hierarchy of the capitalist world-economy; 2) the maneuver of particular states – or which states change position (or not) within the hierarchy; and 3) the explanatory power of variables measuring political and economic inter-state relations in explaining the maneuver behavior of particular states.
Keywords: Context, Geography, Structure, World-Systems Analysis, Maneuver, Learning, Markov model, Functional Logistic Models.
1
In the United States the academic discipline of geography is largely misunderstood by other social scientists, a much different situation than in Europe where geography is an integral part of school and university curricula. This fact requires an explanation of geography’s main contribution to theorizing and analyzing world politics, and is also a vivid illustration of that contribution: Simply, context matters. In this case, whether a social scientist is situated in Europe or the United States is a good indicator of the depth and breadth of their knowledge of geography. But what is context? Is it useful, or arguably necessary, in explaining international relations? And if context can be theorized can it be operationalized for systematic quantitative analysis of international relations? We believe the answers to the last two questions are an emphatic “yes.” To make our case we will define context and provide an operationalization of the concept. Just as with any other discipline, the core content, theoretical framing, and appropriate methodologies of academic geography are contested (Johnston and Sidaway, 2004; Peet, 1998). The discipline, and its numerous sub-disciplines, has not been immune from the spate of handbooks, encyclopedias, etc. that have appeared in recent years. Hence, there is no single definition of context, though the essence of the idea is quite simple. We offer a particular approach to context that is grounded in a structural approach to world politics, and reflects the way international scholars have themselves approached the idea of context. Though our approach is a particular one, we believe it resonates with much empirical international relations research and provides an avenue for future inquiry.
2
Our argument requires a discussion of what is meant by context, and especially an understanding of what a structural definition of context entails. To make such definitions useful for international relations scholars we then show how context has been adopted within empirical international relations, what we believe are the shortcomings of the IR usage of context, and why a structural definition of context is a useful improvement. The term maneuver is then introduced as a concept that enables IR’s focus on agency and decision-making to be connected to our argument regarding the importance of context and structure. Finally, we show the type of modeling that enables a systematic analysis of maneuver, in a way that incorporates the choices and actions of states with a conceptualization of structure and agency that is non-deterministic. Simple Beginnings So what is context? We will begin with two basic concepts: Site and situation. Site identifies the physical attributes of a location, and situation “refers to the location of a place relative to other places and human activities” (Knox and Marston, 1998, pp. 33-34). The general concepts of site and situation have particular expression for world politics through Cohen’s definition of geopolitics “as the analysis of the interaction between, on the one hand, geographical settings and perspectives and, on the other, political processes” (Cohen, 2003, p. 12). In this definition the actors are states, and their attributes and capabilities are equivalent to the notion of site. Situation, for Cohen, is one of geographic setting, which for some critics is too limited and creates a sense of geographic determinism of continental versus maritime powers (Cohen, 2003). For example, the naval capabilities of India are
3
only clearly understood within its geographical position at the center of the Indian Ocean (Brewster, 2014). The over-emphasis upon physical geographic features in Cohen’s definition makes it inappropriate for international relations research that aims to uncover the causal social processes that drive political behavior. At the same time, denying the role of geographic situation in political decisions leads to an understanding of politics that is too abstract. For example, Poland’s decisions regarding alliance formation can only be understood through its situation in central Europe that have defined its relationships with Russia and Germany for centuries. Hence, physical geographic setting must be part, but only part, of a conceptualization of context. The inclusion of social processes in an understanding of situation originates from Tobler’s (1970, p.236) classic first law of geography: “Everything is related to everything else, but near things are more related than distant things.” For international relations scholars the primary forms of relationships are political alliances, enduring rivalries, and economic ties that have been successfully operationalized under the umbrella of the Correlates of War data collection project (Sarkees and Wayman, 2010). Though this project has been very successful in explaining the behavior of states through the application of a dyadic unit of analysis, the shortcomings of such a conceptualization of actors has been recognized by the operationalization of what we can call situation that reflect a state’s position within a network of relations (Flint et al., 2009; Maoz, 2006) or in a neighborhood of states (Gleditsch, 2002, following the pioneering work of O’Loughlin and Anselin, 1992).
4
The network approach to international relations creates an understanding that the behavior of any particular state is enabled and constrained by the aggregate actions of all other states. Tobler’s (1970) first law of geography, as reflected in studies of dyads and triads in networks (Maoz, 2006) or the physical contiguity of neighbors (Gleditsch, 2002), suggests that those “closer” to a state will have greater influence on its behavior. Closeness, as a matter of connectivity or physical distance, is just one way to operationalize situation. The discipline of geography once defined situation as closeness, often using an abstract sense of space – or an isotropic plane – to model the behavior of actors as rational decision makers (Johnston and Sidaway, 2004). This approach to the concept of situation closely reflects the axioms of contemporary international relations. However, geographers have largely dismissed such a conceptualization. The level of abstraction of both the nature of actors, as simplified and unitary rational entities, and of their situation, in a featureless isotropic plain, did not match developments in social theory that were adopted by the vast majority of social scientists (Simandan, 2016). The challenge to the axioms of what was known in geography as the ‘quantitative revolution’ led to a discipline wide adoption of social theories that were influential across the social sciences (Johnston and Sidaway, 2004). The initial influence was theoretical Marxism, and though its presence has waned in geography, it still remains important. For our argument, its contribution rests in the introduction of social structure in to human geography inquiry and, therefore, the conceptualization of context.
5
Adding Structure The discipline of geography moved from the abstract view of space that was crucial for the analytical assumptions of the ‘quantitative revolution’ to understandings of space that were grounded in social theories. Especially, Marxist frameworks, with their structural approach, played a crucial role in changing the paradigmatic focus of the discipline. This was particularly the case for the revival of political geography in the 1970’s and 1980’s. Two related structural theories dominated: A classic Marxist approach (Harvey, 1982; Cox, 1979) and, the focus of this chapter, a political geography perspective on Immanuel Wallerstein’s world-systems analysis (Wallerstein, 1979; Taylor, 1981a). World-systems analysis was a (arguably the) driving force for a revitalized political geography that was desperately seeking theoretical foundation and relevance with other social sciences (Flint, 2010). Peter Taylor (1981a, 1981b, 1982) was the key figure behind the formulation of a political geographic perspective on world-systems analysis. He defined a political geography informed by the writings of Immanuel Wallerstein and the Annales school. For Taylor (1982), world-systems analysis provoked discussion of two key geographic concepts. The first was scale, with an ontology that moved away from concentration upon the state and toward the historical social system, specifically the capitalist world-economy. The second was the geographic concern with areal differentiation and its connection to the core, periphery, and semiperiphery hierarchy of the capitalist world-economy (Terlouw, 1992). The territorialization of core and peripheral processes within state borders led to an understanding of states holding a position within the capitalist world-economy
6
depending on the prevalence of the two different types of processes within the borders of a state (Dezzani, 2002; Babones, 2009; Snyder and Kick, 1979; Terlouw, 1992; Quah, 1997). States are important institutions within the single capitalist world-economy because of the ability they have for partial control of flows across borders and their influence upon wages and other relations of production (Wallerstein, 1979). The scalar approach to world-systems identified states as key actors, but ones that operated within the imperatives of the world-economy (Taylor, 1981b). Hence, states can make decisions, but their choices are limited, and the range of choices is partially a function of position in the core-periphery hierarchy. The application of world-systems analysis introduces the potential for a structural definition of context (or situation) to be conceptualized as position in the broader historical social system. World-systems analysis also provided a sense of process to political geography. Specific political actions were situated within the dynamics of the capitalist world-economy, through the application of the concept of Kondratieff waves and cycles of investment (Arrighi, 1990, 2010; Taylor, 1985; Flint and Taylor, 2011). The Kondratieff dynamic was subsequently connected to cycles of hegemony that were based on economic strength relative to all states in the system (Arrighi, 1990; Boswell and Sweat, 1991; Flint and Taylor, 2011; Thompson, 1986). The consideration of economic and political cycles introduced a temporal aspect to context. The opportunities and constraints facing a state were partially defined by the particular moment in the Kondratieff and hegemonic cycles.
