The Strategy of Territorial Conflict

David B. Carter The Pennsylvania State University

Many empirical studies have found that disputes over territory are central to the outbreak and intensity of the majority ofinterstate military conflict. However, the existing literature lacks an explicit theoretical link between the role territory playsin disputes and the outbreak of violence as well as an exploration of how the control of territory is related to conventionalmilitary capabilities. This article demonstrates that the targets of territorial claims can consolidate their control over disputedterritory to improve their ability to fight effectively on it. The empirical analysis suggests that when territory is strategicallylocated, target states are more likely to consolidate their position, while challenger states are less likely to escalate militarily.Furthermore, when the presence of territorial characteristics such as strategic location makes consolidation an effectivestrategy, target states are increasingly likely to consolidate as they face stronger opponents.

In the early morning hours of February 17, 1979,China crossed into Vietnam with nearly 100,000troops backed up by heavy artillery and armored

units. The stated purpose of the incursion was to punishVietnam for its behavior related to disputed territory inthe border region between the two countries and somedisputed islands. At the outset of the conflict, China pos-sessed both a numerical and a qualitative advantage in itsforces stationed at the border. Despite these advantagesand a lopsided advantage in overall capabilities, the Chi-nese incurred very high costs and experienced great dif-ficulty penetrating Vietnamese defenses. The Vietnamesewere relatively successful in large part because severalkey characteristics of the disputed territory made consol-idation of their position at the border quite successful.Vietnam had constructed a series of border defenses andtraps in the rugged mountainous terrain. Additionally, theVietnamese implemented a provocative policy of forcedemigration that sent about 200,000 ethnic Chinese fromthe Vietnamese side of the border, fostering chaos and in-stability on the Chinese side while removing a populationpotentially friendly to its opponent. After about a monthof tough and costly fighting, China withdrew its forces.The Sino-Vietnamese case is puzzling in light of the bulkof empirical findings regarding the balance of capabili-

David B. Carter, Department of Political Science, The Pennsylvania State University, Pond Laboratory 211, University Park, PA 16802(dbc10@psu.edu).

For their helpful comments on this project, I am very grateful to Deniz Aksoy, Kevin Clarke, Mark Fey, Hein Goemans, PatrickJames, Jeremy Kedziora, Jacek Kugler, Doug Lemke, Eduardo Leoni, Daniel Morey, Dick Niemi, Arnd Plagge, Matthew Platt, ShawnRamirez, Yoji Sekiya, Curt Signorino, Arthur Spirling, Randy Stone, five anonymous reviewers, and all participants of the WatsonCenter for Conflict and Cooperation Seminar Series at the University of Rochester. I thank Paul Huth and Todd Allee for providingme with the data used in this analysis. Any problems in the article remain my responsibility. Replication materials are available athttp://www.personal.psu.edu/dbc10/territorialstrategy.html.

ties and states’ propensity and ability to escalate and fightmilitarized conflicts. Despite the existence of numeroussimilar cases in which a relatively weak opponent wasable to favorably consolidate its control over territory ina way that increased its military capabilities relative toa stronger power (e.g., the Vietnam War), existing ideasabout how territory relates to military capabilities are notwell developed.

I argue that in cases such as the Sino-Vietnamesedispute, territory plays an important, yet largely unex-plored, role in affecting military capabilities. Theoreti-cally, the idea that the control of territory significantlyaffects states’ military capabilities has been noted (Fearon1995a, 1995b; Powell 2006; Smith 1998; Wagner 2000,2007). However, how the possession of territory affectsmilitary capabilities is largely unexplored. I argue thatthe effect that a piece of territory has on the balance ofmilitary capabilities between disputants is in large part afunction of the actions the state in control of the territorycan take in it. In territorial disputes, it is generally the tar-get of territorial claims that controls disputed territory.Consequently, I argue that territory can augment a targetstate’s capabilities if its control of it can be consolidatedin a way that improves its military position relative to itsopponent.

American Journal of Political Science, Vol. 54, No. 4, October 2010, Pp. 969–987

C©2010, Midwest Political Science Association DOI: 10.1111/j.1540-5907.2010.00471.x

969

970 DAVID B. CARTER

Territorial consolidation specifically refers to at-tempts by the target state to alter the military, economic,or political status quo in disputed territory such that itsmilitary capabilities are augmented. Obviously, territorialconsolidation is not equally effective across all pieces ofterritory. I argue that territory with characteristics directlyrelevant to the military, economic, and political situationis especially ripe for consolidation. Specifically, I focuson the characteristics of strategic location, the presenceof valuable economic resources, and/or the presence of aminority with ties to the challenger state. The presence ofany or all of these three characteristics presents the targetstate with conditions that, all else equal, can be effectivelymanipulated to alter the status quo.1 Thus, two factorshelped Vietnam to successfully battle against long odds:the strategic nature of the mountainous region as the bot-tleneck to the entire Red River valley and the presence ofa sizable population of ethnic Chinese. The Vietnamesewere able to effectively consolidate because key character-istics of the disputed territory facilitated alteration of localconditions in a militarily favorable manner. I show thatthis case is just one in a general, but thus far overlooked,pattern in territorial conflict.

This article demonstrates that territory plays an im-portant endogenous role in shaping military capabilities.The possession of territory with important characteris-tics affects a state’s ability to consolidate and improveits military position. To show the implications of terri-tory’s endogenous role, I develop an intuitive theoreticalframework that illustrates why and when the target of aterritorial claim strategically consolidates its position inthe disputed territory. The model demonstrates that astarget states are weaker relative to their opponents, theyare more likely to consolidate their position in the dis-puted region, conditional on consolidation having a largeenough effect on a target’s military prospects. Targetsconsolidate disputed territory when they face strongeropponents because strong challengers represent more ofa military threat and are more apt to find the applica-tion of military force an attractive means to resolve adispute. I demonstrate empirically that consolidation iseffective in reshaping the military situation when terri-tory is strategically located, as this increases the probabil-ity the target consolidates and simultaneously deters somechallengers from escalating disputes. Furthermore, a clearunderstanding of the logic behind the target’s decision toconsolidate sheds new light on a set of interesting cases inwhich weak states very effectively defended territory whenfacing stronger opponents (e.g., the Sino-Vietnamese dis-pute).

1All three characteristics are specifically defined later in the article.

The Territorial Gap in the ConflictLiterature

Scholars of international relations have long found con-flict over territory to be one of the key factors that leadstates to war (Goertz and Diehl 1992; Hensel 2000; Hill1945; Holsti 1991; Kocs 1995; Luard 1986; Vasquez 1995).For instance, in analyzing the population of interstatewars from 1816 to 1992, Paul Hensel (2000, 65) findsthat over 50% of wars included issues of disputed terri-tory. Additionally, regardless of whether two states have acontiguous border, disputes with a territorial dimensionare much more likely to lead to conflicts with fatalitiesthan disputes with no territorial dimension (Hensel 2000,72–74). Similarly, John Vasquez (1995, 284) finds that infive historical periods from 1648 to 1990, a minimumof 79% of wars were fought over territory-related issues.In spite of a wealth of powerful empirical evidence thatterritorial disputes play a central role in the outbreak ofinternational conflict, no study explores how territorialcharacteristics affect states’ abilities or incentives to con-solidate their military position. Consequently, our under-standing of how consolidation affects military capabilitiesand thus the outbreak and character of military conflictis limited.

A massive amount of scholarship deals with theconnection between disputed territory and the even-tual outbreak of war. One prominent vein of literatureis the territoriality explanation for war. Scholars advo-cating this explanation generally argue that geographicproximity provides states with an opportunity to initiateconflict, while the relative salience of territorial disputesmakes states willing to bear the associated risks and costs(Goertz and Diehl 1992; Senese 2005; Vasquez 1995).Thus, authors such as Senese (2005) investigate the ef-fects of variables such as contiguity and the geographicdistance between disputants. While such analysis is quiteuseful in highlighting broad empirical trends, it only pro-vides a very general analysis of the conventional militarysituation facing states in territorial disputes. Furthermore,distance and contiguity undoubtedly affect states’ mili-tary calculations, but neither are variables that states havemuch ability to alter. Thus, the territoriality viewpointleaves us with a rather static view of the military strategysurrounding territorial disputes. Furthermore, it provideslittle traction in putting forth a rich explanation of casessuch as the Sino-Vietnamese dispute since nearly all ter-ritorial disputes involve contiguous states.

The work of Huth (1996) and Huth and Allee (2002)goes a step further by analyzing the effect that impor-tant territorial characteristics have on dispute initiation

THE STRATEGY OF TERRITORIAL CONFLICT 971

FIGURE 1 Consolidation of Disputed Territory

and escalation. Despite providing several significantadvances, neither of these studies addresses the idea thatterritory itself is a source of power, especially when adisputant can strategically consolidate disputed territory.Territorial characteristics such as strategic location areanalyzed primarily as factors that increase the disputants’valuations of the territory and not as variables that canaffect states’ means to obtain territory (i.e., military capa-bilities). Consequently, the theoretical frameworks em-ployed by Huth (1996) and Huth and Allee (2002) donot make significant distinctions between how territorialcharacteristics affect the behavior of the challenger andtarget states. Rather, both the challenger and target arehypothesized to be more likely to escalate as their relativecapabilities increase and when the territory holds valuablecharacteristics such as strategic location. As I show below,territory with key characteristics can be consolidated ina way that augments the capabilities of target states. Thisfinding has important and previously unexplored impli-cations for how variables such as relative military capabil-ities or strategic location influence challenger and targetbehavior.

