Abstract
While no two democratic states have fought an interstate war against each other, democratic dyads experience militarized disputes with some frequency. Previous research suggests that a large percentage of militarized disputes between two democracies involve fishing and oil resources of the sea. Yet this research selects on cases where militarized conflict occurs, and fails to consider whether democracies have more frequent diplomatic conflicts over maritime areas relative to other regime pairings. Analyzing data from the Issue Correlates of War project, which includes diplomatic conflicts over maritime areas (1900–2007) in the Americas, Europe, Middle East, and Asia, this study finds that pairs of democracies have the highest chance of experiencing diplomatic maritime disputes among all pairs of countries in the same region or dyads involving major powers. Three theoretical explanations were empirically evaluated to account for this pattern: (a) greater opportunities for democratic maritime conflicts given higher levels of economic productivity and the sizes of fishing fleets in democratic states, (b) the increasing securitization of maritime issues, especially after the terrorist attacks of September 2001, and (c) variations in the number of democracies across regional contexts. Several illustrative case studies for each theoretical argument are presented. The authors discuss the implications of these findings for the democratic peace literature and the law of the sea regime.
In 1971, Canada and Denmark began three years of bilateral negotiations to discuss resource usage in the continental shelf area between Canada and Greenland, especially salmon fishing rights in the Northwest Atlantic maritime area (Degenhardt, 1991). The countries could not strike an agreement over their competing maritime claims. In 1977, Denmark issued a 200 nautical mile (nm) exclusive economic zone (EEZ) claim around Greenland that overlapped extensively with Canadian claims and threatened Canadian resource management in the region (New York Times, 1977). Since the 1970s, both countries have made yearly concessions on fishing allowances to help maintain the salmon population in the Grand Banks area, but the contested EEZ claim and the related territorial dispute around Hans Island continue to this day (Wall Street Journal, 2012).
This conflict between two fully democratic countries may seem surprising in light of the extensive empirical findings showing generally peaceful relations between pairs of democracies (Reiter, 2012; Russett and Oneal, 2001). While fishing and EEZ disputes are not typically as contentious as other geopolitical issues such as conflicts over land borders, issues involving maritime areas regularly involve the threat or use of force (Hensel et al., 2008). For example, in another maritime conflict involving Canada in the 1990s, the Canadian government authorized its military to fire upon a Spanish trawler, the Estai, a vessel that was fishing in contested areas. The Canadian government had passed domestic legislation to protect its fishing areas with force if necessary, a law that was repealed through negotiations brokered by the United Kingdom (UK) and the European Union (EU) in 1995 (Song, 1997).
These patterns of democracies contending over maritime areas are also observed in larger samples. In a study of all militarized disputes between two democratic countries, Mitchell and Prins (1999) found that 43% of the 97 dyadic disputes between democracies 1 from 1946–1992 involved fishing, oil, and mineral resources. While this study sheds light on the importance of maritime issues for democratic militarized conflicts, we know less about the prevalence of diplomatic conflicts over maritime spaces in general. Given opportunities for maritime disagreements with neighboring states, are pairs of democracies more likely to engage in such conflicts than pairs of non-democracies or mixed dyads involving one democracy and one non-democracy? What factors help to account for the prevalence of maritime issues in relations between democratic states?
Analyzing data from the Issue Correlates of War (COW) project (Hensel et al., 2008), which includes diplomatic conflicts over maritime areas (1900–2007) in the Americas, Europe, Middle East, and Asia, we show that pairs of democracies have the highest chance of having maritime disputes among all pairs of countries (or dyads) with opportunities for maritime claims. We posit three theoretical explanations to account for this pattern: (a) greater opportunities for democratic maritime conflicts given higher levels of economic productivity and the sizes of fishing fleets in democratic states, (b) the increasing securitization of maritime issues, especially after the terrorist attacks of September 2001, and (c) variations in the number of democracies across regional contexts. Large-N dyadic analyses are presented along with several illustrative case studies for each theoretical argument. Consistent with our economic story, we find that major powers and highly developed states are more likely to experience maritime claims than less developed states. We also find that relative economic capabilities matter; asymmetric energy production in a dyad reduces the chances of conflict. Our security hypothesis finds support as well; pairs of countries experiencing militarized conflicts are more likely to contest each other’s maritime spaces than peaceful dyads. However, we do not observe democracies increasing claims to maritime spaces in the post-2001 period. In contrast, the propensity for maritime conflicts has declined, especially in democratic dyads if both states have ratified the United Nations Convention on the Law of the Sea (UNCLOS) treaty. Thus increasing maritime security issues in the post-2001 environment have not translated into more diplomatic claims over maritime zones. Finally, we show regional variations in these patterns, with joint democracies most likely to fight over maritime resources in the most democratic regions (Europe and the Americas). We conclude by discussing the implications of these findings for the democratic peace and conflict literatures and the law of the sea regime more broadly.
