Critical Review of ‘Offense, Defense and the Causes of War’

This essay will provide a critical evaluation of Stephen Van Evera's article 'Offense, Defense and the Causes of War'. This will be done by beginning with a brief overview of the article and the key points the author makes which lead him to reach his eventual conclusions.

The main arguments put forward by the author will then be placed within the wider context of the literature on the offense-defense argument and causes of war, studying the strengths of this article and then moving on to the weaknesses, continuously comparing Van Evera's thesis to other academia.The critical evaluation will then conclude with the validity of the author's argument in regards to the issues discussed. Van Evera is a defensive realist and in this article he constructs his arguments around the offense-defense theory, and states that the probability of war is increased dramatically when "conquest is easy". However, changes in the offense-defense balance can greatly increase or decrease the chances of war (Van Evera 1998 p5).

He goes on to argue that even if these 'changes' in balance are merely perceptions they still greatly affect the likelihood of war (Van Evera 1998 p41). Van Evera begins with an outline of the 'effects' of offense dominance and lists ten 'war-causing effects' that occur during offensive dominance. This is summarised later in the article in the form of a table which displays how all explanations conclude in war.These are 'opportunistic expansion', 'defensive expansion', 'fierce resistance to expansion by other states', 'moving first is more rewarding', 'windows are larger, and more dangerous', 'faits accomplis are more common and more dangerous', 'states negotiate less and reach fewer agreements', 'secrecy is more common and more dangerous', 'more intense arms racing', and 'offense grows even stronger' (Van Evera 1998 p12). Furthermore, he gives examples of failed offensive countries, such as Germany (Van Evera 1998 p42), to display the flaws in being offensive and validate his own viewpoint.To test his hypothesis he uses three case studies, Europe since 1789, ancient-China, and the United States (US) since 1789, and concludes that the offense-defense theory "has the attributes of a good theory" and a "wide explanatory range," and it also has "wide real-world applicability" (Van Evera 1998 p41).

The solution to this problem of offensive realism, according to the author, is for states to adhere to the defensive branch of realism, as this reduces the risk of war dramatically and makes issues such as pre-emptive war very undesirable (Van Evera 1998 p9).He also implies that during any given period that defensive realism dominates, states are more cooperative with one another, and this thus creates stability in the international realm (Van Evera 1998 p10-11). Van Evera's main argument is that war is more likely to occur when a state believes that conquest is easy (Van Evera 1998 p5). This argument is prominent amongst offensive realists and there is much academia to support this, such as Morgenthau, who claims that states continuously attempt "to accumulate power internationally" (1955 p26).In addition, other offensive realists works support another of Van Evera's main theories, that the offense-defense balance consistently changes, as each state 'races' to surpass the other technologically (Van Evera 1998 p14). This is supported by Mearsheimer, who states that offensive weapons have been developed in response to defensive weapons, such as the tank (Mearsheimer 1983 p25).

As aforementioned, Van Evera is a defensive realist, and thus the aim of the article is to state the various disadvantages of offensive realism.As Van Evera's arguments are still widely accepted it is evident that the article has many strengths, with academics, such as Paul, strongly agreeing with Van Evera that when a state can easily impose its will upon another it will engage in offensive policies (Paul 2004 p5). Additionally, the 1998 Spring edition of 'International Security' was headlined by two articles on the offensive-defensive theory, the first of which being the article under evaluation, the other being by Glaser and Kaufmann.This article concurs with Van Evera's main argument that having an "Offense advantage makes war more likely" (Glaser ; Kaufmann 1998 p48). Furthermore, the theory of defensive realism that is favoured by Van Evera is also supported by various other political scientists, with much academia on the topic that predates the article by a number of years, such as Mastanduno, who argued that "states are not 'gap maximisers' " (Mastanduno 1991 p79).

Furthermore, this is also heavily supported by the "balance of threat" theory put forward by defensive realist Stephen Walt.His thesis claims that if an imbalance of threat occurs then "states will form alliances or increase their internal efforts in order to reduce their vulnerability" (Walt 1987 p263). Walt, as well as Joesph Grieco, also argue that security is the principle goal of the state and as such "only seek the requisite amount of power to ensure their own survival" (Dunne ; Schmidt 2001 p152). This branch of realism is the one adhered to by Van Evera, and according to this view, states are defensive actors and thus do not seek to gain further power if it comes at the cost of its own security.

Thus when conquest is difficult, the prominence of powerful states is accepted by smaller states, as "they adopt sensible and effective balancing strategies" (Hopf 1991 p488), and then become "less aggressive and more willing to accept the status quo" (Van Evera 1998 p7). This can be seen as a strength of the article, as this statement is supported by various academics, such as Dunne and Schmidt, who also argue that this is the case (2001 p152).Van Evera also enjoys much consensus on one of his concluding statements, that the offense-defense theory "has the attributes of a good theory" and a "wide explanatory range," and it also has "wide real-world applicability" (Van Evera 1998 p41). Political theorists, such as Labs, also argue that offensive realism provides the best theory of international relations (Labs 1998 cited by Donnelly 2000 p76).