7
The crucial role of states in territorializing economic processes and engendering competition between states created a single political-economic logic that drove the decision-making of states (Chase-Dunn, 1989). The context of state actions was a space-time context of the dynamics of Kondratieff waves and position within the core-periphery hierarchy (Flint and Taylor, 2011). The range of possible actions available to a state depended upon their capabilities, as largely defined by their position in the capitalist world-economy, and the degree of economic growth and innovation or stagnation occurring at a particular time. However, such a conceptualization emphasized structural context and limited discussion of agency or choice (as an exception see Taylor, 1993). Agency must also be operationalized to make the idea of structural, or time-space, context useful for international relations scholars. A pathway to this goal has been open for a while, and is found within the critique of structural approaches that emerged in the social sciences sometime around the 1980s. As with other Marxist approaches, world-systems analysis created a new geographic understanding of context that emphasized that the cost-benefit rational decision making of actors should be reframed as a set of politicized behaviors that occur within a structure of power relations largely defined by the imperatives of capitalism (Peet, 1998, pp. 112-146). Over time other approaches to social theory were adopted that either challenged or ameliorated the early adoption of Marxist approaches. This trend, called the ‘cultural turn,’ meant that contemporary Marxist analyses in geography have been re-framed to become less deterministic (Cowen
8
and Smith, 2009; Gibson-Graham, 2006; Johnston and Sidaway, 2004; Mercille, 2008). The dilution of Marxism’s influence in human geography inquiry went handin-hand with challenges to structural approaches. In contemporary geography, especially understandings of power and context, the idea and role of structure became implicit rather than explicit. The unease with conceptualizing structures as causal entities existing prior to human behavior means that most geography scholars shy away from a structural approach. However, such concern fails to employ understandings of structure that incorporate the mutuality of actor and structure formation (Martin and Dennis, 2010). Reframing political structures as geopolitical constructs stresses the “construction and maintenance of settings or contexts with broad geographical scope and long-lasting implications for social behavior” (Flint, 2016). The ability of actors to act is still recognized and incorporated in to analysis, but there is recognition that there are significant constraints upon behavior. The constraints are expressed as rules and norms that are the expressions of a social structure that is, in turn, the outcome of and medium for social behavior (Martin and Dennis, 2010). Within the context of structural imperatives, the behavior of actors is dynamic and transformative: the capabilities and goals of actors change within new structurally defined contexts that are themselves the product of the aggregate behavior of actors. Structural contexts are the venues for agency, and the relationship between structure and agent is a recursive one. Hence, the nature of structural contexts and the range of choices held by actors is somewhat fluid. The goal is to avoid the
9
determinism of reified structures while also acknowledging the stability of structural imperatives and, therefore the limits they set on agency. In fact, “it is precisely the ability to understand how enduring institutional patterns are generated and maintained through situated interactions that is the greatest challenge to the ‘sociological imagination’, a challenge which conventional sociology has conspicuously failed to meet” (Martin and Dennis, 2010, p. 16). Contemporary quantitative political science has been much less inclined to accept, or even recognize, this challenge. As a result it has emphasized agency over structure, and the contextual settings produced by the mutuality of agency and structure have been under-theorized. International Relations Antecedents Despite contemporary prioritization of agency, the idea of context has historical foundations in empirical IR. Harold and Margaret Sprout first defined the role of context in explaining the behavior of states in the 1950s and 1960s. The Sprouts called for empirical analysis that modeled the interaction between ‘environmental’ factors and interstate politics. For the Sprouts, the environment meant a complex combination of social, geographical, and historical factors, which they called the milieu (Sprout and Sprout, 1965). Political geographers would readily identify the Sprouts’ milieu as a particular definition of context. Moreover, the Sprouts defined world politics as a limited range of options facing states. Hence, some political actions are more probable than others given the environmental situation in which they are made (termed “environmental possibilism”). Decisionmaking by state elites was still seen as rational, but a limited range of choices was
10
identified within the perceived opportunities and constraints of a state’s milieu (called “cognitive behavioralism”) (Sprout and Sprout, 1957, 1965). For a fuller discussion see Radil et al. (2013). The Sprout’s idea of milieu was refined to allow for the operationalization of the concept in empirical analysis. The resulting adoption of context by IR scholars came through the opportunity and willingness framework (e.g., Starr 1978). This was a reductionist idea of milieu or context that, in practice, focused upon the relative geographic location of a state (Most and Starr, 1989; Siverson and Starr, 1991). Geographers have lamented the way that little of the complexity of the Sprouts’ concern for context has been actually reflected in the empirical analyses since their contribution (Radil, et al., 2013). Starr (1978) created his concept of opportunity by building upon the Sprouts’ (1957) idea of environmental possibilism, to define it as the possibility of interaction between two political entities. However, the multiplicity of interacting factors within the Sprouts’ milieu was limited so that opportunity was narrowly understood as territorial adjacency. The result was an operationalization of context, as opportunity, in which “states that shared a boundary were seen as most likely to interact and the Sproutian environment, the total spectrum of social, geographical, and historical factors, had become reduced to various measures of shared boundaries between territorially defined states” (Radil et al, 2013, p. 1470). Willingness, the agency component of Starr’s (1978) framework, was based upon the Sprouts’ (1965) notion of cognitive behavioralism. It was an operationalization of the tangible outcomes of state-decision making, such
11
as the formation of alliance partnerships or the maintenance of militarized rivalries (Starr, 1978; Most and Starr, 1989; Siverson and Starr, 1991). The opportunity and willingness framework made an important contribution to empirical analyses of the diffusion of war. It was based on the idea that the decisions of states can only be fully understood through the inclusion in the analysis of the opportunities and constraints provided by specific contextual settings. However, there was also a legitimate concern with how context could be operationalized within the parameters of the Correlates of War project to enable systemic and replicable analysis that contributed to the whole field of empirical IR. The outcome was an operationalization of context that focused upon physical contiguity with neighboring states, especially immediate neighbors. The richer notion of the Sprout’s milieu was lost. The concept of willingness has also been reduced to political calculations, rather than the intersection of economic and political needs and goals that is the core of a political economy approach such as world-systems analysis. However, the fact that the notion of milieu has existed in IR literature, and kept alive, to some degree, by the idea of opportunity means the door for introducing other notions of context is slightly ajar, if one is optimistic. Operationalizing Context Context is the core concept for geographers, but is only useful for empirical analysis when it is operationalized. Furthermore, to enable inter-disciplinary conversation such operationalization must speak to the analytical agendas of IR. Such conversation would be further enabled if the IR concepts of milieu and the related concepts of opportunity and willingness are advanced theoretically to match
12
developments across the social sciences. A structural definition of context, we believe, is a social theoretic development of the concept of milieu, and one that can be operationalized for systematic quantitative analysis of state behavior. When it comes to the operationalization of context for quantitative analysis three particular concepts have proven to be useful: Spatial dependency, spatial heterogeneity, and relative position. Spatial dependency is akin to the lack of independence in time-series data (Anselin, 1988). It recognizes that the value of a variable in one spatial unit is often partially dependent upon the value in neighboring spatial units. To incorporate spatial dependency in to empirical IR analysis means that the standard dyadic approach must be modified to consider the geographic proximity between the two states (O’Loughlin and Anselin, 1992). This has been adopted to a considerable degree in empirical IR (see Gleditsch, 2002), but if that is the extent of the conceptualization of context then there is a risk of a geographic determinism, or “spatial fetishism” (Smith, 1981) that denies the way mutual construction of social relations and geographic spaces (Massey, 2005). Spatial heterogeneity is the recognition that behavior is not uniform or universal across time and space (Anselin, 1988). Instead, the situation of an actor, especially the temporal-spatial context of the space-time matrix, says a lot about the political possibilities at hand. The empirical manifestation of spatial heterogeneity is a set of spatial regimes, in which the statistical significance and even the positive or negative direction of the relationship varies across sub-sets of the total number of cases that are identified spatially, usually as regions (Flint, 2002). For example,
13
Chi and Flint (2013) identified regional patterns, or spatial regimes, in the causal factors creating territorial conflicts. The concept of relative position has become prominent with the increasing adoption of social network analysis across the social sciences, including international relations (Maoz, 2006; Hoff and Ward, 2004) and political geography (Radil, et al., 2013), and collaborations between the two (Flint, et al., 2009). A social network is an aggregation of social ties that are formed by the decisions of the actors in the network and, in turn, form the structural nature of the total network. Measures of connectivity and centrality of an actor in the network define their relative position that can be broadly understood as being more or less central, peripheral, or even disconnected from the network (Wasserman and Faust, 1994). To define the context of actors in international relations requires recognizing all three of these specific understandings of context. Such a combination of concepts, if theoretically informed, would bring contextual analysis in empirical IR back to the multi-faceted nature of milieu originally promoted by the Sprouts. Dependency is the role of neighborhood or physical contiguity found in Siverson and Starr’s (1991) definition of opportunity and more recently developed by Gleditsch (2002). Heterogeneity may be incorporated in two ways: The geographical sense of regional structures of behavior (Buzan and Waever, 2003), and also a structural sense of similarity in actors capabilities and needs depending upon their relative position within the whole structure (Dezzani, 2002; Barbones, 2009; Terlouw, 1992). Rather than seeing this as a network structure, world-systems analysis would suggest that a state’s position within the core-semiperiphery-periphery hierarchy of the