As noted above, the theoretical idea that territory isa source of power has been recognized by scholars buthas not been developed. Fearon (1995b, 408–9) and Pow-ell (2006, 186–88) argue that disputes over territory withcharacteristics that make it a source of military power(e.g., valuable economic resources) can lead to militarizedconflict due to commitment problems. In short, becausethe transfer of such territory leads to large shifts in the bal-

ance of capabilities, the potential recipient of the territorycannot credibly commit not to use its increased power togain further concessions.2 However, when territory pos-sesses characteristics that are a source of power, the targethas incentives to consolidate its control of the territory toaugment its capabilities, a point that has been previouslyoverlooked. This article focuses on how the possession ofterritory with key characteristics affects a state’s ability toconsolidate and improve its military position.

The Territorial Dispute Game

In order to show the strategic logic of the consolidation ofterritory, or why and when target states consolidate terri-tory to augment their military capabilities, the completeand perfect information game-theoretic model picturedin Figure 1 is formulated. The game starts with a po-tential challenger state that chooses to keep active (oractivate) a challenge to the territorial status quo or not.If no dispute is active, then the status quo ensues and thechallenger state receives a payoff of 0, while the target statereceives its valuation of the disputed territory, vt(�) > 0.The variable � ∈ [0, 1] measures the territory’s tangiblecharacteristics relevant to the state’s military, political,and/or economic situation. Thus, values of � close to 1

2This line of reasoning seems particularly applicable to disputesover the status of strategically located buffer states that neitherdisputant controls at the outset (Fazal 2004).

972 DAVID B. CARTER

indicate territory that is endowed with many relevant tan-gible characteristics, while values close to 0 indicate thatthe territory has few, if any, identifiable characteristics. Ifa challenge is made, the target decides whether to respondvia an attempt to consolidate its position in the disputedterritory. Consolidation refers to attempts by the targetstate to alter the military, economic, or political statusquo in the disputed territory in its favor. Thus, via con-solidation, the target is trying to strengthen its grip onthe disputed territory in a way that augments its capabil-ities relative to the challenger. Consolidation is intendedto make fighting easier for the target. The final move istaken by the challenger, who after observing target ac-tion, decides whether to attempt to take the territory viamilitary force.3

If the target does not consolidate and the chal-lenger does not choose to use force, the challenger re-ceives a payoff of xc ∈ (−∞, ∞), while the target receivesvt(�) + xt , where xt ∈ (−∞, ∞). The xi ∈ (−∞, ∞),where i ∈ {c , t}, terms capture any additional negative orpositive payoff associated with the two outcomes in whicha dispute is active but fighting does not occur. A reason-able argument can be made for the xi terms being eitherpositive or negative. On the one hand, it is widely notedthat territorial challengers have an incentive to ensure aclaim is kept open, even if they do not aggressively pursuetheir claim. Given the importance of precedents in terri-torial issues, challenger states let claims become inactiveat the risk of losing any solid legal basis for their claims(Murphy 1990; Ratner 1996). It is also plausible that chal-lenger states may face something similar to an “audi-ence cost,” domestically (Fearon 1994) or internationally(Sartori 2003), for not acting aggressively on an activeclaim. Thus, while the literature on territorial claims sug-gests that challenger states receive positive payoff fromkeeping disputes open, the theoretical literature on audi-ence costs suggests that the opposite is also plausible.

If the target does not consolidate and the chal-lenger chooses to fight, then both sides receive apayoff of their valuation of the territory multiplied by theprobability they win, p(m)vc (�) for the challenger and[1 − p(m)]vt(�) for the target, minus the costs of fight-ing, kc > 0 for the challenger and kt > 0 for the target.The function p(m) ∈ [0, 1] measures the probability thechallenger wins a war over the disputed territory, whichis an increasing function of the challenger’s share of thetwo states’ total military capabilities, or m.

3This model is not intended to explain why states initiate militarizedconflicts instead of agreeing to some division of the disputed terri-tory. Relatedly, Carter and Goemans (2010) provide considerableevidence that territory is not treated continuously when divided upby states.

If the target chooses to consolidate its position inthe disputed territory and the challenger chooses to fight,the players’ payoffs change to reflect the effect of targetefforts. The coefficient �(�) ∈ (0, 1] measures how effec-tive consolidation is in altering the military situation thatfaces the two states in the disputed territory. Values thatapproach 0 indicate that consolidation is extraordinarilyeffective in augmenting target capabilities while valuesthat approach 1 indicate that consolidation efforts havelittle effect. We assume that � is a nonincreasing functionof the territory’s relevant military, political, and economiccharacteristics. Thus, while characteristics can increasethe effectiveness of consolidation, they are not assumedto always have such an effect. Given that target consoli-dation is also costly, the target’s utility for fighting givenconsolidation is [1 − p(m)�(�)]vt(�) − kt − k�, wherek� > 0 captures this cost. The challenger’s utility for fight-ing, p(m)�(�)vc (�) − kc , is (possibly) negatively affectedby consolidation.4 If the target consolidates but the chal-lenger chooses not to fight, the challenger receives xc ,while the target receives vt(�) + xt − k�.

Summary of Results

I use the subgame perfect equilibrium (SPE) concept toanalyze the game. Since the game is sequential and playedonce with complete and perfect information, there is aunique equilibrium in pure strategies for any distributionof the parameters in the model (Mas-Colell, Whinston,and Green 1995, 276).5 Complete and perfect informationis appropriate here because I am primarily interested inthe observable physical effects that consolidation has onthe military situation, not in informational asymmetry orhow target consolidation signals “resolve.”

The equilibrium conditions are summarized inTable 1. The table displays the equilibrium behavior ofthe two players in the game with the corresponding con-ditions that must hold for each possible path of play to beoptimal. The table is constructed to correspond with thelogic of backwards induction the challenger and target useto decide what action to choose. The column on the farleft indicates the three possible equilibrium paths of playfor the challenger when it has already activated a dispute.The formal condition for this path of play to be optimal

4The character of the results are not dependent on the assumptionthat p(.) is a linear function of �(�). We make the relation linearin order to keep the equilibrium conditions relatively intuitive.

5For formal demonstration of this uniqueness result, see Proposi-tion 9.B.2 in Mas-Colell, Whinston, and Green (1995, 276). Theonly additional requirement for this result to hold is that no playerbe indifferent over two possible actions. I make the fairly standardassumption that indifferent players take the less conflictual action.

THE STRATEGY OF TERRITORIAL CONFLICT 973

TA

BL

E1

Sum

mar

yof

Res

ult

s

Ch

alle

nge

rA

ctio

nTa

rget

Act

ion

Con

dit

ion

for

Giv

enD

isp

ute

Giv

enD

isp

ute

Act

ive

Dis

pu

te

CA

lway

sFi

ghts

If�

(�)>

x c+

k cp

(m)v

c(�

)=⇒

TC

onso

lidat

esIf

�(�

)<

1−

k �p

(m)v

t(�

)�

(�)>

k cp

(m)v

c(�

)

CFi

ghts

IfN

oC

onso

lidat

ion

If�

(�)<

x c+

k cp

(m)v

c(�

)<

1=⇒

TC

onso

lidat

esto

Det

erIf

1>

k �−

x t−

k tp

(m)v

t(�

)x c

>0

CN

ever

Figh

tsIf

1<

x c+

k cp

(m)v

c(�

)=⇒

TN

ever

Con

solid

ates

x c>

0

for the challenger is stated in the next column to the right.The next column to the right indicates the three possiblepaths of play for the target state given that the challengerplays the strategy listed to the left in the same row. Theprecise condition that must be met is listed to the right ofthe target’s potential equilibrium strategy. Finally, the lastcolumn on the right states the condition that must holdfor the challenger to activate a dispute in the first moveof the game for each equilibrium path of play (i.e., thechallenger and target moves listed in the same row).