Since the publication of Van Evera's article there has been much debate on the issue, with the article coming under heavy criticism.Almost immediately after the publication, Van Evera was subject to much scrutiny in an editorial from the journal 'International Security'. In this critique Davis reveals a strength of Van Evera's work, but presents it as a weakness, in that he claims that Van Evera overstates the extent to which his theory stands up to empirical tests (Davis et al 1998/1999 p181). He also states that Van Evera presents an argument in which every outcome is in some way consistent with the hypothesis (Davis et al 1998/1999 p181).

Furthermore, from the same critique, Finel claims that the works of Van Evera has tightened the conceptual logic of the issue and have thus added much needed refinements to the ongoing debate of the article (Davis et al 1998/1999 p182). Van Evera justifies his claims by responding to this critique, and states that all his concepts correspond to phenomena in the real world, which is the way that International Relations Theory should be studied (Davis et al 1998/1999 p197).However, there are numerous weaknesses of Van Evera's article as he only implicitly argues that defensive realism is the solution by stating the disadvantages of offensive realism. Furthermore, he doesn't implicitly offer much to explain why defensive realism is the solution.

A further criticism that can be made of Van Evera's work is that his definitions are somewhat flawed, and that he needs to narrows his definitions and streamline his hypotheses before he can claim that these have been empirically tested.Additionally, while this article explains the consequences of offensive dominance, arguing that states seek only to conquer and maximise their power, academics such as Donnelly counter this by arguing that states not only seek to increase its relative power, but they do also attempt to increase their "security by aiming to preserve the status quo" (Donnelly 2000 p64). This argument thus incorporates both offensive and defensive branches of realism, as both bolster a states position in the international realm and this is something that Van Evera fails to acknowledge.Additionally, key offensive theorists, such as Mearsheimer, argue that it is necessary for states to maximise their power in relation to other states due to the anarchic system (Mearsheimer 1990 p12). Furthermore, Mearsheimer's arguments are further validated by Zakaria, who supported this theory that a state should "increase its control over that environment through the persistent expansion of its political interests abroad" (Zakaria 1998 cited by Donnelly 2000 p63).

Additionally, Labs argued that offensive realism provided the best theory of international relations (Labs 1998 cited by Donnelly 2000 p64). Furthermore, in the aforementioned critique published in response to Van Evera's article, Davis states that the concept made by Jervis was destroyed by Van Evera by adding diplomatic factors to an already useful concept. In addition, he goes on to say that Van Evera's hypothesis is "imprecise, internally incoherent and as a result cannot be tested in any meaningful fashion" (Davis et al 1998/1999 p180).This highlights one of the articles main weaknesses, in that the offense-defense theory explains all of the case studies provided by Van Evera, and does not provide any other explanation for the case studies aside from his theory. Furthermore, the offense-defense theory that Van Evera discusses cannot be branded as a strong theory as it has failed to measure against competing explanations for the case studies he refers to.

Additionally, Davis states that the biggest weakness of Van Evera's thesis is that he claims that Van Evera is incredibly naive, due to the fact that Van Evera believes that he has the field of research on this topic to himself. Davis describes this by saying that he is "playing something more akin to solitaire than to science" (Davis et al 1998/1999 p181). In addition, Van Evera's main argument, that war is more likely when offensive dominance occurs is also subject to much scrutiny. Mearsheimer argues that this is a very naive argument and thus is not useful (Mearsheimer 1983 p24).

Furthermore, Mearsheimer also argues that the statement also fails to explain why it is that the offense would have an advantage over the defense, and also that any "theory of conventional deterrence must stipulate the conditions under which the offense has an advantage over the defense" (Mearsheimer 1983 p25). Finel furthers this criticism of Van Evera's article, by suggesting that there are three main problems with the offense-defense theory. Firstly, the theory ignores interaction effects in warfare. Secondly, it makes poor judgements in consideration of the links between territorial control and victory in war.

Finally, Finel suggests that the theory is not as well conceptualised as it could be (Davis et al 1998/1999 p182). Finel takes this criticism further by suggesting that Van Evera failed to define what conquest is (Davis et al 1998/1999 p188). This therefore means that the theory generated by Van Evera is "methodologically flawed and conceptually muddled" (Davis et al 1998/1999 p188), which is something also argued by Stacie Goddard, who states that the "theory has suffered from a lack of methodological rigour (Davis et al 1998/1999 p189).What's more, one of the few solutions offered by Van Evera to offensive dominance, which is alliance-making, is also flawed, as Walt recognizes that coalitions are not necessarily defensive measures taken by states and offensive alliances can also be made, for example, an alliance can be "intended either to provide the means for an attack on some third party" (Walt 1997 p157).

Conclusively, the offense-defense issue has split political theorists for years before Van Evera's critique of offensive realism, such as the aforementioned works of offensive realists such as Mearsheimer (1983 and 1990), and defensive realists such as Walt (1987 and 1997). However, since the article the issue has still continued, with academics such as Paul agreeing with Van Evera that when a state can easily impose its will upon another it will engage in offensive policies (Paul 2004 p5).Furthermore, it is evident that Van Evera's article has various strengths and weaknesses, which have been discussed throughout this critical review, and as such the article still provokes comments on the two differing theories of the offense-defense debate and will probably be revisited on numerous occasions in the research of the aforementioned academics to further the development of the existing theories.