Regime type and contentious issues
While related to each other, the democratic peace and issue-based literatures in the study of interstate conflict have not typically been connected theoretically or empirically. The democratic peace literature focuses on how the characteristics of states’ domestic political institutions influence their foreign policy behavior and interactions with other states. When considering interactions between pairs of states, empirical studies show an absence of wars in democratic dyads (Bremer, 1992) and significantly reduced chances of militarized conflict short of war (Maoz and Russett, 1993; Russett and Oneal, 2001). Theoretical explanations for the dyadic democratic peace focus on structures/institutions (Bueno de Mesquita et al., 2003; Huth and Allee, 2002; Russett, 1993), information (Schultz, 1999), norms (Dixon, 1994; Maoz and Russett, 1993; Mitchell, 2002; Raymond, 1994), satisfaction with the systemic status quo (Lemke and Reed, 1996), and economic interdependence (Barbieri, 1996; Hegre, 2014; Mousseau, 2000). It is typically assumed that the underlying issues at stake in a diplomatic disagreement are not essential for predicting war or peace, but rather that institutional qualities have a similar effect on foreign policy behavior across issues. Yet unlike realist theories that assume similar structural conditions across the international system (Waltz, 1979), democratic peace scholars recognize that domestic institutions influence foreign policies and hence explain variation in the outcomes of dyadic interactions in world politics.
The issue-based approach focuses on the issues that countries contest diplomatically and examines how the characteristics of issues influence the chances of war or peaceful settlement (Mansbach and Vasquez, 1981; Rosenau, 1966). Diplomatic issues involve tangible salience with values related to security, survival, and wealth, and intangible salience values of culture/identity, equality/justice, independence, and status/prestige/influence (Hensel et al., 2008). The issue-based approach predicts that issues with high-tangible and high-intangible salience will be more likely to experience militarized conflict and see more frequent bilateral and third-party efforts to resolve the issues peacefully. This literature has also classified the issues at stake in militarized disputes (MIDs) and wars over the past 500 years, including territorial, nationalist, state creation, regime survival, and resource issues (Holsti, 1991). Territorial disputes have been a focal point of the literature because more wars in history have been fought over this issue than any other (Vasquez, 2009). Where regime type is typically considered in the issue approach is when studying conflict management strategies to see if democracies utilize difference peaceful or militarized tools than non-democracies (Ellis et al., 2010; Huth and Allee, 2002). The issue approach literature typically fails to emphasize variation in the frequency of contested issues over time, even though we have observed significant decreases in the amount of ongoing territorial disputes as the number of democratic states has grown over time.
A few studies have made connections between these important strands of the conflict literature. Institutional theories of the democratic peace note that leaders may have different incentives for the kinds of foreign policy issues they pursue depending on the size of the groups that keep them in power at home (Bueno de Mesquita et al., 2003). Leaders in states with large selectorates and large winning coalitions improve their chances of survival by providing public goods to their constituents. Democratic leaders are more likely to focus on foreign policy issues that have diffuse benefits such as policy differences and the overthrow of foreign regimes, whereas states with smaller winning coalitions are more likely to fight over territory because the capture of territory often results in valuable resources that can be distributed as private goods to members of the winning coalition. Mitchell and Prins (1999) evaluated this hypothesis by considering what issues democracies contend over when democratic dyads experience militarized interstate disputes, comparing fully institutionalized democracies that score 10 on the Polity democracy scale to those that score six to nine on the scale. In two major time periods (1816–1945, 1946–1992), the authors found that fully democratic dyads (both states scoring 10 on the democracy scale) were less likely to have MIDs over territorial issues than dyads with democratizing states. In the Cold War period (1946–1992), the authors examined what issues were at stake in jointly democratic MIDs (N = 97) and they found that a large percentage of these cases involved resources of the sea such as fisheries and oil. However, Mitchell and Prins (1999) looked only at conflicts between two democracies that reached the level of a threat, display, or use of militarized force. We know little about whether democratic contention over maritime areas is widespread among the set of diplomatic conflicts that arise between states.
To consider whether democratic dyads have general propensities to contend over maritime areas diplomatically, we created a dyad-year data set that paired all coastal states in a given region (N = 93,047) and also created dyadic pairs of major powers in the system with each coastal state in a region (N = 32,844). The latter group was included because many major powers become involved in maritime conflicts involving their former colonies (e.g., UK–Belize). We included several regions in our data set: the Western Hemisphere (North, Central, and South America), Europe (Western and Eastern), Asia, and the Middle East, with a temporal domain of 1900 to 2007. These dyads represented our opportunity cases for maritime conflicts, although we also considered politically relevant dyads as an alternate opportunity group (contiguous states up to 450 m water or involving a major power). We utilized data from the Issue Correlates of War (ICOW) project on maritime claims (Hensel et al., 2008; Nemeth et al., 2014) to identify years in which a given pair of states had one or more ongoing diplomatic disagreements over the ownership or usage of a maritime area. Of the 125,891 total dyads that could have engaged in maritime claims, only 5794 dyad years (4.6%) witnessed such conflicts of the sea. Of these maritime claim cases, 35.3% occurred in the Western Hemisphere, 19.6% in Europe, 25.1% in Asia, 5% in the Middle East, and 13.3% involved major powers intervening in another region. A cross-tabulation of dyadic democracy 2 with the presence of a maritime claim showed that democratic dyads (7.09%) have a higher chance of maritime conflicts than non-democratic dyads (4.68%; χ2 = 203.1308, p < .001).