The challenger’s decision to militarize a dispute ornot depends upon how large its expected value for fight-ing over the territory, p(m)vc (�), is relative to the sumof the payoff to not fighting and the costs from fighting,xc + kc . In fact, all three of the possible conditions onthe challenger’s decision to fight can be nicely expressedin terms of the ratio xc +kc

p(m)vc (�) . As the first two columnsof the first row in Table 1 demonstrate, if the target con-solidates its position, the challenger will always fight if�(�) > xc +kc

p(m)vc (�) . This is intuitive given that higher valuesof �(�) indicate that consolidation is relatively ineffectiveand fighting is more attractive as xc +kc

p(m)vc (�) decreases. If thetarget faces a challenger that always fights, it consolidatesif �(�) < 1 − k�

p(m)vt (�) . If the relation between how attrac-tive fighting is for the challenger relative to the effective-ness of consolidation is such that �(�) < xc +kc

p(m)vc (�) < 1,the challenger only fights if the target does not consoli-date (i.e., the second row). Thus, in the case representedby the second row, target consolidation is effective enoughat changing the military environment to deter the chal-lenger. However, if the target does not consolidate, thechallenger will only fight if xc +kc

p(m)vc (�) < 1, which meansthat the expected value for fighting, p(m)vc (�), has to begreater than the sum of the costs and the payoff from anonviolent dispute, xc + kc . If xc +kc

p(m)vc (�) > 1 is true, thenthe challenger never chooses to fight and the target willnever consolidate since there is no threat of attack by thechallenger.

To summarize Table 1 in words, the first row showsthat if the challenger is strong enough and values the dis-puted territory highly enough, then it is possible for it toprefer to fight regardless of target actions. Furthermore,the target consolidates if doing so has a great enougheffect on its chances of winning. This is true when thechallenger always prefers fighting (i.e., the first row) aswell as when the challenger only prefers to fight in theabsence of consolidation (i.e., the second row). In thelatter case, consolidation acts as a deterrent to the chal-lenger. The third row depicts the scenario in which thechallenger is too weak or values the territory too little toever prefer a militarized confrontation. If this is the case,

974 DAVID B. CARTER

the target never consolidates as it knows that a militarizedconfrontation will not occur.

Empirical Implications

The model has interesting implications for the effect thattarget consolidation has on the conventional military sit-uation and the outbreak of militarized conflict. Since theimplications of the model are assessed below with data,I begin to introduce some of the variables used in thetheoretical model. Additionally, I introduce the languageof the statistical model by discussing the actors’ choicesprobabilistically.

First of all, the three equations in the second columnof Table 1 indicate that militarily stronger challengers aremore likely to fight, since xc +kc

p(m)vc (�) decreases as the militarybalance, m, increases to favor the challenger.

Expectation 1. The probability the challenger choosesto fight increases in its relative militarystrength.

On the hand, the effect of territorial characteristics,� , is more nuanced. If the target does not attempt con-solidation, then higher values of � increase the attractive-ness of military force to the challenger by increasing itsvaluation of the territory, vc (�). However, if the targetconsolidates, the effect of territorial characteristics is lessobvious. While more relevant characteristics still increasevc (�), they can also serve to decrease the probability thechallenger obtains the territory due to the effect of �(�)on the probability of winning. Thus, whether or not theprobability the challenger chooses to escalate militarilyis increasing in � when the target consolidates dependsupon the influence characteristics have on the rate of de-crease in �(�) relative to the rate of increase in vc (�). Thistension gets to the heart of whether territorial characteris-tics affect states’ valuations of territory (Huth 1996; Huthand Allee 2002) or whether territorial characteristics playa significant role in shaping states’ abilities to alter themilitary environment.

The target’s decision to consolidate or not is con-ditional on whether it faces a challenger that will likelyprefer war, as it will not consolidate if there is no crediblethreat of military action (see the final row of Table 1).All else equal, stronger challengers are more likely to pre-fer war. Thus, target states are more likely to augmenttheir capabilities through consolidation as they face rel-atively stronger challengers. Consequently, target consoli-dation efforts are increasingly likely as the challenger stateis stronger. However, this statement needs qualification,

as the target will not always attempt consolidation evenif facing a challenger that chooses war with certainty. Inorder for a consolidation attempt to be optimal, it mustaffect the military environment enough to justify the cost,k�. Recall that this is likely when the territory is endowedwith more relevant characteristics as well as when thechallenger is relatively strong. Examination of the twoinequalities in the fourth column of Table 1 indicatesthat both expressions are more likely to hold as m and �

increase. In sum, target consolidation is increasingly likelyas it faces stronger challengers, but only if the territorypossesses enough relevant characteristics to make it ripefor consolidation.

Expectation 2. If consolidation is effective, the probabil-ity the target consolidates increases in therelative strength of the challenger.

Given the important role effective consolidation playsin expectation 2 and in the theoretical model generally,I now make explicit the territorial characteristics I fo-cus on with data. Recall that the characteristics relevantto target consolidation efforts are related to the military,economic, and political situation. Thus, we need data oncharacteristics that pertain to these three dimensions. Iutilize data from Huth (1996) that indicate whether apiece of disputed territory (held by the target) is strategi-cally significant for the two states, contains economicallyvaluable resources, or contains a minority population ofthe same ethnicity as the majority in the challenger state.Thus, we have data on the military, economic, and polit-ical dimensions. Comparability to previous work in theliterature by Huth (1996) and Huth and Allee (2002) is animportant added benefit of focusing on these three char-acteristics. If it is indeed the case that any or all of thesecharacteristics significantly affect states’ means to reshapethe military environment, the result can be attributed totheory and speculation about whether the result is just anartifact of different data is avoided.

Before presenting the rest of the expectations, I makeexplicit the definitions used for strategic location, eco-nomic value, and/or a border minority with ties to thechallenger. Territory is defined as strategically located ifit is of significant military value to the target and chal-lenger. Territory can have strategic military value becauseof its geographic features (e.g., a mountain pass) or dueto the presence of key military installations (e.g., majortarget base site). Territory is of economic value if it con-tains natural resources that can generate revenue fromexport. Potentially profitable natural resources includeenergy sources (e.g., oil) or mineral deposits (e.g., ironore). Territory contains a border minority if there is a

THE STRATEGY OF TERRITORIAL CONFLICT 975

population in the disputed territory that is a minority inthe target state and shares ethnic or lingual ties to themajority population in the challenger state.

The relation between strategic territory and effectiveconsolidation is the most direct. The building (or expan-sion) of fortifications and military installations and/or thestationing of troops is generally undertaken to improvefighting conditions for a state. Undertaking such actionsin a location of strategic import is particularly effective.This is a point that is well documented by military histo-rians. For example, in discussing the vital import of goodcommunication lines to battle effectiveness in the eigh-teenth century, Hew Strachan notes that good roads andwaterways were of vital strategic importance to advancingarmies. “Communications became like funnels, throughwhich the army had to pass, and which revealed to its en-emy its likely course of advance. In peacetime, therefore, astate could take precautions by the erection of forts at theconfluences of rivers, at defiles and junctions.” For thisreason, “[t]he dominance of forts . . . was well-nigh crip-pling to the offensive” (1983, 11). The relation betweenstrategically located territory, consolidation, and the mil-itary environment is rather straightforward and leads usto the following two expectations.

Expectation 3a. Consolidation is more likely to be effectivewhen territory is strategically located.

Expectation 3b. The consolidation of strategically locatedterritory decreases the probability thechallenger will militarily escalate a dis-pute.

If valuable economic resources (e.g., coal) are presentin a (potentially) disputed piece of territory, I argue thattarget investment in the development of these resourceshas effects in the military realm. For instance, the miningof coal implies some investment in infrastructure (e.g., totransport the mined coal to the market), which also makesit easier for troops to be quickly moved or stationed at ornear the border.

Historically the connection between the developmentof infrastructure and the movement of troops and sup-plies and the maintenance of communication lines hasbeen both explicitly exploited and a fortunate side effectof the development of economic resources. William Mc-Neill, in discussion of eighteenth-century improvementsto infrastructure, points out that “even when relativelyshort-term economic returns were what governed pri-vate as well as official action, transport improvementsalways had the further effect of facilitating military sup-ply” (1984, 164). In eighteenth-century Prussia, the con-struction of canal systems, which had obvious economic

benefits, was also explicitly linked to strategic militaryplanning. Frederick the Great explained the logic, statingthat “[t]he advantage of navigation is, however, never tobe neglected, for without this convenience, no army canbe abundantly supplied” (quoted in McNeill 1984, 163).The presence of exploitable economic resources only pro-vided further impetus, as was the case in the constructionof a canal system for the Ruhr River, which was intended toexpand coal production (Henderson 1958; McNeill 1984).In many cases, economically valuable resources evenlower the costs of building infrastructure and movingsettlers into the territory, since there is some economicimpetus to such actions.

Expectation 4a. Consolidation is more likely to be effec-tive when territory contains economicallyvaluable resources.

Expectation 4b. The consolidation of economically valu-able territory decreases the probability thechallenger will militarily escalate a dis-pute.

I expect the presence of a border minority to increasethe effectiveness of target consolidation because it givesthe target a potentially hostile political environment thatcan be favorably altered. Thus, I assume that the minorityis relatively friendly to the challenger or is at least sus-ceptible to mobilization. For example, in 1980, shortlyafter initiating a war with Iran over the status of the Shattal-Arab, Iraq developed contacts with the minority ofroughly 200,000 Arabs in the Iranian province of Khuze-stan. This minority population had aired grievances withboth the new Islamic regime and the previous govern-ment. The Iraqis stoked existing resentments among theArab minority by promising to aid in forming an inde-pendent state in Khuzestan, which the Iraqis rhetoricallyreferred to as Arabistan (Day 1987, 238).

