![]()
1 Foreign policy, domestic politics and international relations
A strategic-relational analysis
It is widely agreed that foreign policy occupies a critical, interstitial space in world politics, produced as it is at the porous interface of domestic politics and international relations. Four generations of foreign policy scholars, often from radically different standpoints, have converged around the idea that foreign policy is a ‘boundary’ activity, straddling disciplinary fields as well as the inside and outside of political actors, generally states. Thus, in 1966 Robert B. Farrell characterised the study of foreign policy as a ‘no man’s land’ at the intersection of Political Science and IR (Farrell 1966: vi). Similarly, in the early 1970s William Wallace argued that ‘the study of foreign policy is a boundary problem’ for both the foreign policy student and practitioner (Wallace 1971: 7). A decade later, James Rosenau famously used the metaphor of a ‘bridge’ to identify the location of foreign policy between the ‘domestic’ and the ‘international’ (Rosenau 1987: 1). In the 1990s David Campbell gave vast popularity to Richard Ashley’s notion of foreign policy as a ‘specific sort of boundary-producing political performance’ separating inside from outside, identity from alterity (Campbell 1998: 62). Most recently, analysts have reiterated the call for a multi-causal analysis of foreign policy able to accommodate sources placed on the inside and outside of states (Eun 2012).
Despite this broad agreement, the twin questions of how to conceptualise foreign policy as produced at the cusp of the ‘domestic’ and the ‘international’ and how to couch this conceptualisation into a larger understanding of causation in foreign policy have been subject to a less sustained and self-conscious debate than one may think.1 Attempts have floundered, arguably not just due to the complexity of the task, but also due to the poor propensity towards theory and metatheory within FPA and, in particular, to the conflation and reification of a number of relevant epistemological and ontological issues, only a few of which have been properly understood in terms of their foreign policy implications. However, in so far as foreign policy is understood as a form of action (Carlsnaes 1986: 27; Hill 2003: 283–307), and as a process oriented to some – albeit multiple, contrasting and collectively defined – goals and aims, the issue of how to explain foreign policy can hardly be avoided. Indeed, it echoes a classic question in the analysis of politics.
This chapter seeks to advance our understanding of causation in FPA by providing an overview of the state of affairs, formulating a diagnosis and proposing a way forward. It does so in four steps. First, it surveys and unpacks the ontological, epistemological and conceptual questions at stake in a causal analysis of foreign policy. Second, it offers a taxonomy of empirical theories employed by scholars in FPA and IR to account for the patterns of causation between foreign policy, domestic politics, and international relations, exposing their limitations. Third, drawing on the critical realism of Colin Hay and Bob Jessop, it suggests a way to systematically reconceptualise the problem and redraw the boundaries of the field around a tripartite distinction between monocausal, dualist and dialectical models. Fourth, and last, the chapter fleshes out the contours of the strategic-relational model, a dialectical approach to foreign policy that bears particular promise. Because of the way in which it solves the tension between domestic and international determinants of foreign policy, and between naturalistic and intepretativist epistemology, the strategic-relational model manages to transcend without conflating the unnecessarily rigid cleavages informing alternative approaches. Grounded in an understanding of foreign policy as a dialectic interplay of context, strategy and discourse, the model avoids the limits of monocausal accounts by integrating factors placed at different levels, moves beyond dualist views by highlighting the dynamic interaction among variables, reconciles causal and systemic analysis while foregrounding contingency and complexity, and provides a plausible path out of the interpretative impasse currently characterising the debate on causation in FPA.
Mapping the problem-field
The issue of how foreign policy is predicated on the shifting nexus between international relations and domestic politics has been approached from a variety of angles within contemporary FPA as a subfield of IR. The behavioural revolution which played such a large part in launching FPA as a ‘scientific’ enterprise, however, can hardly claim to have started a debate which has long roots in the history of political thought. By the seventeenth century, the tradition of thought variously inspired by the writings of Niccolò Machiavelli and Giovanni Botero and known as Raison d’Ètat had already initiated a conservative, state-centric turn in historical scholarship – away from its previous focus on peoples, ethnicities and nations – that relied on a specific, prescriptive understanding of the relation between foreign policy, domestic politics and international relations (Butterfield 1975). In the words of one of its most eminent followers, Leopold von Ranke, ‘the degree of independence determines a state’s position in the world, and requires that the state mobilize all its inner resources for the goal of self-preservation. This is its supreme law’ (von Ranke 1950: 169; see also Simms 2003; Smith 1999: 13–17).