This higher density of maritime conflicts among democratic dyads is somewhat surprising because maritime conflicts are often inherently territorial, and because democratic dyads are less likely to become involved in territorial disputes over land borders or islands than non-democratic dyads (Kacowicz, 1995). 3 Some maritime claims, such as the “Cod Wars,” involve one state (Iceland) claiming a territorial sea area that is not recognized by others (the UK). Ecuador and Peru claimed 200 nm territorial sea limits in the 1940s and 1950s that were disputed by other countries (e.g., United States (US), Canada) seeking access to their rich fishing grounds. Other maritime claims arise due to two or more coastal states having competing claims to areas where their economic exclusive zones overlap. The US and Canada disagreed about the boundary line in the Gulf of Maine, a case that was ultimately adjudicated by the International Court of Justice. Another basis for maritime claims that has territorial roots is a disagreement that can arise when countries claim uninhabited islands as baselines for making EEZ claims. Venezuela has made EEZ claims around Aves Island, while several Caribbean states claim that it is merely a “rock” and not a valid territory for EEZ claims. Ownership of resources in maritime spaces can also become problematic if the resources migrate between two territorial areas. Canada fired at the Estai because Spanish fishermen were following migratory fish stocks from open sea areas into Canadian EEZ spaces. Many disputes over land borders also involve disagreements over maritime areas, such as the Falkland Islands dispute between the UK and Argentina. Both countries dispute ownership of the Islands and they also disagree about ownership and access to oil and fishing resources around the Islands. In short, while democracies have been successful in avoiding disputes over land borders more often than autocracies, this behavior has not translated into a lack of territorial claims in other issues areas. Next we explore three explanations for this pattern of democratic maritime contention.
Democratic contention of maritime issues
If democratic dyads experience more maritime conflicts than other dyadic pairings, what helps account for this pattern? This is important to understand given the generally peaceful nature of democratic interactions, especially since maritime claims have increased over time relative to traditional land border disputes, and given that maritime claims have become militarized more frequently than other types of water issues, such as cross-border river claims (Hensel et al., 2008). While 41% of territorial claims from 1816–2001 experienced one or more militarized disputes, 27.3% of maritime claims were militarized; this conflict rate is much higher than river claims (11.2%) (Hensel and Mitchell, 2017). In short, this could become a potential area where a clash of democratizations could occur even more in the future, and understanding the dynamics behind such patterns will be useful for analyzing maritime conflicts and dynamic patterns of the democratic peace. 4
We focused on three factors that help explain these patterns: (a) higher levels of economic development in democratic countries creating more opportunities for maritime conflicts, (b) maritime security issues and how they may have changed following the terrorist attacks in the US in September 2001, and (c) varying patterns of democratic clustering by regions, which may explain the density of maritime conflicts in particular areas. We also explored whether dyadic regime type has similar effects on the propensity for diplomatic conflicts across regions. We present illustrative cases for each theoretical factor from the ICOW project data set.
Economic opportunities
First, we focus on a fairly simple explanation for why democracies might experience maritime conflicts with other democracies. Some studies report a positive correlation between democracy and higher levels of economic development at the state (Lipset, 1994) and regional (Gleditsch, 2002) levels. Most countries do not delimit their maritime spaces until they have reached a certain level of economic development. In the data we presented by regions earlier, we noted that the Middle East has the lowest frequency of dyad years of maritime conflicts across all regions in our data set. This stems primarily from the lack of maritime boundary delimitations by states in the Middle East region. Countries such as Bahrain, Iraq, Jordan, and Kuwait have not determined their EEZ limits, which lowers the opportunity for maritime conflicts with neighboring states. 5
This dynamic relates to economic development because many maritime claims arise when fishing fleets move into areas where fishermen compete for resources with fishermen from other countries. The Cod Wars stemmed from conflicts between British and Icelandic fishermen in areas around Iceland. The Turbot Wars arose when Canada sought to protect its fishing rights from encroachment by Spanish and Portuguese trawlers. A state has to reach a certain level of development for fishing boats to travel long distances away from their own coastlines and EEZ zones. Many of the largest fishing fleets in the world come from wealthy democratic states, including Japan, Spain, India, the UK, and US, or powerful autocratic countries, such as China and Russia (DeSombre and Barkin, 2011: 30–31). 6
Several cases from the ICOW data set illustrate the relationship between economic development and economic interests on states’ decisions to initiate new claims to maritime zones. Until the mid-1970s, Spanish fishermen dominated fishing waters in and around the Bay of Biscay. This changed in 1976 when France and other members of the European Economic Community (EEC) set up a 200 nm EEZ in the Bay. Spain was not yet a member of the EEC at the time, but signed on to the agreement. Spanish fishermen did not comply with newly established zones, creating conflicts between the Spanish and French governments. The European Common Market, EEC, and the EU have played a role in attempting to quell the dispute, now taking a more direct role in establishing fishing rights in the Bay (LexisNexis–Europolitics, 2011). Spain’s entry into the EEC was controversial with respect to fishing rights because its fishing fleet was much larger than all other EEC states’ fleets (London Times, 1983), creating significant economic competition between European governments and Spanish fishermen. Spain dropped its general objection to the use of 200 nm EEZs in the region as part of its condition for joining for the EEC (London Times, 1983), but continued to experience conflict with nearby states such as France and Ireland due to Spanish fishermen encroaching on other states’ EEZ areas.