Given the presence of a (potentially) restive minor-ity population, the target can take a variety of actions tosolidify its political position. For instance, the target canimplement a policy of forced emigration for the minor-ity population. A forced emigration policy can removea population from the territory that can potentially aidthe challenger (or hinder the target) militarily. Anotherpossibility is a policy of forced immigration, where thepopulation is infused with large numbers of the majorityethnicity in the target state.

Consolidation efforts can range from extreme casesof attempted genocide (e.g., Bosnia in 1992) to the morecommon example of relatively nonviolent policy deci-sions. An instance of the latter is observable in the disputebetween Italy and Austria over the South Tirol region. In

976 DAVID B. CARTER

October 1957, Italy implemented a legislative policy tobuild 5,000 houses in the city of Bolzano. This policywas seen by German-speaking South Tyrolese and theAustrians as intended to import Italian workers to up-set the ethnic balance (Calvert 2004, 260–62).6 Such apolicy can even be coupled with forced emigration, orincreased monitoring of the minority population. In anyevent, regardless of the exact mechanism, the basic ideais that the presence of the minority group can increasethe effectiveness of political and military consolidationby simply being a problem or potential problem to tryto “solve.” It is important to note that the claim is notthat, all else equal, target states prefer to have an ethnicminority; rather, when such a minority exists it providesa situation that the target may find beneficial to alter.7

Thus, the presence of a minority provides incentives forthe target to consolidate because such efforts can affectthe probability of winning (i.e., [1 − p(m)�(�)]).

Expectation 5a. Territory that contains a minority is morelikely to make consolidation effective.

Expectation 5b. The consolidation of territory that con-tains a minority decreases the probabilitythe challenger will militarily escalate adispute.

Importantly, expectations 3–5 also allow us to as-sess whether the three characteristics discussed above arenot related to the effectiveness of consolidation. If any ofthese characteristics are not significantly related to effec-tive consolidation, then the probability of consolidationwill not increase as the target faces stronger challengerswhen any or all of these characteristics are present. Con-sequently, if expectation 3a, 4a, or 5a is not supported inthe data analysis, then we can conclude this is the case.Relatedly, if territorial characteristics have greater effecton states’ valuations of the disputed territory than on theeffectiveness of consolidation, we would observe that ex-pectations 3b, 4b, and 5b do not find support. Specifically,recall that if strategic location, the existence of a borderminority, and the presence of economically valuable re-sources have a greater effect on the challenger’s valuationof the disputed territory, vc (�), than on the effectiveness

6This is also a strategy that was utilized by the Chinese in theirborder disputes with the Soviet Union (An 1973; Ginsburgs andPinkele 1978; Jones and Kevill 1985).

7Of course, in many cases the costs of consolidation (i.e., k� in thetheoretical model) may be too high to justify. For instance, a policyof emigration intended to alter the demographics of a region canbackfire and create increased problems with the minority. Whilethis is an important possibility, it is fully consistent with the logicof the model.

of consolidation, �(�), then these characteristics will pos-itively affect the probability that the challenger chooseswar. Thus, in the subsequent empirical analysis I assesswhether strategic location, economic value, and the pres-ence of a border minority are primarily ends that statesseek or whether they also significantly affect states’ meansto forcefully obtain disputed territory.

Statistical Analysis of the TerritorialDispute Game

The statistical model in Figure 2 is structurally the sameas the model depicted in Figure 1. Use of a statisticalstrategic model ensures that we capture the strategic na-ture of target consolidation and that we avoid strategicmisspecification issues (Signorino and Yilmaz 2003). Inorder to statistically analyze the strategic logit equivalentof the game, the plausible assumption that both the chal-lenger and target state are boundedly rational is made(Signorino 1999, 2003). Thus, they play with some errorduring the dispute process. Specifically, we use the logitquantal response equilibrium solution concept (LQRE).Nicely, the subgame perfect equilibrium solution conceptused to solve the theoretical model is a special case of theLQRE in which there is no uncertainty. Thus, while theempirical model is structurally consistent with the theo-retical model, it is also more general as it allows for errorsto be made by challenger and target states.

Logic of the Statistical Model

The statistical model captures the idea that the challengerand target make each decision in the game by weighingtheir expected utilities for each possible action. I startfrom the last move in the game, the challenger’s decisionto fight or not, and move up the game tree to show theplayers’ expected utility calculations. For each observa-tion, i = 1 . . . n, the challenger decides whether or not toexercise the military option after observing whether thetarget consolidated its position. If the target consolidates,the challenger makes the following comparison:8

pi,8 = U ∗C (F |C ) > U ∗

C (¬F |C ) (1)

= UC (F |C ) + �8 > UC (¬F |C ) + �7. (2)

8Note that F stands for fight, C for consolidate, and A for activate.The numbers on the � terms correspond to the numbers assignedto the players’ actions in Figures 1 and 2.

THE STRATEGY OF TERRITORIAL CONFLICT 977

FIGURE 2 Specification of the Model with Regressors

Assume the � terms are independent and identically dis-tributed (i.i.d.) Type 1 Extreme Value, which yields

pi,8 = expUC (F |C)

expUC (F |C) + expUC (¬F |C)(3)

pi,7 = 1 − pi,8. (4)

In deciding whether or not to militarize the dispute whenthe target does not consolidate, the challenger makes asimilar comparison which leads to expressions almostidentical to those in equations (1–4). The target makes itsdecision to consolidate or not by comparing, with someerror, its utility for each possible outcome weighted bythe probability the outcome will be realized. If we againassume that the � terms are i.i.d. Type 1 Extreme Value,we obtain

pi,4 =exp( pi,8UT (C,F )+pi,7UT (C,¬F ))

exp( pi,8UT (C,F )+pi,7UT (C,¬F )) + exp( pi,6UT (¬C,F )+pi,5UT (¬C,¬F ))

(5)

pi,3 = 1 − pi,4. (6)

The probability the challenger activates a dispute for agiven observation, pi,2, is expressed similarly. The chal-lenger compares the utility for the status quo, UC (S Q),to the expected utility for activating a dispute, which iscalculated by multiplying each of the four possible postac-tivation utilities with the probability each outcome is re-alized.

Specification of the Model

Specifying the statistical version of the game with regres-sors is no simple task, as identification issues must bedealt with in a theoretically palatable way. In order for themodel to be identified in this context, both the target andchallenger must have the utility for at least one outcomethat is possible at their initial information set and affectstheir utilities normalized to zero.9 Also, no regressor canbe estimated in every utility. The specification for the sta-tistical model is depicted in Figure 2. The players’ utilitiesfor the outcome in which there is a dispute, the target doesnot consolidate, and the challenger does not choose highescalation (UC (¬C¬F ) and UT (¬C¬F )) are normal-ized to zero. Thus, all coefficients are interpreted relativeto the outcome obtained when a dispute is activated, thetarget does not consolidate, and the challenger choosesnot to escalate militarily. This is the most “benign” dis-putatious outcome and is a nice baseline against which tojudge the estimated coefficients.

The data set used was collected by Huth (1996) andincludes 8,328 observations across 129 nondisputes and129 territorial disputes between 1950 and 1990. The gameis modeled as being played once each year a dispute is ac-tive. Huth (1996, 252–55) collected data on the wholepopulation of territorial disputes between 1950 and 1990but randomly sampled 129 of 447 cases of nondisputes.Thus, the data exhibit choice-based sampling, which will

9An initial information set is just the node at which each respectiveplayer makes its first move in the game (Lewis and Schultz 2003).

978 DAVID B. CARTER

introduce bias into the results without correction. Bas(2009) shows that bias from choice-based sampling canbe quite severe in strategic settings. To avoid bias, the like-lihood function is corrected with the weighted exogenoussampling technique introduced by Manski and Lerman(1977).

Regressors

The specification of the challenger’s utility for the statusquo must include the key variables that affect a state’sdecision to activate a dispute or not. Potential challengerstates generally activate territorial disputes that have his-torical roots (Murphy 1990). Thus, to model potentiallyrevisionist states’ utility for the status quo, historical fac-tors that influence its valuation of a potentially disputedpiece of territory are included. The variables used to cap-ture historical context measure whether the challengerexperienced a previous loss of territory to the target, a pre-vious gain of territory from the target, whether there wasa pre-WWII territorial dispute between the two states leftunresolved, a previous settlement that clearly delimitedthe two states’ border, and whether the disputed territoryis a colonial holding.10 I expect states to be more likelyto activate a dispute when territory was previously lost,is a colonial holding, or if there is a history of territorialdisputes (i.e., unresolved dispute). Potential challengerswill be less likely to activate a dispute if they have previ-ously settled boundary issues in a clear and unambigu-ous way or if they have previously gained territory fromthe potential target. Thus, the utility for the status quois:

UC (S Q) = �11.0 + �11.1PreviousSettle

+ �11.2PreviousGain + �11.3PreviousLoss

+ �11.4Colony + �11.5UnresolvedDispute

+ �11.