14
capitalist world-economy creates heterogeneity in capabilities and behavior. Certainly the United States, Japan, and the countries of western Europe play a different role in world politics compared to very poor and politically weaker states. Defining the heterogeneity of states through their position in the world-economy is a way to formally recognize the inequality of states and operationalize such difference in empirical analysis. Agency, or Siverson and Starr’s (1991) opportunity, must also be incorporated to avoid an analysis of world politics that combines geographic and structural determinism. The structure of the capitalist world-economy creates a set of imperatives, demands, expectations, and limits that constrain what is a viable politics for any given set of actors within the temporal dynamics of the system and their position within the hierarchy of the capitalist world-economy. We advance the concept of maneuver to advance the essence of the opportunity and willingness framework in a way that uses a world-systems analysis of the capitalist worldeconomy to define contextual settings that recognize the historical, social, economic, political, and geographic elements of context, and give a true echo to the Sprout’s idea of milieu (Sprout and Sprout, 1965). From Milieu to Position and from Willingness to Maneuver The structural approach of world-systems analysis provides a particular way to consider context or milieu, and that is position within the capitalist worldeconomy. The primary element of position is the core-periphery hierarchy and the particular moment within the temporal dynamics of economic (Kondratieff) and hegemonic cycles: This is the space-time matrix discussed earlier. However, it is
15
quite possible for other elements of context, such as regional interactions and the role of immediate neighbors, to be added to investigate the full complexity of situatedness as conceived by the Sprouts. Such a complex understanding of position goes some way towards alleviating concerns of structural determinism by adding historical cultural ties, long-standing antagonisms, territorial disputes, etc. Position is not a life-sentence. Though the three-tier hierarchy of the capitalist world-economy is a necessary and permanent feature of the system there is room for movement for individual states. In other words, though the proportion of states that may be classified within each of the categories of core, periphery, and semi-periphery remains fixed, actual states can move up or down the hierarchy. For example, the US entered the system as a semi-peripheral, perhaps even peripheral, state to become core. Portugal was once core but has flirted with the semiperiphery. The categories of the world system hierarchy may also change within a stochastic context such that analysis may reveal significant cohorts as subsets or aggregations of the more traditional tripartite configuration (see Dezzani, 2001; Nemeth and Smith, 1985; Quah, 1997; Smith and White, 1991; Snyder and Kick, 1979). There is room for movement, but not much (Babones, 2009; Dezzani, 2002). This brings us to the concept of maneuver – or the willingness or ability of a state to change its position given the opportunity or milieu defined by its position within the capitalist world-economy and regional settings. Maneuver is an idea born of a necessary economic instrumentalism fostered by the idea of the topological state (Taylor 1989, pp.164-165). The idea was more fully developed in terms of specific actors and their possible decisions in Taylor
16
(1993, pp. 184-186). The nascent notions of maneuver formulated by Taylor (1989, 1993) are tied to the potential responses of capitalists within a state economy that are integrated into a larger interdependent world-economy. The logic of ceaseless capital accumulation in the world-economy means that decision makers and controllers of capital have two basic strategic options: 1) raise prices or, 2) lower wages (Taylor 1993, p. 185). These options represent the incipient form of strategic decisions that initialized the maneuver idea for a topological model of the state. The decision options for capitalists in a state-mediated world-economy are limited by a plethora of state decisions that govern firm behavior, trade and labor relations among other processes of interaction and production. This chain of decision-making is path-dependent and also conditioned on existing structural forms and relations at the times of decision consideration and implementation. As such, the decisions that constitute maneuver are a complex linkage of both sequential and parallel events and policies conditioned on prevailing norms of interaction (i.e., laws, treaties, agreements, tradition, culture, behavioral norms, etc.), that may govern the success or failure of policy intended from the decisions. The outcome of these decisions may then have an effect on the relative structural position of the firm or the state that may either enhance, harm or impart no change. Taylor’s idea of maneuver modeled the state as a collection of firms responding to profit-generating activities both within a particular state and across the border in other states in the quest for opportunity (e.g., topological state). Hence, production and movement of commodities that are both necessary and beneficial to states and firms will be influenced by capitalists’ decisions that may alter the structural
17
position of the firm or the state in the world-economy and result in a “change of position” of the state with respect to others (Taylor 1993, p. 184-186). The growth of new mechanisms of production, such as cross-border commodity chains, reflects maneuver responses to the evolution of state-level decisions in an increasingly globalized world-economy (Brewer, 2011; Grinberg, 2016; Robinson, 2002, 2004). However, capitalists are not the only decision-makers within the modern state that need specific decision instrumentalities owing to the topological nature of the modern state integrated into the world-economy. State bureaucracies and most internal political, social and economic stakeholders must also function within a constrained decision framework. As a result, maneuver may be seen to represent the aggregate decisions that potentially result in state-level actions that can influence the relative position of the state with respect to other states in the hierarchical world-economy. The State as the Proper Unit of Analysis. The state is the key actor within the world-economy. That is, the state is the focus of political and economic forces from within the state itself as well as forces engineering and driving change for state responses in the global and regional arenas. This perspective does not preclude other actors both within the state (e.g., political groups, national groups, economic entities, etc.), trans-state (e.g., transnational and multi-national corporations) and non-territorial entities. However, the idea of maneuver is peculiar primarily to the state as it is the state that most effectively benefits from the outcomes of maneuver. Corporations, nationalistic
18
groups and other non-territorial actors may engage in maneuver but the essential payoff of maneuver is garnered by states. This issue raises the question of pertinence of agency in the world-systems context. Many actors may “play the game,” so to speak, but states encapsulate the greatest advantage because only states exhibit territorial capacity coupled with population, resources, the ability to legitimately project power, create policy, control borders, engage with other states and focus institutional capacity for the greatest human welfare. As such, state behavior is more meaningful for the greatest proportion of the human ecumene with respect to any other actor or unit of analysis. The particular role of the state in the capitalist world-economy is a function of its ability to territorialize social relations. Territorial sovereignty enables political and economic elites to enact policies regarding prices and wages as part of the strategy of maneuver, making the state the instrument and unit of action within the capitalist world-economy. The hierarchical structure of the world-economy suggests a relatively conservative framework that is quite stable over time. This stability might be perceived to be at odds with realist notions of anarchy and relative autonomy. The idea of maneuver permits a variety of economic and political solutions to the problem of geographical variation and replaces relative autonomy with locally flexible alternatives to autarky and anarchy. Using this approach, a variety of political and economic solutions are permissible under the conditions of maneuver as the state functions as intermediary with the world-economy. At the center of the maneuver idea is the state, which serves as a basis of interaction between the
19
interior scale of territoriality and the larger external scale of the global economy and intrastate interaction. The state provides the basis of legitimation and political interaction to incorporate non-territorial actors that require territorialized capital, infrastructure and services. Maneuver functions precisely because there is a plethora of states that constitute a range of possible solutions all striving within a common structural framework (Agnew and Corbridge, 1995; Flint and Taylor 2011). We outline a structural approach with process specification that will enhance the ability of the world-systems perspective to evaluate and inform studies in international relations. The Giddens’ (1984) criticism of the world-systems approach is founded on a perceived lack of rigor in ontological specification that we hope, in part, to amend. We acknowledge the political, military and hegemonic roles of states to act in the global arena but we also recognize the preeminence of capitalism as an organizing principle and conservative force in preserving hierarchy (Taylor, 1996; Wallerstein, 1974, 1981). This does not imply that capitalism need be the only mechanism of economic exchange and redistribution, but it is the dominant mode of economic interaction. In addition, we actively seek to incorporate all of the potential and historically-realized functions of the state in the maneuver process for the purpose of mobility and approach the problem from a synthetic rather than reductionist lens. This statement implies that we also support the contributions of Frank and Gills (1993) and Abu-Lughod (1989) that identify a longer period of development for the “modern” world system than is originally specified by Wallerstein (1974).