Since the mid-1970s, Canada and France have also experienced tensions over competing claims to the continental shelf and EEZ area around St. Pierre and Miquelon. The two sides have captured and searched each other’s fishing vessels in the area. A 1984 agreement calmed tensions by exempting the countries from each other’s fishing laws. In January 1987, Canada offered to allow four years of greatly increased levels of French fishing in the region provided that France would agree to submit the maritime boundary dispute to international arbitration (McDorman, 1990: 158–161).
Accusing the French of exploiting cod resources in March 1987, the Canadian government closed its ports to French vessels. Settlement attempts in September and October 1987 were unsuccessful, and in late October 1987, France announced unilateral quotas for French fishermen that were far beyond those desired by Canada. An arbitration panel’s ruling in June 1992 greatly limited France’s maritime zone around St. Pierre and Miquelon, but left the issue of fishing quotas and continental shelf exploration on the negotiating table. Disagreements over quotas led to the seizure of two French fishing boats in Newfoundland waters in early 1993. France issued a proposal to the UN’s Commission on the Limits of the Continental Shelf claiming a 240 km an expanded offshore economic zone (Canwest News Service, 2009), but Canada does not acknowledge France’s claim (The Globe and Mail, 2009). Competition over and conservation of cod and other marine resources around St. Pierre and Miquelon remains a contentious issue today due to both states’ economic interests.
Sometimes democracies resort to the use of force to protect their economic interests in maritime areas. In 1947, Peru, to be joined later in 1952 by its Latin American neighbors Chile and Ecuador, declared a 200 nm territorial sea limit, claiming exclusive rights to the exploitation of natural resources (especially fish) in the maritime zone. The US rejected the limits and encouraged its fishermen to continue fishing in the contested zone, resulting in heavy fines for US fishing boats captured by Peruvian patrol boats. The maritime dispute between the US and Peru resulted in several dozen US fishing vessels being captured and fined in the 1950s. By the 1960s, Peru’s enforcement of its maritime claims grew more vigorous, giving rise to a series of militarized incidents including a February 1969 attack on US tuna boats by Peruvian gunboats. Throughout the decade, US–Peruvian relations deteriorated, thanks in part to US congressional legislation aimed at Peru, but mostly because of the 1968 coup by Peruvian military officers that installed an anti-American military junta.
In 1979, President Carter’s Freedom of Navigation program included Peru’s 200 nm territorial sea-claim among those maritime zones to be contested by US naval ships, although no US–Peru naval incidents occurred. The last high-profile US–Peru military incident was recorded in 1992 when Peruvian air force jets attacked a US anti-drug flight intruding on its territory. 7 These kinds of clashes were also common in the Cod Wars when British naval vessels would escort British trawlers into claimed Icelandic fishing zones contested by the British government. For countries with large economic stakes in the fishing industry such as Peru, Spain, and Iceland, the development of their fishing fleets can result in a higher propensity for maritime conflicts. This leads to our first primary hypothesis. 8
Maritime security interests
The second factor we examine focuses on maritime security issues, which have arguably become more important in the post-Cold War era (Klare, 2002; Klein, 2011). Maritime areas are often valued for their strategic location and states make diplomatic claims to protect their security interests. Maritime security includes: ensuring the freedom of navigation, the flow of commerce and the protection of ocean resources, as well as securing “the maritime domain from nation-state threats, terrorism, drug trafficking, and other forms of transnational crime, piracy, environmental destruction, and illegal seaborne immigration” (Klein, 2011: 8).
Modern threats to maritime security include piracy, terrorist acts in maritime areas, illicit trafficking of drugs, arms, or people, illegal fishing, and environmental destruction (Klein, 2011: 10). A number of initiatives have been undertaken to deal with these issues, especially as the threat of terrorism on the high seas became more pronounced after the September 2001 terrorist attacks and other prominent port attacks in Yemen and the Philippines.
How do security concerns potentially result in jointly democratic maritime conflicts? As noted above, democracies have more opportunities for maritime claims given their more developed economies with more extensive fishing operations. Yet democracies may also face greater risks of transnational terrorist attacks. Some studies have found that democracies face higher risks of terrorist attacks (Eubank and Weinberg, 1994; Li and Schaub, 2004), while other studies show a negative relationship (Li, 2005; Sandler, 1995). Democratic governments have certainly responded to the perceived increased environment for terrorist attacks and the proliferation of weapons of mass destruction (WMD) (and WMD technology) by creating several mechanisms to protect their maritime areas. Given that 90% of the world’s goods are transported by sea, governments also recognize their economic vulnerability for failing to protect major shipping lanes and ports (Klein, 2011: 150).