Recall that the three key territorial characteristics wediscuss are whether the territory is strategically located,contains a minority population with ties to the challenger,and/or is imbued with economic value. Thus, these threevariables are included in both players’ utilities for the mil-itarized outcomes, as each of them are expected to affectthe states’ valuations of the territory as well as the effec-tiveness of target consolidation efforts. Territory is codedas strategically located if at least one of the following sixmilitarily relevant characteristics is present: (1) “it is inclose proximity to major shipping lanes or choke pointsof narrow straits; (2) it is located in close proximity to

10For all regressors, see the article’s supporting information or Huth(1996) for details on coding.

military bases of the challenger; (3) it would provide anoutlet to the sea for an otherwise landlocked country; (4)it was being used as a military base site for the target;(5) it could be used to establish a second military frontagainst the target; or (6) control of disputed territoryblocked the principal route through which a challengercould attack a target” (Huth 1996, 256). Territory is eco-nomically valuable if it contains natural resources thatcan produce export revenues.11 Finally, territory containsa border minority if a minority in the disputed territoryshares language or ethnicity with the largest group in thechallenger state. The military balance, or m in the the-oretical model, is also included in both players’ utilitiesover militarized outcomes as it is an essential componentof the probability the challenger is victorious militarily,p(m).

Several additional variables not explicitly discussedabove that are related to the parameters in the theoreticalmodel are also included. Whether the target is a memberof a deterrent alliance is included in both players’ utilitiesto capture the effects of anticipated military help in theevent that the two states end up fighting over the disputedterritory. Militarily relevant aid to the target from an allyis not captured in the measure of the military balance;thus, including this variable ensures that we capture theeffect such aid may have on the military situation. Addi-tionally, if an ally of the target will contribute help in theevent of an armed conflict, this is likely to make fightingmore costly for the challenger while diverting some costsfor the target. The number of times the target defeated thechallenger is included to both account for rivalry and tocapture the notion that challengers who have previouslyfought the target but still do not possess the disputedterritory will have learned something about target ca-pabilities, fighting prowess, and how costly fighting thetarget is. The time since the target last consolidated is alsoincluded to account for decreasing marginal returns tobuilding more fortifications, solidifying political control,and any other such measure. Thus, if the target has justconsolidated its position, this will undoubtedly affect howeffective further consolidation can be above and beyondwhat was accomplished in the previous year. Thus, theutilities for the militarized outcomes are:12

UC (C F ) = �18.1MilitaryBalance + �18.2StrategicLoc

+ �18.3EconValue + �18.4BorderMinority

+ �18.5DeterrentAlly + �18.6Defeats + �18

11See Huth (1996) for an exact list of qualifying resources.

12Note that I omit the two players’ utilities for fighting when thetarget does not consolidate because their specification is identicalto those shown.

THE STRATEGY OF TERRITORIAL CONFLICT 979

UT (C F ) = �28.1MilitaryBalance + �28.2StrategicLoc

+ �28.3EconValue + �28.4BorderMinority

+ �28.5DeterrentAlly + �28.6Victories

+ �28.7PreviousCons + �28

A constant is estimated for both players’ utilities whenthe target has consolidated but no fighting occurs. Theseconstant terms can be interpreted as the xc and xt termsdiscussed in the theoretical model. Given that these pa-rameters are specified quite generally and can be positiveor negative, representing them with a constant is sensible.

Dependent Variables

In order to estimate the strategic model, data indicatingwhat decision is made at each node by the challenger andtarget are required. The action variables are then usedto create binary variables, di, j , that indicate the outcomej ∈ {1, 5, 6, 7, 8} for each observation i = 1, . . . , n in thedata.13

The target state’s only decision in the game is whetherto consolidate its position in the disputed territory (i.e.,play Consolidate) or not (i.e., play ¬Cons ol idate). Thetarget is coded as playing Consolidate if it creates militaryestablishments in the disputed territory, actively pursuesthe extraction or development of valuable natural re-sources from the disputed territory, or takes legislativeor constitutional action to either further incorporate theterritory into the target or hinder any future attempts bythe challenger to take the disputed territory (Huth 1996,259–60).

There are three dependent variables for challengeractions. The first variable is a binary variable that indi-cates whether a dispute was activated or kept active in agiven year. The second indicates whether the challengerstate chooses to militarize a dispute after the target con-solidates, while the third indicates the challenger’s actionif the target does not consolidate. The challenger is codedas not fighting following the target’s decision if it chooseslow- or moderate-level diplomatic conflict. In contrast, ifthe challenger confronts the target militarily with crediblethreats and/or the actual use of military force, it is codedas having chosen to fight.14

13The j correspond to the action probabilities depicted in Figure 2.For example, j = 1 indicates that the challenger decides not to acti-vate a dispute, just as p1 is the probability the challenger chooses notto activate a dispute. Note that outcomes are a result of challengerand target decisions.

14The decision to fight is essentially equivalent to a decision to starta war, as war breaks out in the data 95% of the time when it ischosen.

I combine the target and challenger action variablesto create a dependent variable, di, j , that takes a value of 1if a given terminal node j is reached in observation i and 0otherwise. Thus, if a dispute is active and the target playsConsolidate while the challenger plays Fight in observa-tion i, then di,8 = 1, while all other terminal node dummyvariables would be given a value of 0 for this observation.Given dummy variables di, j and a probability pi, j for eachterminal node, the log-likelihood to be maximized is:

ln(L ) =n∑

i=1

�No Disputedi,1ln( pi,1) + �Disputedi,5ln( pi,5)

+ �Disputedi,6ln( pi,6) + �Disputedi,7ln( pi,7)

+ �Disputedi,8ln( pi,8).

The probabilities correspond to the probabilities esti-mated for each observation as shown in equations (1–6).The � terms are weights to correct for the choice-basedsample. The � terms ensure that the imbalance in thesample between nondispute and dispute cases is explic-itly taken into account in estimation of the coefficients.

Results

The regression results in Table 2 imply considerable sup-port for several of the key expectations derived from thetheoretical model. To account for a variety of potential is-sues with the standard errors, I bootstrap with samplingclustered by dyad. A known issue with the standard er-rors is bias induced from the use of a choice-based sample,while other possible issues are a violation of independenceacross within-group observations. The average bootstrapsample size is 7,800 across 1,000 iterations.15 Since thechallenger and target play the game by figuring out eachothers’ moves via backwards induction, the results arepresented moving “up the game tree.”

Challenger Escalation Utilities

Unsurprisingly, the balance of military forces is posi-tive and significant regardless of whether the target hasconsolidated or not. Thus, more powerful challengers gethigher utility from fighting relative to not fighting inthe no consolidation case. This result is what we wouldexpect given expectation 1, as possessing more militaryforces relative to the target should increase the benefits ofmilitary confrontation. The presence of a deterrent target

15The sample size is an average because the number of observationsacross groups differs according to the number of years a disputebetween two states is active.

980 DAVID B. CARTER

TABLE 2 Player Utilities for Statistical Strategic Model

UC (F ) UC (C¬F ) UC (C F ) UT (F ) UT (C¬F ) UT (C F ) UC (S Q)

Constant −2.189∗∗ 13.174∗∗ 24.475∗∗ 0.626 2.995∗∗

(0.122) (0.797) (6.481) (0.418) (0.166)Military Balance 1.203∗∗ 11.929∗∗ 21.935∗∗ −4.856∗∗

(0.223) (2.169) (7.729) (2.435)Strategic Location −1.077∗∗ −1.290∗ 6.584 4.120∗∗

(0.132) (0.747) (4.499) (1.823)Economic Value −0.819∗∗ 3.972∗∗ 18.445∗∗ 7.340∗∗

(0.239) (1.354) (7.345) (3.151)Border Minority 0.155 2.325 −4.525 −2.681∗∗

(0.133) (1.571) (3.696) (1.075)Deterrent Alliance −0.925∗∗ −0.921 −6.772 −1.381

(0.192) (1.234) (7.394) (1.656)Target Victories −0.082 2.173∗∗ −13.394∗∗ 0.215

(0.070) (0.522) (2.050) (0.744)Time Since Consolidation 0.644∗∗ −0.119∗∗

(0.070) (0.066)Previous Settlement 3.330∗∗

(0.211)Previous Loss −1.076∗∗

(0.193)Previous Gain 1.815∗∗

(0.196)Decolonization Norm −0.673∗∗

(0.223)Unresolved Dispute −2.863∗∗

(0.313)

Number of Observations: 8,328.Log-Likelihood: 2105.458.88.070% Observations Correctly Predicted.Bootstrapped Standard Errors in Parentheses.∗∗ Indicates Significance at .05 Level; ∗ Indicates Significance at 0.10 Level.

alliance decreases challenger payoffs for both militarizedoutcomes, although it is not significant if the target hasconsolidated. Previous military defeats at the hands of thetarget do not have any significant effect when the targethas not consolidated. However, when the target consol-idates its position, the challenger derives positive utilityfrom escalation of the dispute.