20
Signaling and “Learning” The essential precursors whereby states engage in maneuver are signaling/perception and ‘learning.’ There are many forms of signaling to the state or state governing agencies such as diplomatic protest, formal treaties and compliance, relative performance in the international economy as reflected in statistical assessment, threats of civil conflict and declarations of war, to name just a few. The manner in which we apply the term ‘signaling’ is similar to the idea of signaling theory in evolutionary biology and political science (e.g., Banks, 1991; McCarty and Meirowitz, 2007, pp.214-235), but is closer to the approach of dynamic Bayesian games with imperfect information. We consider states to be grouprational semantic systems capable of information processing and acting on that information to maximize economic or power advantage and, thus, position in the global hierarchy. State perception functions can be centralized and institutionalized in the state as with formal intelligence gathering agencies and also with information-collecting organizations or through non-state and informal information gathering (Powell, 1999). The signaling concept is specified in Modelski’s (1987) mechanism of interstate competition over hegemonic cycles. Modelski posits a Parsonian learning framework that is used to describe the ability of a state to perceive and respond to changes in secular cycles, hegemonic cycles and immediate crises or threats. The performance of the state responding to these information inputs is a function of economic and coercion capacity as well as the presence of coping structures and the willingness to engage solutions. We can think of information flowing into a state
21
through many channels (see Figure 1) and consider a sequence of “channels” for this information corresponding to state-level structures for information processing and decision-making. While the traditional world-systems approach does not reinforce or provide a framework for the analysis of such strategic phenomena, the structural approach we posit does significantly account for a variety of learning processes. Learning is not necessarily just Parsonian (i.e., differentiation and phase-cycle/process), but may also be mimetic, (e.g., by states attempting to emulate core or semi-periphery successes or forced with the imposition of Rostow-type policies that reflect inherent core-behavior mimesis), and also reflect a variety of positional logics, both geographical/geopolitical and hierarchical. Our structural framework models the state as a semantic entity that uses contextual and group perceptions to balance specific bargaining postures into the decision-making process. The state, the key unit of analysis in empirical IR, is situated within contexts defined by the structural imperatives of the capitalist world-economy and conceptualized as an actor with the ability to ‘learn’ and act within its contextualized constraints and opportunities. The maneuver approach suggests that states can choose to selectively use external and internal inputs from a variety of actors territorialized within the state and across state boundaries and represent their interests. This process will vary over many states and will also impact the final maneuver decisions and thus, maneuver success or failure. In other words, the process of maneuver is relational as the decision-making process of one state is partially defined by the decisionmaking of all other states, and the action of one state change the structural context
22
for all other state decision-making. In this way we can combine the structural imperatives of the capitalist world-economy with consideration of both the internal state and external regional contextual setting to provide a multi-faceted and multiscalar notion of the Sprouts’ milieu; one that shows that “opportunity” is a limited set of actions that can be thought of as state maneuver. States, as containers of power and legitimacy, are used by state elites for their own purpose but also by other groups or actors seeking to engage their own interests (Flint and Taylor 2011, p.142). The policy instruments, interactions and relationships formed to accomplish goals and engender conflict all convolve to influence the maneuver position of a state. As such, a variety of factors from path dependent process effects, (i.e., Markov dependence), to internal power politics and resource factor endowments can be manipulated in a variety of ways that influence decisions, policies and actions which result in a state’s maneuver posture. Not only endogenized factors but also global and regional processes influence the maneuver position of a state. Perceptions of hegemonic stature, and capital accumulation and concentration cycles (i.e., Kondratieff cycles) may all provide input into the maneuver decision process. State learning is ongoing as possibilities for particular action are engaged – this Parsonian learning, as described by Modelski (1987) provides opportunities for states and other elite actors and corporations to attempt to “understand” the mechanism of the world-economy and to generate alternatives that introduce enhanced efficiency further improving the outcomes of maneuver. Evolutionary learning that occurs over decades and centuries may also inform the decision basis
23
for maneuver, or, alternatively, create a dialectic that may result in a reformulation of the gauge parameters in the decision matrix that will also influence state-level maneuver (Flint and Taylor, 2011; Modelski, 1990, 1996). The decision for a core state to contend for hegemony is the result of positional logic inferences reflective of either current hierarchical position or a perceived trend in capabilities for hegemonic contention (Modelski, 1987, 1996). The outcome will depend on many factors some of which have computable likelihoods. It is postulated that structural determination of these likelihoods is feasible for covariates that are effective measures of the state-level decision processes or their outcomes. As the hierarchical structure is Markov dependent between any two sequential time periods, computation of likelihoods are determined by probability counting rules and the maximum likelihood principle (Dezzani, 2002; Sayrs, 1993; Wilkinson and Tsirel, 2005). Aggregate mobility fluxes of countries across levels of the hierarchy are defined as transitions and can be assessed using probabilities or likelihood of movement. Transition probabilities are used to measure state mobility as determined from hierarchical classification over a sequence of time periods and assessed for persistence or transition: this procedure provides a complete description of change and state-level movement in the system. Transition probabilities, used to analyze state-level mobility, can also be used as the basis for structural explanations of the transition process. “Explanation” as expressed by covariate relationships between a measure of mobility, expressed as a probability, with process and decision measures occurs through the logistic expansion of the transition matrix and coupling mobility
24
with structural covariates that are associated or correlated with maneuver (Dezzani, 2012). Structural Analysis of the Hierarchical World Economy The basis of a system is to capture the individual changes as part of the entire system mechanism. This argument is best expressed in a defense of world history by William McNeill: “…Many historians, indeed, refuse to interest themselves in world history because they feel it involves so much vagueness and generality that testable statements about the past simply slip away. Such a view is quite wrong. World history depends on sources in exactly the same way as national or any other scale of history depends on sources; and the effort to corroborate or refute a particular hypothesis is the same, whether the hypothesis in question pertains to the entire world, to a civilization, to a nation or to some little village in the Pyrenees. What constitutes adequate evidence is always problematical. One-toone correspondence between a historian’s statements and what “really” happened is unattainable; and if it were attainable would be undesirable, since it would simply preserve the buzzing, blooming confusion of everyday experience that impinges differently on every human being, hour by hour and minute by minute. A total recording of an individual consciousness is impossible, as novelists’ experiments of the early twentieth century surely suffice to prove. What is needed – always – is a suitable shorthand: a system of terms that classifies experience into meaningful, usable, and satisfying patterns. Only so can we understand the world around us. Only by leaving things out, and lumping varying individual instances together into categories and classes of things, can we hope to navigate successfully amidst the infinitely various actual encounters humans have with one another and with the world around (McNeill, 1986, pp. 82-83).”