The 1985 hijacking of an Italian vessel, the Achille Lauro, by the Palestinian Liberation Organization in 1985 led to the creation of the 1988 Suppression of Unlawful Acts (SUA) Convention, which identified several acts taken against ships as unlawful (e.g., destroying cargo, harming persons on board). The International Maritime Organization (IMO) sought to enhance port security through adoption of the International Ship and Port Facility Security Code (ISPS Code), which came into force in 2004. 9 The US pushed for further regulations in the Container Security Initiative (CSI) and a cargo strategy to protect the global trade network from terrorist attacks. This included 24 hour advanced notice of goods being shipped into US ports prior to the landing of a foreign vessel (Klein, 2011: Chapter 4). The SUA Protocol was also enhanced in 2005 by the IMO to give countries greater rights for boarding potentially threatening vessels, even in areas of the open sea. These general maritime security initiatives were also playing out in interstate diplomatic relations as states sought to protect their security interests through maritime claims. While the perceived increased threat from terrorist attacks has influenced the development of maritime law, countries have also sought to protect security interests closer to home when involved in wars or militarized conflicts with neighboring states.
The lanes in the Straits of Hormuz demonstrate the security threats posed by overlapping ownership of a maritime area. Oman, seeing itself as the regional enforcer of proper shipping lanes, began to build up its navy following the outbreak of the Iran–Iraq War in 1980. At the same time, Iran attempted to restrict the entry and arrival of aid to Iraq via the Strait of Hormuz. Iran stopped commercial vessels in waters previously negotiated as Omani in order to accomplish this task. Oman responded by attempting to prevent Iranian ships and reconnaissance planes from entering their territorial space with the aid of US and British forces. Many of the Iranian searches of commercial vessels were also taking place in Omani territorial waters. According to the envoys, “The Omanis are talking on a daily basis with the Iranian Navy, telling them to keep their business out of Omani waters” (Chicago Tribune, 1985). While the conflict subsided, the Oman government protected its economic interests from security threats in the region.
The Northwest Passage claim also shows how security issues affect maritime issues. The claim arose in 1969 when a US military vessel, the Polar Sea, went through the Passage without obtaining the permission of the Canadian government (although the Canadian government eventually gave the vessel permission to traverse the Passage). Bilateral talks between the two governments from 1985–1987 produced an agreement signed in January 1988, called the Arctic Cooperation Agreement. The US agreed that all navigation by American icebreakers within waters claimed by Canada to be internal would be undertaken with the consent of the Canadian government (Huebert, 1995: 359). While this functional settlement diminished conflict between the two sides, the US never recognized Canada’s claims of sovereignty over the Northwest Passage and hence the claim is ongoing. The US’s interest in passing through the area is primarily for military purposes of moving naval vessels through the region in the winter.
Recent clashes between the Japanese and Chinese governments over the sovereignty of the Senkaku/Diaoyu Islands date back to at least 1879, when Japan annexed Ryukyus and attempted to incorporate the islands’ administration into Okinawa. Japan gained complete control with the annexation of Taiwan during the Sino–Japanese War in 1895. Japan administered the islands until their defeat in World War II in 1945. According to the terms of the Cairo and Potsdam treaties, all territories stolen from China were to be returned. The 1951 Treaty of Peace between Japan and China declared all agreements prior to 1941 between the two nations to be null and void. Japan contends that because the islands came under their control prior to the 1895 treaty that ended the Sino–Japanese War, the Senkaku/Diaoyu Islands are exempt from the treaty. The 1952 Treaty of San Francisco included the US undertaking the administration of Okinawa and included the islands (and they were jointly patrolled by US and Japanese forces). When the US returned Okinawa to Japan, Japan argued that the islands were returned as well.
Prior to 1972, China was not often directly involved in the dispute; instead the dispute was between Taiwan and Japan. After 1972, however, the dispute became primarily one between China and Japan, as relations normalized and China reiterated its claims to Taiwan and the islands. While both governments attempt to downplay the dispute, this has proven difficult as it relates to China’s resolve over all maritime claims and ownership of Taiwan, and to Japan’s resolve of its own alternative maritime claims. The issue has not led to direct military confrontation, but both sides have employed coast guard vessels to protect their maritime claims (Blanchard, 2000; Dzurek, 1996; Pratt and Schofield, 2000). China’s declaration of an Air Defense Identification Zone around the islands in November 2013 also increased tensions.
North and South Korea have also engaged in several threats and uses of military force to protect their maritime security interests. In 1981, North Korea proclaimed a 50 m “military sea boundary” and a 150 nm expansion of economic zones. South Korea refused to recognize the proclamation. The dispute initially stems from the Korean War and the maritime boundaries established after the war ended. While this dispute has been ongoing, the Korean conflict intensified as both sides threatened military retaliation after North Korea launched the first artillery attack on South Korean territory in nearly 60 years in 2010. The Pyongyang regime fired 50 heavy shells onto an island off South Korea’s northwest coast, killing two marines, and setting the homes of 60 civilians on fire. Seventeen marines and three civilians were also injured. The South responded by firing 80 shells into North Korea and scrambling F-16 fighter jets to protect its airspace. In a day of rising tensions, South Korea said its troops would safeguard its territory with “stern responses” if Pyongyang launched any further attacks. But it said it would not “escalate” the conflict on the Peninsula (The Age, 2010).