This law came to be known as Primat der Aussenpolitik – the primacy of foreign policy (Czempiel 1963; Schulin 1987) – a model which was to inspire generations of historians as well as political scientists to produce sweeping accounts of modern European and international history in terms of the international dynamics of security and power politics (see, for instance, Hintze 1962 [1906] and Tilly 1975; and, in IR, Kennedy 1986; Kennedy 1988; Gourevitch 1978). It was not until the 1930s, but especially from the 1960s, that the traditional primacy of foreign policy in historical scholarship was challenged from ‘below’, as it were, by its very mirror image. Eckart Kehr initiated a ‘revisionist’ strand of scholarship centred around the Primat der Innenpolitik (the primacy of domestic politics) which foregrounded the crucial role of economic interests and élites, instead of supposed international necessities, to explain German foreign policy during the Wilhelmine period and in the run-up to the First World War (Kehr 1977). Drawing on Kehr, Hans Ulrich Wehler explicitly used the category of imperialism to explain Bismarck’s foreign policy, thus openly calling on Marxist and critical theory (Wehler 1984, 1973; Hillgruber 1987). But it was Fritz Fischer’s 1961 book on the causes of the First World War which, finally, gave the greatest popularity to a domestic- and economic-based analysis of foreign policy (Fischer 1975 [1961]), thus igniting a controversy which was to divide scholars across Europe, and then across the Atlantic, in two opposing camps: ‘orthodox’ vs. ‘revisionists’ (Stephanson 1994).
At the heart of this historical debate was arguably an antithetical understanding of the sources of foreign policy behaviour, namely a distinction between international and domestic causes, as well as the espousal of opposite traditions in international political theory. On the one hand, the ‘Rankean’ penchant for political Realism – stretching from Thucydides to Rousseau through Machiavelli and Hobbes – and its international accounts of foreign policy, its typical cross-historical generalisations and its privileged focus on power and security. On the other, the appeal of Marxism and Liberalism to Kehr’s followers, interested in uncovering the domestic sources of foreign policy, and concerned about issues of political responsibility and the possibility of change.
If this is arguably the pre-history of contemporary attempts to arrive at satisfactory conceptualisations of foreign policy in its international and domestic determinants, a lot more layers have calcified around the issue since then. These need bringing into sharper focus in order to gain a clear picture of what is at stake in an account of foreign policy from the point of view of international and domestic politics. Of ontological, epistemological and disciplinary kind, these layers have been often understood as coterminous and confluent – if not wholly blurred – in contemporary FPA. For my argument, however, it is important to consider them separately.
The first and most notorious question revolves around the issue of levels of analysis. Originally conceptualised by David Singer, anticipated by Kenneth Waltz and revisited by Martin Hollis and Steve Smith, the level-of-analysis debate addressed the issue of which level to choose in order to explain events in world politics (Singer 1961; Waltz 1959; Hollis and Smith 1990, 1991, 1992; Buzan 1985; Wendt 1991, 1992). Arguably, the focus of this debate was never foreign policy, but rather the outcomes of interaction among states, i.e. the aggregate effects of interaction. In fact, in the level-of-analysis framework foreign policy was simply bypassed, with the three levels of the individual, the state and the system said to exhaust the universe of world politics. While Waltz and Singer raised and solved the question of how to explain aggregate outcomes by choosing the third level (that of the international system), this solution never had much purchase in FPA (for an exception, see Hollis and Smith 1986). The mission of the then emerging discipline was to unpack notions (such as that of the state) that the levels of analysis approach used precisely as assumptions. The levels of analysis framework was hugely influential and certainly helpful in explaining ‘outside in’ or ‘inside out’ dynamics, i.e. how systemic events shape unit-level attributes (also called ‘second image-reversed’ perspectives), or how unit-level attributes influence the nature/characteristics of the system (‘second image’) (Gourevitch 1978). However, it was virtually silent on how these dynamics were mediated by foreign policy – in fact, it was silent about foreign policy tout court.