With regard to empirical IR, the challenge is to operationalize the behavior of specific actors, in our case states, within broader systemic patterns and show how the behavior of an individual actor maintains and/or alters the patterns. In other
25
words, behavior and context are recursively related and any distinction between “opportunity” and “willingness” is false. To deny and avoid such a false distinction between agent, structure, and context modeling of state behavior must operationalize the fact that changes in one part of the world-economy can induce changes in another because they are connected. One way to accomplish this is to model the integrating characteristic for each part of the world-economy. Call the characteristic Ri which is a measure of the maneuverability or transition change in in each part/zone (i.e., classification for i = 1, 2, … n). Then the entire system can be described as a series of differential equations in terms of every other zonal/group component as:
∂R1/∂t = f1(R1, R2, . . . , Rn) ∂R2/∂t = f2(R1, R2, . . . , Rn) ... ∂Rn/∂t = fn(R1, R2, . . . , Rn)
That is the partial time rate of change of net change is a unique zonal function of change within zone as well as change across all other zones. This is necessary since the world-system and the world-economy represent a single interacting mechanism such that changes that occur in one subsystem induce commensurate changes in other subsystems. For the sake of argument, assume only three possible states of the world-economy as explicated in the original world-systems framework (Wallerstein, 1974). The hierarchical state classifications for the n countries under
26
analysis are core, semi-periphery and periphery with indices c, s, and p, respectively. Generally speaking, the grouping of states using prevailing commonalities that reflect similar processes while permitting contexts to vary, can provide for many more potential groups of states with statistically-similar characteristics even if processes may vary somewhat. Thus, this analysis is not limited to the traditional tripartite division of groups of states in the hierarchical world-economy. As such, a structural hierarchical matrix of groups reflecting similar processes creating the hierarchical structure is directly reflective of the dynamic mechanism of change. This is a particular operationalization of heterogeneity discussed earlier. Theoretically, every state could form a unique “group” reflective of its degenerate situation in the world-economy; this is, in fact the degenerate situation and states do exhibit similar characteristics across a range of criteria that are reflective of statistically significant groupings of territorial entities of the world-economy (Arrighi and Drangel, 1986; Dezzani, 2001, 2002; Fingleton, 199X; Quah, 200x, etc.). For example, assume a valid statistical classification procedure isolates k groups of states constituting the hierarchy of the world-economy such that k ≤ n, where n is the total number of states in the study and n ≤ N with N representing the total number of possible state in the world-economy at the time of measurement. Then, the possible transition matrix may be represented as: 𝐹11 𝐹12 . . . 𝐹1𝑣 𝐹21 𝐹22 . . . 𝐹2𝑣 ... ( 𝐹𝑣1 𝐹𝑣2 . . . 𝐹𝑣𝑐 )
27
Where the rows exhibit the frequencies Fi*, such that i = 1,…v, of states occupying the group i at time t, and the columns F*j, such that j = 1,…v, of state occupying group j at time t+k, where v is some interval of time specified by the study parameters. That is, k is the time interval for which mobility or persistence is to be measured. Using this specification, any number of classes, not just the original Wallerstein tripartite hierarchy, can be employed to examine the hierarchical structure of the world-economy. The marginal totals, either row or column, represent the total states occupying the grouping at time t (for rows) or t+k (for columns). Basic counts of the dynamics of state movement or persistence in the worldeconomy can be assessed directly from the transition matrix. For a complete description of movement indices in a hierarchical system over time see Boudon (1973). The proportion of mobile states among hierarchical groupings can be computed as: 𝑃𝑟𝑜𝑝(𝑚𝑜𝑏𝑖𝑙𝑒) =
𝑡𝑜𝑡𝑎𝑙 𝑚𝑜𝑏𝑖𝑙𝑖𝑡𝑦 − 𝑠𝑡𝑟𝑢𝑐𝑡𝑢𝑟𝑎𝑙 𝑚𝑜𝑏𝑖𝑙𝑖𝑡𝑦 𝑡𝑜𝑡𝑎𝑙 𝑚𝑜𝑏𝑖𝑙𝑖𝑡𝑦
where total mobility = [n - ∑ Fii ] and structural mobility =[ n - ∑ min(Fi*, F*i) ] for all i. One can think of the structural mobility as the minimum number of mobile states across groupings as determined by the marginals for the time interval k. Similarly, we can call the maximum mobility as the maximum number of mobile states across hierarchical groupings for the time interval k. These counting rules operate consistently for any dimension of 2 groups or greater and serve as basic descriptors of the change of states across the v groupings of the hierarchical world-economy.
28
These descriptors serve to permit comparisons of the mobility of states between time periods and across different intervals to assess the time sensitivity of statelevel mobility in the world-economy. For the sake of brevity and simplicity we will employ the conceit of a tripartite hierarchy, (e.g, Wallersteinian groupings – core, semi-periphery and periphery), of the world-economy but note that this framework can be generalized to any number of statistically-significant classes of states producing the hierarchical structure and that the dimension of the hierarchy can change at any time (Dezzani, 2002). A few simple inferences of the transition matrices will demonstrate the veracity of this generality to the reader. For the specific case of three structural elements constituting the hierarchy of the world-economy we have: 𝐹𝑐𝑐 𝐹𝑐𝑠 𝐹𝑐𝑝 𝐹𝑠𝑐 𝐹𝑠𝑠 𝐹𝑠𝑝 … (𝐹𝑝𝑐 𝐹𝑝𝑠 𝐹𝑝𝑝) The maximum likelihood estimates for transition probabilities are estimated as the quotient of the observed state frequencies with the total number of countries traversing or persisting in the particular state for the time periods considered (Fij, where i is the row index for hierarchical position at the succeeding time t and j is the column index for hierarchical position at the preceding time t-k for some interval k):
29
𝐹𝑐𝑐 ∑∀𝑖 𝐹𝑐𝑖 𝐹𝑠𝑐 𝑃𝑠𝑐 = ∑∀𝑖 𝐹𝑠𝑖 𝐹𝑝𝑐 𝑃𝑝𝑐 = ∑∀𝑖 𝐹𝑝𝑖 𝑃𝑐𝑐 =
𝐹𝑐𝑠 ∑∀𝑖 𝐹𝑐𝑖 𝐹𝑠𝑠 𝑃𝑠𝑠 = ∑∀𝑖 𝐹𝑠𝑖 𝐹𝑝𝑠 𝑃𝑝𝑠 = ∑∀𝑖 𝐹𝑝𝑖 𝑃𝑐𝑠 =
𝐹𝑐𝑝 ∑∀𝑖 𝐹𝑐𝑖 𝐹𝑠𝑝 𝑃𝑠𝑝 = ∑∀𝑖 𝐹𝑠𝑖 𝐹𝑝𝑝 𝑃𝑝𝑝 = ∑∀𝑖 𝐹𝑝𝑖 𝑃𝑐𝑝 =
for all I = 1, 2, …,n countries. The maximum likelihood computation produces the final Markov transition probability matrix: 𝑃𝑐𝑐 Mt-k,t-i = ( 𝑃𝑠𝑐 𝑃𝑝𝑐
𝑃𝑐𝑠 𝑃𝑠𝑠 𝑃𝑝𝑠
𝑃𝑐𝑝 𝑃𝑠𝑝 ) where k > i 𝑃𝑝𝑝
The vector of occupation properties for core, semi-periphery and periphery after the time of the estimation interval (t-k, t-i such that k > i) is: P(c,s,p,t) = (Pc,t Ps,t Pp,t) = P(t)Mt-k,t-i such that t > t-i where Pcc, Pss, Ppp are probabilities of persistence within the world-system hierarchy reflecting Pii , where i ={c, s, p}, while Pij such that i,j ={c, s, p} but i≠j. Then, a logistic formulation is derived to capture structural relationships for Markov transition probabilities for persistence (i.e, i,i), and mobility (i.e., i,j): Pii (β, x(i,i)) = exp(β’x(i,i))/[1 + exp(β’x(i,i))] (persistence) Pij(χ, x(i,j)) = exp(χ’x(i,j))/[1 + exp(χ’x(i,j))] (mobility) where β and χ are the corresponding parameter vectors and x is the corresponding structural/covariate design matrix. The binary response likelihood function for any particular world-system Markov transition configuration is: L = ∏q Pcc Fcc,q ( 1–Pcc)Fcs,q + Fcp,q PssFss,q (1–Pss)Fsc,q + Fsp,q PppFpp,q (1–Ppp)Fpc,q + Fps,q