Another example of security claims can be seen in the conflict between Indonesia and Malaysia in the Celebas Sea. Both countries laid claims to the Sipadan and Ligitan Islands and the Ambalat Block in the Celebas Sea. The International Court of Justice awarded the Sipadan and Ligitan Islands to Malaysia; however, maritime borders remain undetermined in part because of overlapping EEZ claims on the Ambalat Block. Indonesia dispatched three warships to a disputed oil-rich area in the Sulawesi Sea also claimed by Malaysia in a clear sign of strong nationalist stirrings within the armed forces. The show of force was seen as a robust response after Jakarta’s embarrassing loss in the territorial dispute over Sipadan and Ligitan (Straits Times, 2005). Malaysia sent another warship across the sea border near northeastern Borneo in 2009. Indonesia responded by stepping up naval patrols in the area and ordering extra sea power to stand by at the main port of Surabaya. The two states are still negotiating the issue although they have engaged in joint border patrols in the area to combat terrorism, smuggling, and human trafficking.
These examples show how security issues can increase the likelihood of interstate maritime claims. To evaluate this general idea, we look at two factors: (a) a dyad’s history with militarized conflict to capture the general security situation between the two states, and (b) an evaluation of whether maritime claims have increased following 2001 to capture systemic changes in maritime security. Our expectation is that states will be more likely to compete over maritime spaces when security issues are more acute.
Regional contexts
Finally, we consider whether there are significant regional differences in the frequency of maritime claims. We noted that high levels of economic development in a region increase the opportunities for maritime claims. Some of the regions we analyze (e.g., Europe) have much higher development levels than other regions (e.g., Middle East). There are also variations in the number of democratic states in each of our regions for analysis, with Europe and the Western Hemisphere having more democracies than the Asian or Middle Eastern regions. This increases the opportunity set for maritime conflicts if democracies are more apt to engage in maritime claims. We address the potential for regional variation by (a) including dummy variables in our analyses of all regions combined, and (b) estimating separate models for each region. This will give us a sense of whether regime type, economic development, and security issues have similar effects across regions or whether we see varying dynamic patterns.
Research Design
The construction of the dyadic data set (N = 125, 891) was described earlier. We generated the dyad-year data set using Version 3.204 of the EUGene program (Bennett and Stam, 2000). Here we describe the construction of other variables utilized in our multivariate analyses. First, we included two measures for dyadic regime type: the first is a dummy variable for joint democracy that equals one if both states score six or higher on the Polity IV democracy scale (26.5%). The second variable captures mixed dyads, where one state is democratic and one state is not (42.2%). The omitted reference category is autocratic dyads. Our measure of economic development comes from version (4.0) of the COW capabilities data set, which includes data from 1816 to 2007. 10 We measured relative dyadic energy production by taking each country’s energy production estimate, which is based on consumption of coal, petroleum, electricity, and natural gas, and then generated a dyadic measure by dividing the larger energy using state’s value by the sum of the dyadic total. This creates a variable that ranges theoretically from 0.5 (the two states have equal energy production) to 1.0 (the stronger state has all energy production in the dyad). The mean in our data set was 0.892 with a standard deviation of 0.138. Because the ICOW data set extends back to 1900, using the COW data gave us the longest time span coverage for economic data possible, since most data sets with GDP per capita values cover the post-1945 period. We also reported models with the energy scores entered monadically for each state to test if higher economic development levels increased states’ maritime claims. We included a dummy variable for whether the dyad contains a major power (39%), an important control given that we pair major powers from other regions with coastal states.
Security environments were captured by two variables. First, we included a dummy variable that indicated if the two states had experienced a militarized interstate dispute in the current dyad-year (1.54%). Second, we included a dummy variable for the years 2001 through 2007 to capture the systemic shift in maritime law that occurred following the September 2001 terrorist attacks (16.2%). We expected both factors to increase the frequency of dyadic maritime claims.
Regional differences were captured by including dummy variables for each region (the Americas, Europe, Asia, and the Middle East) as coded by the COW project. The omitted category captured major powers paired with states from other regions. To capture potential interdependencies across time and space in our data, we included measures for year and year squared (Carter and Signorino, 2010). 11 We also included a dummy variable that equaled one in 1994 and later years to examine the effects of the UNCLOS treaty coming into force. We estimated this variable separately from the post-September 2001 variable, given the perfect overlap after 2001. Because the UNCLOS treaty creates explicit guidelines for limits to states’ claims to territorial sea, contiguous zone, or EEZ areas, states that have jointly ratified the convention should be less likely to challenge each other’s maritime claims. We also used interaction terms with these interventions and the democratic-dyad variable to see whether the effect of democracy on maritime claims varied over time. Given that our dependent variable (the presence of a maritime claim) was dyadic (4.6%), all models were estimated as logit models.