Although primarily concerned with ontology, the level-of-analysis framework re-ignited the classic epistemological issue of causation and explanation in IR and, to a lesser extent, FPA. This is a classic debate in the social sciences, with roots that extend as far back as the sociological diatribe between the holism of Emile Durkheim to the methodological individualism of Max Weber (Taylor 1985; Dallmayr and McCarthy 1977; Hollis and Smith 1990). Since the advent of the so-called post-positivist turn the debate has become familiar to the IR or FPA scholar (Wight 2012; Waever 1996), especially in its incarnation as ‘agency-structure’ debate (Giddens 1984; Wendt 1987, 1999; Dessler 1989; Hollis and Smith 1990; Doty 1997; Wight 1999, 2006; Bieler and Morton 2001; Suganami 1999; Carlsnaes 1992, 1994, 2013). At its most basic level, the debate focused on the question of how to account for action, whether in terms of the environment (system, structure) in which a particular actor is embedded, or in terms of the properties of the actor (motives, beliefs, intentions, aims, etc.). In so far as they deal with action, and to the extent that they are interested in looking for its causes, all social sciences need to grapple with this question – foreign policy included.
Figure 1.1 | Foreign policy, images and levels of analysis. While arrow a represents how international political events are influenced by individual-level properties, arrow b represents how domestic politics influences international politics: these are the only two causal arrows analysed in the original level-of-analysis discussion by Singer (‘images’ à la Waltz). Arrow c represents ‘second-image-reversed’ influences, studied by Gourevitch, inter alia. Arrows d, e and f, on the other hand, represent international, domestic and individual influences on foreign policy. It is only the latter that are of concern here. |
As some scholars have pointed out, the application of this frame of analysis to issues of foreign policy is intuitively fitting and heuristically fruitful, albeit not entirely straightforward (Hill 2003: 25–30; Carlsnaes 2013). After all, isn’t foreign policy caught between the Scylla of structural, environmental forces and the Charibdes of agential, unit attributes? Naturally, depending upon the unit of analysis considered (e.g. the state, the government, the individual foreign policy maker, etc.), structures will be of different kinds (e.g. the international environment, the bureaucracy, the Cabinet, etc.), as will the unit’s attributes (e.g. the domestic political process, group dynamics, the individual’s psychology, etc.). Thus, the ‘agency-structure’ framework can be applied at each step of the foreign policy-making ladder, from the individual level to the aggregate level of the state, and vice versa (cf. Hollis and Smith 1990: 1–10 and 196–216; see also Wendt 1991: 384).
However, whether we settle questions of ontology in terms of levels of analysis or agency and structure, the epistemological choice about how we actually treat the explanans and construct an explanation of foreign policy necessarily remains a separate one. This epistemological choice has often been subsumed, erroneously, in peremptory decisions about the alleged primacy of structure or agency. Such a move, however, is obfuscating in so far as it forecloses epistemological questions of four different kinds.
First, there is the issue of whether to adopt a naturalistic or interpretative approach. With Gabriel Almond, this is often framed in terms of the choice between an external, naturalistic, ‘clock-like’ type of inquiry and an internal, interpretative, ‘cloud-like’ approach (Almond with Genco 1990). Or, in the Weberian terms utilised by Hollis and Smith, the two ‘stories’ of ‘explaining’ or ‘understanding’ (Hollis and Smith 1990). This choice evidently presents itself at every level of analysis, for both agents and structures, as all of these ontological elements are open to a naturalistic or interpretative treatment.2
Second, there is the question of what time frame our account is supposed to accommodate, and what strategies we utilise to conceptualise change. The opposition of Jean Braudel’s longue durée with the conjuncture comes to mind here as a particularly relevant example of the tension between diachronic and synchronic modes of investigation (Braudel 1972; Aron 1964). From this perspective, it has often been concluded that the structural level is not just the realm of ‘inevitability’ and structural laws, as already said, but also the domain of general, long-term developments, of those ‘vast impersonal forces’ that run their course through decades, if not centuries. On the other hand, analyses placed at the level of agency are supposed to offer snapshots of more short-term, conjunctural developments whose effects may be exhausted in a very brief span of time. This confluence of ontological and epistemological strategies is, however, far from inevitable or necessary.
Third, there is the fundamental question of what passes or counts for cause in FPA and IR – a question which is implicit in much of the literature, yet is hardly ever discussed explicitly (Suganami 1996; Kurki 2008). Essentially, the choice is whether a cause is only to be understood as identifying empirically observable, objective and efficient entities that bring about change directly, mechanically a...