30
Where Fij,q are the number/frequency of transitions of each type observed in the qth country. When the expectations of the logistic functions are embedded in the likelihood then the log-likelihood becomes ln(L) = L0 + L1 where L0 is the likelihood for persistence and L1 is the likelihood for mobility. For brevity, we will only illustrate functions for persistence: L0 = ∑𝑛𝑞=1{𝐹(𝑐𝑐, 𝑞) 𝜷′ 𝒙(𝑞) − (𝐹(𝑐𝑐, 𝑞) + [𝐹(𝑐𝑠, 𝑞) + 𝐹(𝑐𝑝, 𝑞)]) ln[1 + exp(𝜷′ 𝒙(𝑞))]} Maximize L0 + L1 by Bayesian hierarchical Newton-Raphson iteration and estimate the parameter vectors β and χ provided the design matrix x is of full rank and the logistic for persistence and mobility can be estimated separately as there are no terms involving both parameter sets simultaneously: ∂L0/∂βs = ∑q=1,n x(qs)[Fcc,q – (Fcs,q + Fcp,q) Pcc(β’ x(q))] ∂2L0/∂βs∂βr = − ∑𝑛𝑞=1 x(qs)x(qr)(Fcc,q + [Fcs,q + Fcp,q]) Pcc (β’ x(q))[1 –Pcc(β’ x(q))] Hence, there exists a feasible covariate solution set for the explanation of worldsystem transition. The result of this functional explication is the derivation of the logistic function. The logistic can be used as a framework of stochastic “explanation” for persistence and/or mobility of countries across the states of the world hierarchy. The logistic function is derived from the log-likelihood for the core persistence expression and is delineated as: ln [Pcc / (1 – Pcc)] = Z which implies [ Pcc / (1 – Pcc)] = eZ So that Pcc = F(Z) = F(β’x) = F(β0 + ∑m=1,k βmxm) = 1 + e-(β’x) Validation of maneuver decisions can be evaluated by assessing the significance of parameters associated with independent variable selected to capture maneuver
31
behavior that would “explain” the variation in the transition probabilities. This analytical framework should provide a statistically feasible assessment of maneuver behavior that might be expected to account for movement of countries in the hierarchy; hence, provide a stochastic explanation of maneuver as it results in transition probabilities (Dezzani, 2012). This approach can be reasonably extended to explicit hierarchy connectin evaluations using the representation of Markov random graphs employing spatial connection matrices in the logistic (Pattison and Wasserman 1999). Up to now we have emphasized how the actions of one state are partially defined by its context, and how the aggregate of state actions maintain and alter structural context. In other words, we have operationalized a dynamic sense of milieu (Sprout and Sprout, 1957, 1965) or opportunity (Starr, 1978). To satisfy IR’s concentration on state level decision-making we must also operationalize behavior or ‘willingness’ (Starr, 1978). A mixed logistic framework can be used to generalize the parameter specification and evaluate country-level behaviors with respect to persistence and/or movement in the hierarchy (Mcfadden and Train, 2000; Revelt and Train 1998, Tsutakawa 1988, Wang and Puterman, 1998). Choice probabilities dictating the transition rate and probabilities between time periods are induced by both structural variables which can change as secular (slow) rates over longer time periods or, alternatively, represent agent decisions responding to short-term and local contexts or, some combination of these. Thus, individual country decisions may vary significantly within a hierarchical class and this variance can provide information on unique maneuver behaviors. Variables can be combinations of
32
economic, political, conflict and policy measures while trade and capital flows may be modeled as modified Hirschmann indices measuring state-level global contribution (see Dezzani 2001 for a description). These variables may be found in the collection of datasets in the Correlates of War project, and complementary sources. Transition probabilities would require variables that capture dynamic change in measured levels of economic and political processes or conflict/social behaviors within and across state boundaries. Through the implementation of creative and theoretically-derived covariates, a variety of hypotheses evaluating persistence or transition may be evaluated. While, the logistic framework can provide information on group average transition behaviors between time periods, the mixed logistic framework will permit the evaluation of individual countries. However, the data restrictions are more severe for the parameter estimation of the mixed logistic regression function (ref here..).
Pcc,i = Fi(Zi) = Fi(β’x) = Fi(β0,i + ∑m=1,k βm,i xm,i) = 1 + e-(β’x)i
for countries I – 1, 2, …,n and parameters m = 1, 2, …,k where k is the number of covariates employed to explain the transition of persistence probability.. Markov transition matrices provide a complete description of state-level and aggregate mobility but do not explain maneuver mechanisms or processes. The logistic expansion framework, derived from the transition probability partitions, provides an analytical basis for “explaining” maneuver mechanisms through the
33
inclusion of structural and contextual covariates as well as components of statelevel change. This analytical framework permits both specific and general maneuver hypotheses to be evaluated. In other words, we can model the degree of persistence and change in contextual settings as well as the individual behavior of states, and the recursive interaction between agency and context. The Markov methods provide system description for the hierarchical arrangement of the world-economy for a time interval and the logit decomposition of the transition probabilities, as persistence or mobility behaviors, coupled with functional logistic covariate analysis provide the statistical explanation of specific and general maneuver hypotheses. Potential Problems and Issues with the Analytical Approach As with all quantitative methodologies there is a necessity for generalization and the need to make hypotheses tractable to the data available. Models, by definition, are simplifications of reality. However, by choosing units of analysis that are meaningful at a scale for which reliable data is available, such methods can provide needed information with a resolution that is useful for both theory validation and policy analysis. These two goals are well within the capacities of the proposed method. However, covariates in the logistic model must be shown to reflect measures of maneuver-based processes, policies and effects otherwise, the “explanatory” logistic model will not capture the salient features of the maneuver behavior. Nevertheless, if the independent variables constituting the covariate effects on transition can be shown to be effective measures of maneuver behavior, then specific maneuver hypotheses can be evaluated.
34
Data limitations are the major issue with most global, long-term studies. Most reliable data at the state level begins in the 1950s and extends to the present. However, variable coverage can be intermittent. As such, many studies limit the number of possible cases to be included (see Dezzani, 2001, 2002). Structural criticisms which argue that the study can only be state-centered are not necessarily valid as other “actors” for which comparable data is available may be included in the structural and explanatory components of the analysis. However, the data types across actors are assumed to be complimentary. Hence, the major limitations of this approach may be data quality and availability. This is also the case, however, with most quantitative approaches to complex social science studies.