Empirical analyses
We begin with an analysis of maritime conflicts in the set of opportunity dyads based on coastal states in regions and major powers. Model 1 in Table 1 shows the effects of regime type, economic, security, and regional variables on maritime claim occurrence in all regions from 1900–2007. We calculated substantive effects using Clarify (Tomz et al., 2003). First, we see that the bivariate finding comparing democratic and non-democratic dyads holds in the multivariate models when we also include mixed dyads. Democratic dyads are significantly more likely to experience maritime claims than autocratic dyads, while mixed dyads are less likely to experience claims than autocratic dyads. The typical dyad in our data set has a 0.026 (0.024–0.029) chance of starting a new maritime claim in a given year. 12 Joint democracies’ risk of maritime disputes is 46% higher (0.038 (0.034–0.041)) than autocratic dyads’ chances of conflict. When looking at the risks in different regions, we found that pairs of democracies are most likely to experience maritime claims in the Americas (0.051 (0.047–0.055)) and least likely to experience conflict in Europe (0.009 (0.008–0.010)). Mixed dyads face a 19% lower risk of maritime claims than autocratic dyads have (0.021 (0.019–0.023)).
Multivariate analyses of maritime claim occurrence (1900–2007).
Entries are coefficients followed by standard errors: * p < .05, ** p < .01.
Hypothesis 1 postulates that states with higher levels of economic development are more prone to maritime conflict. The results in Table 1 (Models 1 and 2) demonstrate support for the hypothesis and show that relative economic size is also important. Dyads with asymmetric energy production levels are much less likely to experience maritime claims than countries that are more similar in energy production levels. The most symmetric energy dyads have a 225% greater risk of maritime claims (0.065 (0.057–0.074)) than the most asymmetric energy dyads (0.020 (0.018–0.022)). In Table 1, Model 2, we added separate measures for each state’s energy production and we see that the effect for relative energy production holds, while only one state’s (A) energy production is statistically significant and positive as predicted. 13 Increasing state A’s energy production from the minimum to the maximum value increases the chances of maritime claims by 343%. 14 Having one or two major powers in a dyad significantly increases the chances of maritime claims (by 94% in Table 1, Model 1). In short, capabilities and relative power influence the chances of maritime conflicts, consistent with our theory that more developed states have greater opportunities for maritime claims. 15
Hypothesis 2 predicts that states that face higher dyadic or systemic security threats are more likely to experience maritime disputes. We found support for this hypothesis (Table 1, all models) because pairs of countries that experience militarized disputes are much more likely to have maritime claims than dyads without militarized disputes. This security factor produced the largest substantive change in the predicted probability of maritime claims, raising the risk by 577% above the baseline (0.176 (0.155–0.200)). Table 1, Model 1 includes a dummy variable to capture the post-September 2001 environment, yet it is not statistically significant. On the other hand, the years since UNCLOS came into force in 1994 have witnessed a significant drop in the chances of maritime claims (Table 1, Model 4), showing the positive influence this multilateral institution has had on the management of oceanic resources (see also Nemeth et al., 2014). Prior to 1994, the dyadic risk of maritime claims (0.028 (0.026–0.031)) was 40% higher than in the post-1993 period (0.020 (0.018–0.024)). That dyads with militarized disputes and major powers are more likely to experience maritime claims is consistent with the general idea that states in more acute security environments are more likely to contest maritime spaces.
Our third theoretical factor focused on differences that may arise regionally due to distinct levels of economic development and democracy. We see in Table 1 that all of the regional variables are positive and statistically significant, showing us that dyads in the same region are more likely to experience maritime claims than major powers paired with states in other regions. As noted earlier, the Americas (0.051 (0.047–0.055)) have the highest risk of new maritime claims among the regions we analyzed, followed by Asia (0.039 (0.035–0.043)), the Middle East (0.024 (0.020-0.028)), and Europe (0.009 (0.008–0.010)). Table 2 presents models where we estimated sub-samples for each region and the major power sample. First, we see that the positive relationship between joint democracy and maritime claims is being driven by Europe and by the major power sample. This makes sense in terms of these two sub-groups having the highest overall levels of economic development. Thus while European states have lower overall risks of maritime conflicts (baseline probability in Table 2, Model 2 is 0.006 (0.005–0.008)), pairs of democracies are 300% (0.024 (0.021–0.028)) more likely to experience maritime disputes with European neighbors than autocratic dyads in the region. Interestingly, joint democracy (0.068 (0.059–0.078)) has a negative and statistically significant effect on maritime claim occurrence in Asia (Model 3), reducing claim risk by 240% compared with autocratic dyads (0.020 (0.015–0.025)). Mixed dyads have varying effects across regions, increasing the chances of maritime conflict in Europe and the major power sample, and decreasing conflict in the Americas, the Middle East, and Asia. These differences reflect the varying degrees of democracy that exist across these regions at various points in time. Asymmetric energy production discourages maritime claims in all regions except for Europe (Table 2, Model 2).