The existence of ethnic brethren does not signifi-cantly affect the challenger’s utility for fighting regardlessof whether the target has consolidated. The coefficientsare both positive, which suggests that the presence of aminority is not related to effective target consolidation assuggested by expectation 5b, or at least implies that anyeffects from consolidation are washed out by its influenceon the challenger’s valuation of the territory. The insignif-icance of ethnic brethren for the challenger’s decision tofight contrasts sharply with Huth’s finding that a border

minority is the most substantively significant factor thatincreases the probability a challenger uses force (Huth1996, 107–9).

The presence of economic resources has a positive ef-fect on the challenger’s propensity to militarize a disputewhen the target has consolidated, although it negativelyaffects the challenger’s utility for military escalation inthe absence of consolidation. Thus, counter to expec-tation 4b the development of economic resources doesnot seem to be related to effective consolidation; rather,the coefficients suggest that the impact upon the chal-lenger’s valuation of the territory outweighs any changein the military situation. At first glance, the fact that thechallenger’s propensity to militarily contest territory ofeconomic value increases when the target consolidatesand decreases when the target does not consolidate ispuzzling. This finding suggests that the consolidation of

THE STRATEGY OF TERRITORIAL CONFLICT 981

economically valuable territory (in the absence of otherterritorial characteristics such as strategic location) mayincrease the value of the territory in addition to strength-ening the target’s hold over it. For economically valu-able territory this makes sense, as the value of natural re-sources is in large part a function of their development.16

The finding that the presence of economically valuableresources decreases the challenger’s propensity to fightwhen the target does not consolidate is consistent withthe findings of Huth (1996). Huth argues that challengersare reluctant to escalate when disputed territory is eco-nomically valuable because joint gains from trade are lost(which is thus how they affect vc (�)). Interestingly, thisis unlikely to be a concern if the target has already actedaggressively in consolidating the valuable resources.

Strategically located territory has a negative and sig-nificant effect on challenger utilities for both violent out-comes. This finding indicates that when the target’s strate-gic consolidation decision is modeled, challengers do notseem to pursue strategically located territory more vigor-ously, which is the opposite effect found by Huth (1996)and Huth and Allee (2002). Neither Huth (1996), Huthand Allee (2002), nor any other work that examines ter-ritorial disputes models the strategic choices available tothe target to use territory to improve its military standing.Not modeling target actions ensures that prior studies donot pick up whether consolidated territory that is strategi-cally located augments target capabilities enough to deterchallengers from militarily contesting it, which is whatI find here. While these results are very promising andconsistent with expectation 3b, examination of predictedprobabilities and target utilities is necessary to fully un-derstand the relationship between strategic territory andthe challenger’s escalation decision.

The top two graphs in Figure 3 (i.e., Figures 3(a)and 3(b)) substantiate the support for expectations 1 and3b. The graph in Figure 3(a) shows that when the targetdoes not consolidate, the probability that the challengerchooses military escalation is increasing in its strength,although the probability decreases if the territory is strate-gically located or economically valuable and barely in-

16This is a mechanism that is unlikely to be present for either ofthe other two characteristics but is quite plausible here. Currently,I assume in the theoretical model that the value of territory is in-creasing in its valuable characteristics. If I also allow the value ofterritory to possibly increase further as a result of target consoli-dation efforts, the character of our expectations does not change.Rather, allowing the challenger’s valuation to increase as a func-tion of characteristics as well as the target’s consolidation of thesecharacteristics provides another way for a characteristic such aseconomic value to have a greater effect on the challenger’s valua-tion than it does on the effectiveness of target consolidation effortsbut does not change the existing expectations.

creases if there is a border minority. Figure 3(b) dis-plays the probability that challengers of differing relativestrength choose military escalation when the target hasconsolidated its position in the disputed territory. It showsthat the probability the challenger fights is essentially zerountil it is overwhelmingly powerful if the territory hasno relevant characteristics. Interestingly, the presence ofeconomic resources greatly increases the propensity ofstrong challengers to resort to military means in resolu-tion of disputes. This substantiates the finding above thatthe consolidation of economically valuable territory ac-tually increases its value in ways that offset any securitygains for the target. However, if the territory is also strate-gically located, the probability that the challenger resortsto military means plummets and lingers close to 0, evenwhen challengers are overwhelmingly powerful. Thus, thenegative effect of strategic location (i.e., expectation 3b)overpowers the positive effect of economic value.

In general, Figure 3 shows that target consolidationsignificantly alters the strategic calculus of the challenger,as escalation is more clearly associated with greater mili-tary strength than when the target does not consolidate.Strikingly different dynamics for military capabilities af-ter target consolidation suggests that consolidation is sig-nificantly changing the military calculus of both states asthe theory suggests. Also, the existence of strategic ter-ritory always lowers the probability that the challengerhighly escalates when the target has consolidated, whichprovides considerable support for the notion that con-solidation is more effective when territory is strategicallylocated.

The changes in probability associated with each re-gressor shown in the first four columns of Table 3 affirmthe observation that when targets consolidate, challengersare significantly less likely to attempt favorable resolutionof disputes militarily. Comparison of the baseline cases,in which all regressors are held at their median values, isparticularly telling, as the probability of military escala-tion is almost zero postconsolidation while it is 0.166 inthe absence of target consolidation.

Target Escalation Utilities

At this stage, the target has observed challenger activationof a dispute and now must decide whether to consoli-date conditional on the expected escalation decision ofthe challenger. Interpretation of the coefficients here isdifficult, as the effect of any variable is also influenced byits effect on challenger utilities. For this reason, predictedprobabilities and graphical illustrations are necessary toassess the meaning of the results.

982 DAVID B. CARTER

FIGURE 3 Predicted Probabilities When a Dispute Is Active

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.2

0.4

0.6

0.8

1.0

Military Balance

Pro

babili

ty C

Fig

hts

Giv

en N

o C

onsolid

ation

Nothing

Strategically Important

Economic Value

Border Minority

(a) Effect of Strategic Location After No Consolidation

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.2

0.4

0.6

0.8

1.0

Military Balance

Pro

babili

ty C

Fig

hts

Giv

en T

Consolid

ation

Nothing

Strategic Location & Economic Value

Economic Value

(b) Effect of Strategic Location After Consolidation

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.2

0.4

0.6

0.8

1.0

Military Balance

Pro

babili

ty T

Will

Consolid

ate

Giv

en D

ispute

Nothing

Strategic Location

Border Minority

Strategic Location & Border Minority

(c) Effects of Strategic Location and Border Minority

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.2

0.4

0.6

0.8

1.0

Military Balance

Pro

babili

ty T

Will

Consolid

ate

Giv

en D

ispute

Nothing

Economic Value

Economic Value & Strategic Location

(d) Effect of Strategic Location with Economic Value

The last two columns of Table 3 provide a goodoverview of the effect of each variable on the target’sconsolidation decision. The substantive effects for eachvariable are calculated holding all other variables at theirmedian value, where the first row is a baseline case withall variables at their median. First, notice that of the threevariables hypothesized to increase the effectiveness of tar-get consolidation, only strategic location has a particu-larly strong substantive effect. Both economic value andborder minority have positive effects, but these effectsare comparatively small. The number of previous victo-

ries over the challenger also has large positive substantiveeffects on the target’s decision to consolidate.

The two graphs in the second row of Figure 3 (i.e.,Figures 3(c) and 3(d)) illustrate nicely that strategicallylocated territory has significant and positive influenceon the effectiveness of consolidation. Figures 3(c) and3(d) show that when the disputed territory does notpossess any economic value, strategic importance, orethnic brethren of the challenger, the probability thatthe target consolidates is generally low and decreasesslightly in challenger strength. This is true even when the

THE STRATEGY OF TERRITORIAL CONFLICT 983

TABLE 3 Change in Probability Associated with Regressors

Challenger Probabilities Target Probabilities

Pr(F ight|¬Cons) � in Pr Pr(F ight|Cons) � in Pr Pr(Consolidate) � in Pr

Median Values 0.166 NA 0.001 NA 0.036 NAMilitary Balance = .9 0.249 +0.083 0.080 +0.079 0.030 −0.006Military Balance = .1 0.112 −0.044 0.000 −0.001 0.049 +0.013Strategic Location = 1 0.064 −0.102 0.000 −0.001 0.215 +0.179Economic Value = 1 0.081 −0.085 0.030 +0.029 0.063 +0.027Border Minority = 1 0.046 +0.010Deterrent Ally = 1 0.073 −0.093 0.352 +0.316Challenger Defeats = 1 0.005 +0.004 0.272 +0.236

territory is economically valuable and/or contains a mi-nority, which confirms the weak effects found for thesetwo variables in Table 3. At this point, these weak effectsfor economic value and border minority are encouragingbecause they lend overall coherence to the results of thestatistical model. The findings for the challenger implythat strategic territory is associated with effective consol-idation, while economic value and border minority donot have such effects. Thus, these findings increase ourconfidence that the expectations derived from the the-oretical model present a meaningful characterization ofthe strategic situation facing disputants.