Conclusion That the behavior of states, and other social actors, is context specific is axiomatic to the discipline of geography and the ways IR scholars adopt spatial analysis. However, the idea of context has been under-theorized. The structural approach to context, using world-systems analysis, enables an extension of the Sprouts (1957, 1965) idea of milieu and Starr’s (1978) opportunity and willingness framework, in a way that better reflects the political-economy logic of the capitalist world-economy within which states act. Though our approach is structural it should be noted that the structuralism of today is very different from the structuralism of environmental determinism that ushered in modern geography. Critiques of structuralism demand a place for agency (Martin and Dennis, 2010). However, the trend in social science has emphasized agency at the expense of structure, such as the epistemology of
35
rational choice. Poststructuralist argue that they account for structure but alter the arguments that focus on experiential units. The idea of maneuver allows for an exploration of the possibilities of state action within the structural constraints of the capitalist world-economy. Such an approach can build upon the ideas of milieu and opportunity that have been used in IR in a number of ways. One way is by adding economic structure to the dominant emphasis upon political calculations. The other way is to expand what is meant by milieu or opportunity through the idea of contextual setting that identifies and analyzes multiple processes operating simultaneously at multiple scales (neighborhood and regional). The result is a contextualization of state actions, both economic and political, that show the ability of states to learn and act within a limited set of structurally imposed constraints. In this way, the geographic tradition of identifying the contextual setting of a social actor is blended with IR’s traditional focus on decision-making, and a world-systems analysis tradition of structural constants, or the longue durée (Braudel, 1984). We can summarize this approach by proposing that change within the world system hierarchy is measured using the mobility of countries, as political-economic territorial entities with the capacity for growth and change conferred by such entities, across the levels of hierarchy (Agnew and Corbridge, 1995, Tilly 1984). This movement may be rapid or slow depending on a variety of factors primarily dependent upon maneuver decisions that encompass economic actions, policy actions, diplomacy and conflict. These actions may have both internal and external
36
effects on country’s interaction with other actors, both territorial and nonterritorial. Maneuver thus forms a basis for this state-level change within the hierarchy of the world-economy. Maneuver also provides the rational basis of structural inference of transition probabilities that describe structural change but provide no explanation for change. The logistic expansion of transition probabilities using process-based outcomes and effects to “measure” the relative change tied to the changing structural position provides a basis of explanation. The analytical approach consisting of the logistic expansion of the Markov transition probability matrix can provide structural models (spatially explicit), to examine specific factors in either mobility or persistence behaviors. In this way, maneuver is useful tool with a strong analytical component for the analysis of state-level change in a hierarchical world-economy. Complexities of hegemony and non-territorial actors may also be included as the imagination of the analyst permits. The inclusion of context, space, geography, distance, and other key concepts within the discipline of geography are a welcome sign of inter-disciplinary with an established history. However, to theoretically advance the collaboration these concepts need to be as rigorously theorized as the relations that have become axiomatic within IR, such as alliances, rivalry, etc. We hope this essay is a step in that direction through a theorization and operationalization of context that, though embedded within a political-economy logic, is of use to the puzzles addressed by IR scholars.
37
Citations Abu-Lughod, J. L. 1989. Before European Hegemony: The World System A.D. 1251350 (New York: Oxford University Press). Agnew, J. and Corbridge, S. (1995). Mastering Space: Hegemony, Territory, and International Political Economy (London and New York: Routledge). Anselin, L. (1988). Spatial Econometrics: Methods and Models (Dordrecht, Netherlands: Kluwer Academic Publishers). Arrighi, G. (1990). “The Three Hegemonies of Historical Capitalism.” Review XIII, 3: 365-408. Arrighi, G. (2010). The Long Twentieth Century (New York: Verso). Arrighi, G. and Drangel, J. (1986). “The Stratification of the World-Economy: An Exploration of the Semiperipheral Zone.” Review X, 1: 9-74. Babones, S. (2009). The International Structure of Income (Saarbrucken, Germany: VDM Verlag Dr. Muller). Banks, J. S. (1991) Signaling Games in Political Science. (Chur, Switzerland: Harwood Academic Publishers). Boswell, T. and Sweat, M. (1991). “Hegemony, Long Waves, and Major Wars: A Time Series Analysis of Systemic Dynamics, 1496-1967.” International Studies Quarterly 35(2): 123-149. Boudon, R. (1973) Mathematical Structures of Social Mobility. (Amsterdam: JosseyBass/Elsevier). Braudel, F. (1984). The Perspective on the World (London: Collins). Brewer, B. D. (2011). “Global Commodity Chains & World Income Inequalities: The Missing Link Of Inequality & The ‘Upgrading’ Paradox.” Journal of World Systems Research XVII (2): 308-327.
38
Brewster, D. (2014). India’s Ocean: The Story of India’s Bid for Regional Leadership (New York: Routledge). Buzan, B. and Waever, O. (2003). Regions and Powers: The Structure of International Security (Cambridge: Cambridge University Press). Chase-Dunn, C. (1989). Global Formation (Cambridge, MA and Oxford, UK: Blackwell). Chi, S-H. and Flint, C. (2013). “Standing Different Grounds: The Spatial Heterogeneity of Territorial Disputes.” Geojournal 78(3): 553-573. Cohen, S.B. (2003). Geopolitics of the World System (Lanham, MD: Rowman and Littlefield). Cowen, D. and Smith, N. “After Geopolitics? From the Geopolitical Social to Geoeconomics.” Antipode 41(2009): 22-48. Cox, K. R. (1979). Location and Public Problems: A Political Geography of the Contemporary World (Chicago, IL: Maaroufa). Dezzani, R. J. (2001). “Classification Analysis of World Economic Regions.” Geographical Analysis 33:330-352. Dezzani, R. J. (2002). “Measuring Transition and Mobility in the Hierarchical WorldEconomy.” Journal of Regional Science 42:595-625. Dezzani, R. J. (2012). “Measuring transition and hierarchy of states within the worldsystems paradigm.” Pp 129-138 in: Babones, S. J. and Chase-Dunn, C. (editors) Routledge Handbook of World Systems Analysis (London, UK. Routledge). Fingleton, B. (1999) “Estimates of Time to Economic Convergence: An Analysis of Regions of the European Union.” International Regional Science Review 22 (1): 5-34. Flint, C. (2002). “The Theoretical and Methodological Utility of Space and Spatial Statistics: The Nazi Party in Geographical Context.” Historical Methods 35 (1): 32-42. Flint, C. (2010). “Geographic Perspectives on World-Systems Theory,” Pp 2828-2845 in R. Denemark (Ed.) International Studies Association Compendium (Malden, MA: Wiley-Blackwell).
39
Flint, C. (2016). Geopolitical Constructs: Mulberry Harbours, Operation Bolero, and the Making of the Militarized Trans-Atlantic (Lanham, MD: Rowman and Littlefield). Flint, C. and Taylor, P.J. (2011). Political Geography: World-economy, Nation-State, and Locality (sixth edition) (Harlow, Essex: Pearson Education). Flint, C., Diehl, P., Scheffran, J., Vasquez, J., and Chi, S-H. (2009). “Conceptualizing ConflictSpace: Toward a Geography of Relational Power and Embeddedness in the Analysis of Interstate Conflict.” Annals of the Association of American Geographers 99 (5): 827-835. Frank, A. G. and Gills, B. (editors). 1993. The World System Five Hundred Years Or Five Thousand? (London, UK. Routledge). Gibson-Graham, J.K. (2006). Postcapitalist Politics (Minneapolis, MN: University of Minnesota Press). Giddens, A. 1984. The Constitution of Society (Cambridge: Polity Press). Gleditsch, K. (2002). All International Politics is Local: The Diffusion of Conflict, Integration, and Democratization (Ann Arbor: University of Michigan Press). Grinberg, N. (2016). “Global Commodity Chains and the Production of Surplus-value on a Global Scale: Bringing Back the New International Division of Labor Theory.” Journal of World Systems Research, 22(1): 247-278. Harvey, D. (1982). The Limits to Capital (Oxford: Basil Blackwell). Hoff, P. and Ward, M. (2004.) “Modeling Dependencies in International Relations Networks.” Political Analysis 12: 160-175. Johnston, R. J. and Sidaway, J. (2004). Geography and Geographers: Anglo-American Human Geography Since 1945 (London: Hodder Arnold). Knox, P. and Marston, S. (1998). Human Geography: Places and Regions in Global Context (Upper Saddle River, NJ: Prentice Hall). Lake, D. A. 2009. Hierarchy in International Relations (Ithaca, NY. Cornell University Press).