Regional analyses of maritime claim occurrence (1900–2007).
Entries are coefficients followed by standard errors: * p < .05, ** p < .01.
Rel. Energy Prod. = Relative Energy Production.
States’ involvement in a militarized dispute in a dyad-year increases the chances of maritime claims in all groups. This is consistent with other studies that show linkages across contested diplomatic issues, with militarization of early territorial issues heightening risks of maritime or river claims at a later point in time (Dreyer, 2010; Mitchell and Thies, 2011). The post-2001 variable was not significant in any of the five models. We observed that the control variables for year and year squared changed signs in the Asian and Middle Eastern regions, showing different temporal hazard rates (U-shaped: Middle East) for maritime claims compared with the other regions (inverted-U: the Americas, Europe). This could reflect in part that many democratic countries in the Western Hemisphere and Europe have signed and ratified UNCLOS, which has lowered the likelihood of new maritime conflicts in more recent years.
Finally, we created an interaction between the post-2001 period and the years that UNCLOS has been in force with the democratic-dyad variable to see how the behavior of democracies may have changed over time. We theorized that maritime security issues have become more acute over time, especially for democracies, since they have regularly been targets of terrorist attacks in ports and open sea areas. We see in Table 3, Model 1 that the effect of UNCLOS is negative and significantly different from zero when interacted with democracy, suggesting that the propensity for democracies to engage in maritime conflicts has been diminished by ratification of the UNCLOS convention. Joint democracies reduce maritime conflict risk (0.058 (0.051–0.064)) by 81% if both states have ratified UNCLOS (0.032 (0.027–0.036)). In Model 2, we interacted joint democracy with the post-2001 years. Neither the post-2001 variable nor the interaction term was statistically significant. Thus it appears that the presence of UNCLOS has been more significant than the post-September 2001 environment for changing the likelihood of maritime conflicts arising between states.
Interactive effects for democratic dyads.
Entries are coefficients followed by standard errors: * p < .05, ** p < .01.
Conclusion
This study sought to explain a puzzling positive correlation between joint democracy and the occurrence of maritime conflicts. While the democratic peace process has reduced the chances of traditional land border disputes, democracies have not avoided other forms of territorial conflict with their democratic peers, such as contestation of maritime areas that overlap with other states’ claimed zones. Maritime conflicts have increased in frequency over the past several decades and the fact that such disputes often lead to threats or use of military force gives us good reason to understand whether these conflicts represent a potential threat to the democratic peace. We generated a set of dyads that could potentially experience maritime conflicts and then examined how regime type, economic, security, and regional factors influence the occurrence of diplomatic disagreements over the ownership or usage of maritime areas.
Our results confirmed earlier studies that found democracies to have conflict over maritime issues regularly (Mitchell and Prins, 1999). We showed that democratic dyads were more likely than mixed or autocratic dyads to experience diplomatic maritime claims. We also found that economic capabilities influenced the chances of maritime conflicts, with major powers and highly developed states being more active in pursuing claims, and asymmetric economic ties depressing the chances of conflict. We also theorized that threatening security environments increases states’ aggressive tendencies to compete for maritime resources, especially in the aftermath of the September 2001 terrorist attacks when states and international organizations sought to enhance security around ports and strategic choke points. Maritime claims often arise when militarized conflicts and wars affect states’ abilities to extract resources and move in maritime spaces. Our large-N analyses confirmed this pattern, whereby the experience of militarized conflict significantly increased the chances of diplomatic disagreements over maritime areas. However, overall dyadic risks of maritime claims did not increase in the 2001–2007 period, despite a series of global efforts to increase maritime security.
We also found that regional contexts explained some of the variation in maritime conflicts around the globe. Jointly democratic conflicts have occurred frequently in democratic regions such as Europe and the Americas and in dyads with democratic major powers. Interestingly, joint democracy had a depressing effect on maritime conflicts in Asia, although we will need to do more research on the frequency of such dyads relative to the whole to see whether these findings are artifacts of a small group of dyads. We also observed that the Americas had the highest rate of maritime conflict in general, which is also consistent with the pattern of more democratic regions experiencing maritime conflict frequently.
Finally, we looked at the presence of the UNCLOS and its entry into force in 1994. We found that UNCLOS significantly reduced the chances of new maritime conflicts and also helped resolve some claims that had existed prior to 1994. Pairs of democracies in the UNCLOS era are less likely to experience maritime claims, thus this institution has been effective at reducing the chances of conflict among the highest risk group. Future analyses may examine commitments at the dyadic level since other studies suggest that joint UNCLOS membership in a dyad reduces the chances of maritime claim onset (Nemeth et al., 2014). We could also explore the role of the IMO in greater detail to see how it may have helped reduce conflicts and set standards for maritime claims. That international organizations have played an important role in the management of maritime resources shows how the democratic peace is part of the larger Kantian peace, whereby economic interdependence and international organizations are important factors in preserving peace between democracies.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest
The authors have no conflict of interest to declare.