In Figure 3(c), the probability of consolidation in-creases considerably when the disputed territory is strate-gically located, although the probability still decreasesslightly in challenger strength for most of the range of themilitary balance. Figure 3(d) demonstrates that when ter-ritory is both economically valuable and strategically im-portant, the probability of target consolidation not onlyincreases, but increases markedly in challenger strength.If the majority of the population in the territory is alsoof the same ethnicity as the dominant one in the chal-lenger, a similar trend is shown in Figure 3(c), althoughthe probability decreases somewhat after the challengercontrols roughly 80% of total dyadic capabilities. Theseresults, taken in tandem with the results for the challenger,strongly suggest that strategic territory facilitates effectiveconsolidation. Additionally, they imply that while strate-gic location is the key factor, the presence of economicallyvaluable resources and/or a border minority can furtherincrease the effectiveness of consolidation. When strategiclocation is present along with one or more of these otherfactors, the probability generally increases in challengerstrength once targets are paired with challengers strongerthan themselves.

In sum, examination of target utilities and actionprobabilities provides powerful support for several of the

theoretical expectations. In particular, expectation 2 inconjunction with expectation 3a found strong support.In contrast, economically valuable territory and territorywith a border minority did not have strong individualeffects on target consolidation. Thus, the territorial char-acteristic most directly related to the military situation,strategic location, had the expected effects, while the moreindirect mechanisms via the development of economic re-sources and the political consolidation of territory with aborder minority did not have strong independent effectson target consolidation.

Challenger Utility for Activation

All of the coefficients in the challenger’s utility for thestatus quo work in the expected way and are statisticallysignificant. If a challenger has previously settled its bor-der issues with the target unambiguously, this exerts apositive influence on UC (S Q) relative to UC (¬F ). Hav-ing previously experienced a net gain of territory fromthe target also has a positive effect. Furthermore, poten-tial challengers that have either signed agreements clearlydelimiting their borders with potential targets or expe-rienced a net gain of territory from the potential targetgenerally prefer the status quo to the activation of a dis-pute. Conversely, previously losing territory to the targetstate has a negative effect on the potential challenger’sutility for the status quo; thus, states that have lost terri-tory to their potential target generally prefer to challengethe status quo. If the target is a colonial power, this alsounsurprisingly exerts a negative effect.

The Sino-Vietnamese Dispute

The dispute between Vietnam (Democratic Republic ofVietnam) and China discussed briefly in the introduction

984 DAVID B. CARTER

nicely illustrates some of the key aspects of the theoreticalmodel that found support in the empirical analysis.17 Thisis an especially interesting case because Vietnam, the tar-get of military action in early 1979, is much weaker thanChina. Additionally, the disputed territory was strategi-cally located and also contained a border minority ofethnic Chinese. Thus, the model would predict (and theresults of the empirical analysis would suggest) that Viet-nam undertake consolidation if it thought that Chinawould attack militarily. Indeed, Vietnam undertook ex-tensive consolidation efforts in the territory that had realmilitary consequences.

Although active, the dispute was relatively quietthroughout the early 1970s as the Vietnamese were preoc-cupied with the fight against the United States while Chinaplayed an important role as benefactor in this effort. Afterthe conflict in South Vietnam ended, the territorial dis-pute became increasingly contentious in each year after1975. Throughout 1978, the Vietnamese became increas-ingly aggressive in the border region and consolidatedtheir position. They consolidated in two ways. First, theyattempted to solidify their military and political controlover the region through expulsion of thousands of eth-nic Chinese. By the end of 1978, approximately 200,000ethnic Chinese were expelled from the Sino-Vietnameseborder region (Chang 1985; Chen 1987). The Vietnameseundertook this massive forced emigration not only to con-solidate their position in the disputed border region, butalso to destabilize Chinese control over its border region,which was quite successful (Chang 1985; Chen 1987). Ad-ditionally, the Vietnamese built extensive border defenseinstallations that were meant to present the Chinese withan “impregnable fortress” (Chen 1987, 107). It is impor-tant to note that the mountainous border region was ofparticular strategic value, as it had historically providedthe gateway to control of the entire Red River Delta for in-vaders. The defensive structures built across the mountainrange were “characterized by tunnels, caves, and trenchesmanned by well-trained and well-armed militia” (Chen1987, 107). The Vietnamese had also planted numerousland mines, booby traps, and anything that generally in-creased the difficulty of the already difficult terrain.

On February 17, 1979, Chinese forces launched amassive attack that was stated to be in retaliation for nu-merous Vietnamese-instigated border incidents.18 Eventhough China initiated attacks with around 100,000

17The two states disputed three pieces of territory in total: theirnearly 800-mile border, the Gulf of Tonkin, and the Paracel andSpratly Islands.

18Some analysts such as Ross (1988) argue that the Chinese attackwas actually in response to Vietnamese actions in Cambodia andnot a direct consequence of territorial issues. While this view is not

troops (a third of all troops massed at the border) againstlocal Vietnamese forces and militias, they met with con-siderable difficulty (Womack 2006). China was eventuallyable to take the important strategic border city of LangSon on March 4 after several weeks of tough and surpris-ingly costly fighting. However, the immense difficultiesencountered in fighting the Vietnamese made the Chi-nese even more eager to present the fall of Lang Son asbeing a symbolic victory important enough to lead to ahalt in fighting (Duiker 1986, 87). It should be noted thatthe fall of Lang Son came after two bloody weeks thatcost the Chinese 25,000 troops to the Vietnamese loss of20,000 (Womack 2006, 200). This is particularly strikinggiven that the Chinese were not even fighting Vietnameseregulars; rather, the DRV’s elite troops were either fight-ing in Cambodia or stationed around Hanoi. Thus, theVietnamese defenses had performed quite well despite thefact that local forces were used and the DRV’s best troopswere not even involved.

This case nicely demonstrates how strategic consol-idation of strategic territory was quite beneficial for theVietnamese against their considerably stronger Chinesecounterpart. The forced emigration of ethnic Chinesestabilized the political and military situation for the Viet-namese and simultaneously fostered chaos and disorderon the Chinese side of the border. Additionally, Viet-namese consolidation of their military defenses madeChinese incursions very difficult, making clear that Chinawould not be able to easily defeat and occupy Vietnam. Itis also clear that the defensive fortifications made fight-ing easier for the Vietnamese, as they were able to defendquite well against a superior power without even usingtheir best troops.19

It is clear from this discussion that the model shouldpredict that Vietnam consolidate and China fight (i.e.,row 1 of Table 1). Since the Sino-Vietnamese dispute isa case in our data, we now examine the predictions ofthe model given the data. Specifically, we examine themodel’s predictions in 1979, the year discussed above.Given that the model makes rather clear predictions with

wholly implausible given that all analysts acknowledge that Cam-bodia was an important factor, it is not of much consequence for mypurposes. I am interested primarily in the impact that Vietnameseconsolidation of strategic territory had on thwarting attacks froma stronger challenging enemy. In any event, I think Ross’s analy-sis goes too far in putting Cambodia front and center given thatthe Chinese and Vietnamese had been having militarized incidentsalong the border since as early as 1974.

19The Chinese asserted that they were just punishing the Viet-namese for numerous border violations and did not intend topenetrate deeply into Vietnam. However, the tough fighting con-ditions ensured that this was the only sensible course of action forChina.

THE STRATEGY OF TERRITORIAL CONFLICT 985

FIGURE 4 Predictions for Sino-Vietnamese Dispute

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.2

0.4

0.6

0.8

1.0

Military Balance

Pro

babili

ty V

ietn

am

Consolid

ate

s G

iven A

ctive D

ispute Strategic & Border Minority

Not Strategic

Actual Prediction

0.0 0.2 0.4 0.6 0.8 1.0

0.0

0.2

0.4

0.6

0.8

1.0

Military Balance

Pro

babili

ty C

hin

a F

ights

Giv

en V

ietn

am

ese C

onsolid

ation

Strategic & Border Minority

Not Strategic

Actual Prediction

regard to this case and that our discussion of this caseindicates that it should fit the patterns found in the data,this is a nice way to assess the face validity of the empiricalmodel. First, we examine predicted Vietnamese behaviorin 1979, holding all variables at their actual values exceptmilitary balance. The dashed line in the leftmost graphin Figure 4 depicts the probability Vietnam consolidatesacross the entire range of the military balance, holdingall other variables at their actual values. Given that theactual value of the military balance is 0.78 (i.e., the “x” inthe graph), the model predicts that Vietnam consolidatedits position with probability 0.87, a very strong predic-tion that is well above the typical threshold of 0.50. Ifwe entertain the counterfactual that the territory is notstrategic, as depicted by the solid line, we see a starkly dif-ferent relationship. If the territory was not strategic, themodel predicts that consolidation is decreasingly likelyas China gains in strength relative to Vietnam. For theactual value of military balance, the model predicts Viet-namese consolidation with probability 0.25, which is anull prediction.