40
Maoz, Z. (2006). “Network Polarization, Network Interdependence and International Conflict.” Journal of Peace Research 43: 391-411. Martin, P. J. and Dennis, A. (2010). “Introduction: The Opposition of Structure and Agency” Pp. 3-16 in P. J. Martin and A. Dennis (eds.) Human Agents and Social Structures (Manchester and New York: Manchester University Press). Massey, D. (2005). For Space (Thousand Oaks, CA: Sage Publications). McCarty, N. and Meirowitz, A. (2007). Political Game Theory (New York, NY: Cambridge University Press). McFadden, D. and K. Train (1997) “Mixed Multinomial Logit Models for Discrete Response.” working paper, Department of Economics, University of California, Berkeley. McNeill, W. (1986). Mythistory and Other Essays (Chicago: University of Chicago Press). Mercille, J. (2008). “The Radical Geopolitics of US Foreign Policy: Geopolitical and Geoeconomic Logics of Power.” Political Geography 27: 570-586. Modelski, G. (1987). Long Cycles in World Politics (Seattle, WA. University of Washington Press). Modelski, G. (1990). “Is World Politics Evolutionary Learning?” International Organization 44(1): 1-24. Modelski, G. (1996). “Evolutionary Paradigm for Global Politics.” International Studies Quarterly 40(3): 321-342. Most, B. and Starr, H. (1989). Inquiry, Logic and International Politics (Columbia, SC: University of South Carolina Press). Nemeth, R. and Smith, D. A. (1985). “International Trade and World-System Structure: A Multiple Network Analysis.” Review 8: 517-560. O’Loughlin, J. and Anselin, L. (1992). “Geography of International Cooperation and Conflict: Theory and Methods.” Pp. 11-38 in M.D. Ward (ed.) The New Geopolitics (Philadelphia: Gordon and Breach).
41
Pattison, P. and Wasserman, S. (1999) “Logit models and logistic regressions for social networks: II. Multivariate relations” British Journal of Mathematical and Statistical Psychology 52: pp 169– 193 Peet, R. (1998). Modern Geographical Thought (Oxford: Blackwell). Powell, R. (1999). In the Shadow of Power: States and Strategies in International Politics (Princeton, NJ: Princeton University Press). Quah, D. T. (1997). “Empirics for Growth and Distribution: Stratification, Polarization and Convergence Clubs.” Journal of Economic Growth 2: 27-59. Radil, S, Flint, C., and Chi, S-H. (2013). “A Relational Geography of War: ActorContext Interaction and the Spread of World War I.” Annals of the Association of American Geographers 103 (6): 1468-1484. Revelt, D., & Train, K. (1998). “Mixed logit with repeated choices: households' choices of appliance efficiency level.” Review of economics and statistics, 80(4): pp 647-657. Robinson, W. I. (2002). “Capitalist Globalization and the Transnationalization of the State” Pp. 210-229 in Rupert, M. and Smith, H. (editors) Historical Materialism and Globalization (New York, NY: Routledge). Robinson, W. I. (2004). A Theory of Global Capitalism: Production, Class, and State in a Transnational World (Baltimore, MD: Johns Hopkins University Press). Sarkees, M.R. and Wayman, F. (2010). Resort to War: 1816-2007 (Washington, DC: CQ Press). Sayrs, Lois W. (1993). “The Long Cycle in International Relations: A Markov Specification.” International Studies Quarterly 37(2): 215-237. Simandan, D. (2016). “Proximity, Subjectivity, and Space: Rethinking Distance in Human Geography.” Geoforum 75: 249-252. Siverson. R., and H. Starr. (1991). The Diffusion of War: A Study of Opportunity and Willingness (Ann Arbor, MI: University of Michigan Press).
42
Smith, D. A. and White, D. R. (1991).”Structure and Dynamics of the Global Economy: Network Analysis of International Trade 1965-80.” Social Forces 70: 857-893. Smith, N. (1981). “Degeneracy in Theory and Practice.” Progress in Human Geography 5: 111-118. Snyder, D. and Kick, E. (1979). “Structural Position in the World-System and Economic Growth, 1955-1970: A Multiple-Network Analysis of Transnational Interactions.” American Journal of Sociology 84: 1096-1126. Sprout, H. and Sprout, M. (1957). “Environmental Factors in the Study of International Politics.” Journal of Conflict Resolution 1: 309–328. Sprout, H., and Sprout, M. (1965). The Ecological Perspective on Human Affairs, with Special References to International Politics (Princeton, NJ: Princeton University Press). Starr, H. (1978). “'Opportunity' and 'Willingness' as Ordering Concepts in the Study of War.” International Interactions 4: 363–387. Taylor, P.J. (1981a). “Political Geography and the World-Economy.” Pp. 157-172 in A.D. Burnett and P.J. Taylor (eds.) Political Studies from Spatial Perspectives (New York: Wiley). Taylor, P. J. (1981b). “Geographical Scales Within the World-Economy Approach.” Review 5: 3–11. Taylor, P. J. (1982). “A Materialist Framework for Political Geography.” Transactions, Institute of British Geographer NS7: 15–34. Taylor, P.J. (1985). Political Geography: World-Economy, Nation-State, and Locality (Harlow, Essex: Longman Scientific and Technical). Taylor, P.J. (1993). Britain and the Cold War: 1945 as Geopolitical Transition (London: Continuum). Taylor, P. J. (1996) The Way the Modern World Works: World Hegemony to World Impasse (Chichester, UK: John Wiley & Sons, Ltd.).
43
Terlouw, C.P. (1992). The Regional Geography of the World-System (Utrecht: Rijksuniversiteit). Thompson, W. R. (1986). “Polarity, the Long Cycle and Global Power Warfare.” Journal of Conflict Resolution 30 (4): 587-615. Tilly, C. (1984) Big Structures, Large Processes, Huge Comparisons (New York: Russell Sage Foundation). Tobler, W. (1970) “A Computer Movie Simulating Urban Growth in the Detroit Region.” Economic Geography 46: 234-240. Tsutakawa, R. K. (1988) “Mixed Model for Analyzing Geographic Variability in Mortality Rates.” Journal of the American Statistical Association, 83 (401): 37-42. Wallerstein, I. (1974). The Modern World-System: Capitalist Agriculture and the Origins of the European World-Economy in the Sixteenth Century (New York: Academic Press). Wallerstein, I. (1979). The Capitalist World-Economy (Cambridge: Cambridge University Press). Wang, P. and Puterman, M.L. (1998) “Mixed Logistic Regression Models.” Journal of Agricultural, Biological, and Environmental Statistics 3 (2): 175-200. Wasserman, S. and Faust, K. (1994). Social Network Analysis: Methods and Applications (New York: Cambridge University Press). Wellman, B. (1997). “Structural Analysis: From Method and Metaphor to Theory and Substance.” Contemporary Studies in Sociology 15, 19-61. Wilkinson, D. and Tsirel, S. V. (2005). “Analysis of Power-Structure Fluctuations in the “Longue Duree” of the South Asian World System.” Structure and Dynamics 1(2). http://escholarship.org.uc/uc/item/1j3063fd
44
Figure 1 Causal information flows for the maneuver process
45