The empirical model also performs well in predictingChinese behavior. The dashed line again represents pre-dictions across the entire range of military balance withall regressors at their actual values. Note that the modelpredicts China would attack with probability 0.55, whichis above the 0.50 threshold. As depicted by the solid line,if the territory was not strategic the model predicts thatChina fights with probability 0.82 at the actual value ofthe military balance. Thus, the strategic location of theterritory depresses the probability China fights consider-ably, although not enough to lead to a null prediction.

In sum, the empirical model makes the correct predic-tions for both Vietnam and China. In conjunction withthe qualitative evidence discussed above, this providesa quite compelling explanation of what occurred in theSino-Vietnamese war fought in 1979.

Conclusion

Numerous historical examples corroborate the findingthat the relatively weak augment their capabilities throughthe consolidation of territory. Herodotus, referred to bymany as the “father of history,” notes a similar dynamicin the war between the Phocians and the stronger Thes-salians of ancient Greece. In an important engagement,the Phocians took advantage of the location of strategi-cally significant mountain passes that Thessalian forcesneeded to advance. In order to thwart the advance ofThessalian cavalry, they dug a large trench in the passand filled it with empty wine bottles, subsequently re-covering the trench to camouflage it. The Thessalian cav-alry advanced quickly upon seeing the Phocian forcesand became messily entangled in the trench, which broketheir horses’ legs and made them susceptible to ambush(Asprey 1975, 32). Of course, without a strategically lo-cated mountain pass, the Phocians would have likelyfound it exceedingly difficult to effectively consolidatetheir position in such a cunning manner. This rather il-lustrious case shows how the use of territory to augmentcapabilities has been effective for centuries despite beingabsent from quantitative studies.

986 DAVID B. CARTER

I generate a new set of hypotheses about how militarycapabilities and territorial characteristics (e.g., strategiclocation) affect state behavior in territorial disputes byfocusing on the role territory plays in disputes. Ratherthan simply defining how much states value territory,characteristics also affect states’ abilities to obtain dis-puted territory through armed conflict. Territorial char-acteristics affect states’ means to obtain territory by forcebecause they define how effective target state consolida-tion is in reshaping the conventional military situation. Byincorporating this insight into a game-theoretic model, Icome to the conclusion that relatively weaker target statesare more likely to consolidate their position when terri-torial characteristics imply that consolidation will signif-icantly change the military environment. I hypothesizethat consolidation is effective when territory holds im-portant characteristics such as strategic location. Usinga statistical estimator that is the structural equivalent ofthe theoretical model, I find that the consolidation of ter-ritory is a significant source of power, especially when itis strategically located. The results for strategic territoryand disputant behavior vis-a-vis the balance of militarycapabilities represent new and interesting findings thatare at odds with the findings of previous studies by Huth(1996), Huth and Allee (2002), and Walter (2006), butconsistent with numerous historical cases ranging fromthe Phocians of ancient Greece to the Vietnamese in thetwentieth century.

References

An, Tai Sung. 1973. The Sino-Soviet Territorial Dispute. Philadel-phia, PA: Westminster.

Asprey, Robert B. 1975. War in the Shadows: The Guerilla inHistory, Volume I . Garden City, NY: Doubleday & Company.

Bas, Muhammet A. 2009. “Democratic Inefficiency: RegimeType and the Tendency to Make Foreign Policy Mistakes.”Unpublished manuscript. Harvard University.

Calvert, Peter, ed. 2004. Border and Territorial Disputes of theWorld. London: John Harper.

Carter, David B., and H. E. Goemans. 2010. “The Making of theTerritorial Order: New Borders and the Emergence of Inter-state Conflict.” International Organization. Forthcoming.

Chang, Pao-min. 1985. The Sino-Vietnamese Territorial Dispute.New York: Praeger.

Chen, King C. 1987. China’s War with Vietnam, 1979. Stanford,CA: Hoover Institution Press.

Day, Alan J. 1987. Border and Territorial Disputes. 2nd ed. Essex,UK: Longman.

Duiker, William J. 1986. China and Vietnam: The Roots of Con-flict . Berkeley, CA: Institute of East Asian Studies.

Fazal, Tanisha M. 2004. “State Death in the International Sys-tem.” International Organization 58(2): 311–44.

Fearon, James D. 1994. “Domestic Political Audiences and theEscalation of International Disputes.” American Political Sci-ence Review 88(3): 577–92.

Fearon, James D. 1995a. “Bargaining Over Objects That Influ-ence Future Bargaining Power.” Unpublished manuscript.University of Chicago.

Fearon, James D. 1995b. “Rationalist Explanations for War.”International Organization 49(3): 379–414.

Ginsburgs, George, and Carl F. Pinkele. 1978. The Sino-SovietTerritorial Dispute, 1949–64. New York: Praeger.

Goertz, Gary, and Paul Diehl. 1992. Territorial Changes andInternational Conflict . London: Routledge.

Henderson, W. O. 1958. The State and the Industrial Revolutionin Prussia. Liverpool, UK: Liverpool.

Hensel, Paul R. 2000. “Territory: Theory and Evidence on Ge-ography and Conflict.” In What Do We Know about War?ed. John Vasquez. Lanham, MD: Rowman and Littlefield,57–64.

Hill, Norman. 1945. Claims to Territory in International Lawand Relations. New York: Oxford University Press.

Holsti, Kalevi J. 1991. Peace and War: Armed Conflicts andInternational Order. New York: Cambridge UniversityPress.

Huth, Paul K. 1996. Standing Your Ground. Ann Arbor: Univer-sity of Michigan Press.

Huth, Paul K., and Todd L. Allee. 2002. The Democratic Peaceand Territorial Conflict in the Twentieth Century. Cambridge:Cambridge University Press.

Jones, Peter, and Sian Kevill, eds. 1985. China and the SovietUnion, 1949–1984. New York: Facts On File.

Kocs, Stephen A. 1995. “Territorial Disputes and Interstate War,1945–1987.” Journal of Politics 57(1): 159–75.

Lewis, Jeffrey B., and Kenneth A. Schultz. 2003. “Reveal-ing Preferences: Empirical Estimation of a Crisis Bargain-ing Game with Incomplete Information.” Political Analysis11(4): 345–67.

Luard, Evan. 1986. War in International Society. London: I.B.Tauris and Company.

Manski, Charles F., and Steven R. Lerman. 1977. “The Estima-tion of Choice Probabilities from Choice-Based Samples.”Econometrica 45(8): 1977–88.

Mas-Colell, Andreu, Michael Whinston, and Jerry Green.1995. Microeconomic Theory. New York: Oxford UniversityPress.

McNeill, William. 1984. The Pursuit of Power: Technology, ArmedForce, and Society since A.D. 1000. Chicago: University ofChicago Press.

Murphy, Alexander B. 1990. “Historical Justifications for Ter-ritorial Claims.” Annals of the Association of American Geog-raphers 80(4): 531–48.

Powell, Robert. 2006. “War as a Commitment Problem.” Inter-national Organization 60(1): 169–203.

Ratner, Steven R. 1996. “Drawing a Better Line: UTI Possidetisand the Borders of New States.” The American Journal ofInternational Law 90(4): 590–624.

Ross, Robert S. 1988. The Indochina Tangle: China’s Viet-nam Policy, 1975–1979. New York: Columbia UniversityPress.

THE STRATEGY OF TERRITORIAL CONFLICT 987

Sartori, Anne E. 2003. “The Might of the Pen: A ReputationalTheory of Communication in International Disputes.” In-ternational Organization 56(1): 121–49.

Senese, Paul D. 2005. “Territory, Contiguity, and InternationalConflict: Assessing a New Joint Explanation.” American Jour-nal of Political Science 49(4): 769–79.

Signorino, Curtis S. 1999. “Strategic Interaction and the Statis-tical Analysis of International Conflict.” American PoliticalScience Review 93(2): 279–97.

Signorino, Curtis S. 2003. “Structure and Uncertainty in Dis-crete Choice Models.” Political Analysis 11(4): 316–44.

Signorino, Curtis S., and Kuzey Yilmaz. 2003. “Strategic Mis-specification in Regression Models.” American Journal ofPolitical Science 47(3): 551–66.

Smith, Alastair. 1998. “Fighting Battles, Winning Wars.” Journalof Conflict Resolution 42(3): 301–20.

Strachan, Hew. 1983. European Armies and the Conduct of War.London: George Allen & Unwin.

Vasquez, John A. 1995. “Why Do Neighbors Fight? Proxim-ity, Interaction, or Territoriality.” Journal of Peace Research32(3): 277–93.

Wagner, R. Harrison. 2000. “Bargaining and War.” AmericanJournal of Political Science 44(3): 469–84.

Wagner, R. Harrison. 2007. War and the State: The Theory ofInternational Politics. Ann Arbor: University of MichiganPress.

Walter, Barbara F. 2006. “War as a Reputation Problem: Explain-ing Government Responses to Self-Determination Move-ments.” American Journal of Political Science 50(2): 313–30.

Womack, Brantly. 2006. China and Vietnam: The Politics ofAsymmetry. Cambridge: Cambridge University Press.